Niki Zacharias^1,2, Napapon Sailasuta³, Henry Chan³, Meng Wei³, Robert W Grubbs¹, Brian D Ross³, and Pratip Bhattacharya^3
1California Institute of Technology, Pasadena, CA, United States, ²Enhanced Magnetic Resonance Laboratory, Huntington Medical Research Institutes, Pasadena, CA, United States, ³Enhanced Megnetic Resonance Laboratory, Huntington Medical Research Institutes

All the molecules we have hyperpolarized using parahydrogen induced polarization (PHIP) and have employed for rodent in vivo applications have physiological barriers for clinical utility. We have recently hyperpolarized diethyl succinate using PHIP. Using tail vein injections of hyperpolarized diethyl-3-¹³C-4,5-d₂-succinate in mice, we have been able to achieve ultrafast ¹³C MRI, MRS, and successfully interrogate the TCA cycle in real time. Diethyl succinate is quickly taken up by cells, is water soluble, can be hyperpolarized at neutral pH, and is known to be nontoxic. It’s a good candidate for future in vivo metabolic imaging in humans.

Kayvan R. Keshari¹, Robert Bok², Subramaniam Sukumar², Mark Van Criekinge², Daniel Vigneron², and John Kurhanewicz^2
1UCSF, San Francisco, CA, United States, ²UCSF

In this study a 14T multi-parametric (T2-MRI, DWI, HP-13C MR) imaging protocol was used to compare MR findings with histopathology and immunohistochemistry of normal and TRAMP mice at various stages of cancer progression. It was determined that, similar to the human situation, ADC was reduced in TRAMP tumors relative to normal prostate. Both increased proliferation and hypoxia were related to increased uptake of pyruvate and lactate to pyruvate ratio. Late stage TRAMP tumors had overall increased perfusion relative to the normal prostate. However, perfusion was heterogeneous and there were regions of low perfusion in the tumor with high metabolic activity.

Rosa Tamara Branca¹, Simone Degan², John Nouls³, Challa Kumar⁴, Sanchita Biswas⁴, and Bastiaan Driehuys^3
1Chemistry, Duke University, Durham, North Carolina, United States, ²Chemistry, Duke University, Durham, North Carolina, ³Center for In Vivo Microscopy, Duke University, ⁴Center for Advanced Microstructures and Devices, Louisiana State University, United States

The recently introduced combination of hyperpolarized (HP) gas MRI with targeted superparamagnetic iron oxide nanoparticles (SPIONs) could open new avenues for MR-based molecular imaging of the lung. However, because the lungs’ fine capillary network readily traps particles larger than ∼10μm, it is critically important to develop stable, monodisperse SPION formulations that do not accumulate unless the pathology of interest is present. The current study was designed to evaluate the in vivo bio-distribution of several SPION formulations injected intraveinously, and to test their effects on 3He MRI in normal control mice.

Katherine Louise Parcell^1,2, Tammy Louise Kalber^1,2, Paul Southern³, Quentin A Pankhurst³, Sam M Janes², and Mark F Lythgoe^1
1UCL Centre for Advanced Biomedical Imaging, Division of Medicine and Institute of Child Health, University College London, London, United Kingdom, ²Centre for Respiratory Research, University College London, London, United Kingdom, ³Davy-Faraday Research Laboratories, The Royal Institution of Great Britain, London, United Kingdom

Superparamagnetic iron oxide nanoparticles (SPIO) have diverse applications in biomedicine, and have been used for cell tracking of mesenchymal stem cells (MSCs) within the body using MRI. SPIO labelled MSCs will home to and incorporate themselves within lung metastases in vivo, carrying SPIO particles into tumours. The application of an alternating magnetic field causes rapid heating of the SPIO, damaging the surrounding cancer cells. In this study we investigate the potential of SPIO as bifunctional nanoparticles for hyperthermia treatment in a subcutaneous murine tumour model and the possibility of MR monitoring of SPIO labelled MSCs within the tumour.

Simon Walker-Samuel¹, Peter Johnson², Barbara Pedley², Mark F Lythgoe*¹, and Xavier Golay*^3
1UCL Centre for Advanced Biomedical Imaging, Department of Medicine and Institute of Child Health, University College London, London, United Kingdom, ²Institute of Cancer, University College London, United Kingdom, ³Institute of Neurology, University College London, United Kingdom

Chemical exchange saturation transfer (CEST) imaging has been used experimentally in a broad range of applications. However, full quantification of CEST effects in vivo using standard imaging sequences is time consuming as a large number of saturation frequency offsets are required to define a z-spectrum. Furthermore, outside the brain, the presence of fat can confound the interpretation of z-spectra. A novel acquisition and post-processing technique is presented in this study, named exchange-modulated point resolved spectroscopy (EXPRESS), which aims to address these limitations and enables spatially localised, high signal-to-noise measurements of CEST effects in vivo.

Nai-Wei Yao^1,2, Chiao-Chi V Chen^1,3, and Chen Chang^1,3
1Functional and Micro-magnetic Resonance Imaging Center, Academic Sinica, Taipei, Taiwan, ²Institute of Zoology, National Taiwan University, Taipei, Taiwan, ³Institute of Biomedical Sciences, Academia Sinica, Taipei, Taiwan

The targeted migration of neural progenitor cells (NPCs) towards brain injuries is mediated by chemokines, such as stromal cell¡Vderived factor-1 (SDF-1). It is thus hypothesize that the transplantation of NPCs with an elevated SDF-1 gradient leads to stronger migration to the tumor site, which may cause a different growing pattern and morphology of the tumor. The present study demonstrates that combinative treatment of NPC with SDF-1 rendered tumors exhibit rapid growth with sever hemorrhage. The finding identified an important role of NPCs facilitated by SDF-1 in tumor expansion.

Yann Jamin¹, Jessica KR Boult¹, Jeffrey C Bamber¹, Ralph Sinkus², and Simon P Robinson^1
1CRUK and EPSRC Cancer Imaging Centre, The Institute of Cancer Research and Royal Marsden NHS Trust, Sutton, United Kingdom, ²INSERM U773, CRB3, Centre de Recherches Biomédicales Bichat-Beaujon, Paris, France

In this proof-of-principle study, we demonstrate that magnetic resonance elastography (MRE) affords noninvasive measurements of the visco-elastic properties of orthotopic murine gliomas. We have specifically assessed the visco-elastic properties of orthotopically propagated RG2 gliomas in mouse brain, which are a faithful representation of the most common primary brain tumour, astrocytoma. MRE revealed that RG2 gliomas were softer and less viscous than the surrounding brain tissue, which is consistent with the described appearance of the RG2 glioma model and astrocytomas in the clinic.

Yang Guo¹, Yue Zhang^2,3, Ning Jin^2,4, Jodi Nicolai², Rachel Klein², Guang-Yu Yang⁵, Reed Omary^2,4, and Andrew Larson^2,4
1Radiology, Northwestern University, Chicago, IL, United States, ²Department of Radiology, Northwestern University, Chicago, IL, United States, ³Biomedical Engineering, University of Illinois at Chicago, Chicago, IL, United States, ⁴Department of Biomedical Engineering, Northwestern University, Chicago, IL, United States, ⁵Department of Pathology, Northwestern University, Chicago, IL, United States

Electropermeabilization involves application of electrical pulses to increase cell membrane permeability; electrochemotherapy (ECT) takes advantage of this phenomenon to increase tumor uptake of chemotherapeutic drugs. The timing of drug infusion and application of electrical pulses is critical to optimize the ECT procedure. The purpose of this study in rat hepatocellular carcinoma (HCC) model was to demonstrate the potential to use dynamic contrast-enhanced MRI for intra-procedural optimization and monitoring during intra-hepatic ECT.We found that ECT treated tumors showed reduction in tumor perfusion and a significant increase of cisplatin uptake compared to the internal control tumors. Our findings suggest that ECT permits superior chemotherapeutic drug uptake within targeted HCC compared to conventional chemotherapy in the N1-S1 rodent model. Dynamic contrast enhance MRI shows the potential to allow patient-specific adjustments to ECT timing parameters for optimal drug delivery to targeted hepatic tumors and monitoring of electroporation related vascular lock perfusion changes.

Lu Jiang¹, Kamila Chughtai², Dmitri Artemov¹, Paul Winnard Jr.¹, Venu Raman¹, Zaver Bhujwalla¹, Ron Heeren², and Kristine Glunde^1
1ICMIC, Russell H. Morgan Department of Radiology and Radiological Science, Johns Hopkins Medical School, Baltimore, MD, United States, ²FOM-Institute for Atomic and Molecular Physics, Amsterdam, Netherlands

The intensity of the lipid signal in both MR spectroscopic images and mass spectrometric images of tumors is spatially heterogeneous. The likewise heterogeneous physiologic tumor microenvironment, characterized by regional hypoxia, may contribute to this heterogeneity. We therefore investigated the relationship between hypoxia and lipid metabolites in a human breast cancer model. To this end, we combined in vivo magnetic resonance imaging (MRI) and magnetic resonance spectroscopic imaging (MRSI) with ex vivo mass spectrometric imaging (MSI) and optical imaging of hypoxia and necrosis.

Ellen Ackerstaff¹, Makiko Suehiro¹, Natalia Kruchevsky¹, Sean Carlin¹, Eric H. Rosenfeld¹, Paul Burgman¹, Guangbin Yang¹, Geralda Torchon¹, Ouathek Ouerfelli¹, Pat B. Zanzonico¹, Kristen L. Zakian¹, Clifton C Ling¹, and Jason A. Koutcher^1
1Memorial Sloan-Kettering Cancer Center, New York, NY, United States

Tumor hypoxia is related to treatment response and outcome. We evaluated the ability of trifluoromisonidazole (TFmiso) to detect hypoxia in vivo and in vitro. In vivo, TFmiso could reproducibly measure whole-tumor hypoxia and the effect of oxygenation changes (carbogen, oxygen breathing) on tumor hypoxia. Its accumulation in hypoxic tumor areas was validated ex vivo using ¹⁸F-TFmiso autoradiography and pimonidazole immunohistochemistry. In vitro, ¹⁸F-TFmiso is taken up preferentially at [O₂]<1% with the highest uptake in anoxic tumor cells and <50% of cellular TFmiso is protein bound. TFmiso imaging may be useful in identifying tumors that can be successfully reoxygenated and radiosensitized.

Heather H. Cornnell¹, Ihor Luhach¹, Gary Martinez¹, Arig Ibrahim Hashim¹, Robert A Gatenby², and Robert J Gillies^1
1Department of Imaging, Moffitt Cancer Center and Research Instutite, Tampa, FL, United States, ²Department of Radiology, Moffitt Cancer Center and Research Instutite, Tampa, FL, United States

Measurement of pH using MRI has the potential to help aide in disease diagnosis and evaluation; notably in cancers, which have been shown to have an acidic microenvironment. In small animal models, IEPA (2-imidazole-1-yl-3-ethoxycarbonylpropionic acid) has been used as a MRSI agent to noninvasively map tumor pH. However, IEPA has a pKa of ~ 7 and we have recently shown that chronic oral administration of IEPA can inhibit spontaneous metastases. These results were accompanied by an increase in tumor and urine pH, which illustrates that IEPA has the potential to alter the pH of the tissue being measured by MRSI.

Simon Walker-Samuel¹, Peter Johnson², Barbara Pedley², Mark F Lythgoe*¹, and Xavier Golay*^3
1UCL Centre for Advanced Biomedical Imaging, Department of Medicine and Institute of Child Health, University College London, London, United Kingdom, ²Institute of Cancer, University College London, United Kingdom, ³Institute of Neurology, University College London, United Kingdom

Tumours typically rely on glycolytic metabolism rather than oxidative phosphorylation and, as such, their rate of glucose uptake and metabolism is generally greater than most other tissues. It has previously been shown that chemical exchange saturation transfer (CEST) can be used to detect glycogen in liver, via the selective saturation of exchangeable protons in –OH groups. In this study, the ability of gluco-CEST to detect exogenously administered glucose in subcutaneous tumour xenograft models was evaluated, using two acquisition sequences. A significant, spatially heterogeneous enhancement was observed in all but one tumour.

Lu Jiang¹, Dmitri Artemov¹, Paul Winnard Jr.¹, Venu Raman¹, Zaver Bhujwalla¹, and Kristine Glunde^1
1ICMIC, Russell H. Morgan Department of Radiology and Radiological Science, Johns Hopkins Medical School, Baltimore, MD, United States

Tumor hypoxia triggers signaling cascades that significantly impact on biological outcomes resulting in resistance to radio- and chemotherapy. Therefore, understanding the hypoxic response of tumors is critical. In this study, we have investigated the relationship between hypoxia, necrosis and several metabolites in a human breast cancer model by combining in vivo magnetic resonance imaging (MRI), magnetic resonance spectroscopic imaging (MRSI) with optical imaging of hypoxia and necrosis ex vivo. This multi-modal molecular imaging technology proved useful for delineating the effects of tumor hypoxia and necrosis on MRS-detectable metabolites, some of which may serve as markers for hypoxia and/or necrosis.

Marie-France Penet¹, Paul T. Winnard Jr.¹, Radharani Marik¹, Sridhar Nimmagadda¹, Martin G. Pomper¹, and Zaver M. Bhujwalla^1
1JHU ICMIC Program, Russell H. Morgan Department of Radiology and Radiological Science, The Johns Hopkins University School of Medicine, Baltimore, MD, United States

Cancer-induced cachexia is an under explored problem. There are no known cures for this condition and its multi-faceted nature makes it difficult to investigate. Multimodality imaging approaches are ideally suited to understand critical pathways in cachexia. MRS studies revealed differences in the choline metabolism of cachectic tumor and positron emission tomography experiments demonstrated a higher glucose uptake in cachectic tumors. We are developing an optical reporter to detect the onset of cachectic signals following tumor inoculation. These studies will provide further insight into the ‘cachectic phenotype’, which will be used to define new targets and improve treatment efficacy.

Peter Edward Thelwall¹, Nicholas E Simpson², Zahid N Rabbani³, Daniel Clark², Roxana Pourdeyhimi⁴, Jeffrey M Macdonald⁴, Stephen J Blackband², and Michael P Gamcsik^4
1Newcastle Magnetic Resonance Centre, Newcastle University, Newcastle upon Tyne, Tyne and Wear, United Kingdom, ²University of Florida, United States, ³Duke University Medical Centre, United States, ⁴University of North Carolina / NC State University, United States

We performed 13C MR spectroscopy and imaging of in vivo rat tumour models to non-invasively monitor the metabolism of [2- ¹³C]-glycine into glutathione. Glutathione is an important intracellular antioxidant that plays a role in tumor response to therapy and survival. We observed lower glutathione concentrations and synthesis rates than in our previous studies on fibrosarcoma tumour models. We also observed the metabolic fate of ¹³C label from glycine into serine and cysteine, providing an insight into serine hydroymethyltransferase, glycine cleavage and transulfuration pathways. Our studies demonstrate the use of MRI and MRS for monitoring metabolic processes central to oxidative stress defence.

Samata M Kakkad¹, Marie-France Penet¹, Arvind Pathak¹, Meiyappan Solaiyappan¹, Venu Raman¹, Kristine Glunde¹, and Zaver M Bhujwalla^1
1JHU ICMIC Program, Russell H. Morgan Department of Radiology and Radiological Science, Johns Hopkins University School of Medicine, Baltimore, MD, United States

Solid tumors are characterized by hypoxic environments. Hypoxia stimulates the gene expression of a cluster of hydroxylases used for collagen I fiber formation. Hypoxic environments in tumors may lead to abnormal collagen deposits either by cancer cells or by fibroblasts within the tumor stroma. In healthy tissue, collagen fibers direct interstitial fluid into lymphatic channels. In tumors, these fibers may not be structured for efficient flow of fluid, especially in hypoxic areas. Our purpose is to understand the role of hypoxia in modifying macromolecular fluid transport using MRI, and collagen I fiber distribution using second harmonic generation microscopy.

Xiaoen Wang¹, Liang Zhang², Michael P Collins³, Brittany Bahamon³, Sabina Signoretti³, Michael B Atkins², David C Alsop¹, and Rupal S Bhatt^2
1Department of Radiology, Beth Israel Deaconess Medical Center, Harvard Medical School, Boston, Massachusetts, United States, ²Division of Hematology-Oncology and Cancer Biology, Beth Israel Deaconess Medical Center, Boston, Massachusetts, United States, ³Department of Pathology, Brigham and Women's Hospital, Harvard Medical School, Boston, Massachusetts, United States

This study investigates the relationship between ASL-MRI assessed tumor perfusion and histologic microvessel density (MVD) before and after antiangiogenic therapy in a renal cell carcinoma mouse model. The average tumor perfusion was 67.9 ± 28.2 ml/100g/min (n=27), MVD was 159.9 ± 87.7 vessels/mm2. The correlation between ASL tumor perfusion and MVD was very good (r=0.92). Tumor perfusion was significantly lower in tumors treated with antiangiogenic therapy as compared with that of untreated tumors (49.3 ± 15.0 ml/100g/min vs 102.6 ± 6.0 ml/100g/min, p<0.001). ASL-MRI may serve as an accurate and noninvasive method for mornitoring tumor angiogenesis and response to treatment.

Jessica K.R. Boult¹, Lara Perryman², Gary Box³, Chris Jones², Suzanne A Eccles³, and Simon P Robinson^1
1Cancer Research UK and EPSRC Cancer Imaging Centre, The Institute of Cancer Research and Royal Marsden NHS Trust, Sutton, Surrey, United Kingdom, ²Paediatric Oncology, The Institute of Cancer Research, Sutton, Surrey, United Kingdom, ³Cancer Research UK Cancer Therapeutics Unit, The Institute of Cancer Research, Sutton, Surrey, United Kingdom

Delineation of infiltrative gliomas, in which the blood brain barrier remains intact, using conventional Gd-DTPA-enhanced MRI can be problematic. The use of intravascular USPIO contrast agents may facilitate the detection of such infiltrative areas. In this study, quantitative Gd-DTPA and USPIO-enhanced MRI was performed in orthotopically propagated RG2 gliomas and the spatial relationship compared. Tumours were highly heterogeneous and a displayed a substantially higher blood volume than the surrounding brain. An association between ÄR1 with USPIO and fractional blood volume established with USPIO was observed. Regions of infiltration identified histologically presented with a relatively low blood volume and low ÄR1.

Leanne Bell¹, Davina Honess², Dominick McIntyre², and David Tuveson^2
1CRUK Cambridge Research Institute, Cambridge, United Kingdom, ²CRUK Cambridge Research Institute, United Kingdom

This preclinical study uses dynamic contrast-enhanced MRI to investigate the comparative perfusion of spontaneous and transplanted pancreatic tumours in genetically engineered mice. Spontaneous tumours, accurately recapitulating human disease, have substantial desmoplastic stroma while transplanted tumours of the same genomic background have virtually none. The study demonstrates that the Gd-DTPA volume transfer constant falls as both tumour types grow from 100–600 mm3, but in spontaneous tumours it is consistently approximately half that measured in transplantable tumours of comparable size. This is consistent with relatively poor drug delivery in tumours with a high stromal content, contributing to their typical drug resistance.

Xiaobing Fan¹, Sanaz A Jansen², Erica J Markiewicz¹, Gillian M Newstead¹, and Gregory S Karczmar^1
1Radiology, The University of Chicago, Chicago, IL, United States, ²Mouse Cancer Genetics Program, National Cancer Institute, Frederick, MD, United States

Clinical DCE-MRI is less sensitive to ductal carcinoma in situ (DCIS) than invasive breast cancer. This could be because region-of-interest (ROI) analysis is generally used in clinical practice to improve signal-to-noise ratio. Here we performed pixel-by-pixel analysis of contrast media uptake by pure intraductal cancers in a mouse model of breast cancer and compared this approach with ROI-based analysis. DEC-MRI data were acquired at 4.7T with temporal resolution of ~4 s. Contrast uptake by intraductal cancers was heterogeneous compared to lymph nodes and there was a significant difference between pixel-based analysis and whole ROI analysis.

Stefan Zwick¹, Wilfried Reichardt¹, Claudia Weidensteiner¹, and Dominik von Elverfeldt^1
1Dept. of Radiology, Medical Physics, University Medical Center, Freiburg, Germany

Aim of this work (supported by the 2010 ISMRM Seed-Grant) was to implement and evaluate a spin echo gradient echo sequence to realize pharmacokinetic modeling and VSI simultaneously, both based on R2 and R2* measurements. The sequence with EPI readouts was tested on phantoms consisting of several compartments with different CA concentrations. We evaluated the accuracy of our sequence to quantify the differences in R2 and R2* and compared the results to R2 and R2* measurements with standard sequences. The sequence is able to quantify the changes in relaxation rates and thus should enable pharmacokinetic modeling and VSI simultaneously.

Daisuke Kokuryo¹, Ichio Aoki¹, and Tsuneo Saga^1
1Molecular Imaging Center, National Institute of Radiological Sciences, Chiba, Chiba, Japan

Ultra-short echo-time (UTE) imaging promises to be a powerful tool for lung parenchyma imaging. In this paper, signal changes in lung parenchyma and metastatic tumor induced by a positive contrast agent were evaluated using 3D UTE imaging for several flip angles. The signal intensity of UTE images in lung increased after Gd-DTPA administration. The enhancement was larger for FAs  10°. The signal intensity of UTE images in tumor was also enhanced by Gd-DTPA. The results will be useful for optimizing chemotherapy with tumor targeting nanoparticles.

Kai Henrik Barck¹, Anthony Lima², Tim Cao¹, Rafael Molina², William F Forrest³, Weilan Ye⁴, Leisa Johnson², and Richard A. D. Carano^1
1Biomedical Imaging, Genentech, South San Francisco, CA, United States, ²Molecular Biology, Genentech, South San Francisco, CA, United States, ³Biostatistics, Genentech, South San Francisco, CA, United States, ⁴Tumor Biology and Angiogenesis, Genentech, South San Francisco, CA, United States

Our goal was to evaluate the vascular effects of anti-VEGF-A treatment in a Genetically Engineered Mouse Model of Non-small Cell Lung Cancer with respiratory-gated DCE-MRI. Small (Gd-DTPA) and intermediate (Gadomer 17) molecular weight contrast agents were compared. With Gadomer 17, both Ktrans (mean±SEM: -45±8% vs. 7±9%, p<0.001) and IAUC180 (-29±5% vs. 4±6%, p<0.001) decreased significantly in mice treated with anti-VEGF compared to control, but not with Gd-DTPA. This strong anti-VEGF response is consistent with drug efficacy in this model. Gadomer 17 was found to be more suitable for DCE-MRI in this model, likely due to its slower extravasation.

Marte Thuen¹, Jien Wang², Per Øyvind Enger², Aurelie Poli^3,4, Guro Løkka², Else Marie Huuse¹, Frits Thorsen², Cecilie Brekke Rygh², and Martha Chekenya^2
1Dep of Circulation and Medical Imaging, Norwegian University of Science and Technology, Trondheim, Norway, ²Department of Biomedicine, University of Bergen, Bergen, Norway,³Translational cancer research, University of Bergen, Norway, ⁴Lab for immunology and allergology, CRP sante, Luxembourg, Luxembourg

Glioblastoma multiform (GBM) is a highly aggressive brain tumor. In this study we target NG2-receptor (a marker for aggressive development) with monoclonal antibodies (mAb) and natural killer cells (NK). Rats were implanted with GBM cells and treated with mAb, NK and NK+mAb. MRI (DCE-MRI) and diffusion weighted MRI was performed 7 and 17 days after treatment. Overall survival and histological analysis illustrated that the combined treatment with NK+mAb had greatest therapeutic effect. Volume of extra-vascular cellular space was significantly reduced in rats treated with NK compared to control, due to the present of NK-cells in addition to tumor cells. The NK+mAb combined treatment was not different from controls, indicating a reduction of tumor cells due to the effect of the treatment.

Kimberly R Pechman^1,2, Deborah L Donohoe^2,3, Devyani Bedekar^2,3, and Kathleen M Schmainda^2,4
1Neurosurgery, Medical College of Wisconsin, Milwaukee, WI, United States, ²Translational Brain Tumor Research Program, Medical College of Wisconsin, Milwaukee, WI, United States,³Radiology, Medical College of Wisconsin, Milwaukee, WI, United States, ⁴Radiology and Biophysics, Medical College of Wisconsin, Milwaukee, WI, United States

Few systematic studies examined the influence of timing on efficacy of anti-angiogenic therapy for brain tumors. MRI measures of enhancing tumor volume are unreliable indicators of response since anti-angiogenic drugs decrease contrast enhancement. The purpose of this study was to evaluate rCBV, derived from DSC imaging, and ADC to determine whether the treatment effect is altered when given in later stages of tumor growth. The studies, performed in the U87 xenograft brain tumor model, demonstrate treating tumors earlier provides greater tumor inhibition but treating later maximally decreases rCBV. Each marker is sensitive to different or complementary aspects of tumor response.

Kimberly R Pechman^1,2, Deborah L Donohoe^2,3, Devyani P Bedekar^2,3, and Kathleen M Schmainda^2,4
1Neurosurgery, Medical College of Wisconsin, Millwaukee, WI, United States, ²Translational Brain Tumor Research Program, Medical College of Wisconsin, Milwaukee, WI, United States,³Radiology, Medical College of Wisconsin, Milwaukee, WI, United States, ⁴Radiology and Biophysics, Medical College of Wisconsin, Milwaukee, WI, United States

Few systematic studies have examined optimal timing for combining anti-angiogenic therapy with chemotherapy for brain tumor treatment. 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 use rCBV, derived from DSC-MRI, to optimize combination of anti-VEGF agent, B20, and chemotherapeutic, temozolomide, for treatment of U87 brain tumor models. The studies demonstrate optimal combination occurs when temozolomide is started two or four days after starting B20 treatment. Inhibition of tumor volume demonstrated different trends, which trend is more predictive of optimal response needs further exploration and validation.

Drew R DeBay¹, Sarah A LeBlanc¹, Genevieve M Weir², Marc Mansour², and Chris V Bowen^1,3
1National Research Council - Institute for Biodiagnostics (Atlantic), Halifax, Nova Scotia, Canada, ²Immunovaccine Inc., Halifax, Nova Scotia, Canada, ³Department of Physics and Atmospheric Science, Dalhousie University, Halifax, Nova Scotia, Canada

Due to the inherent composition of depot vaccine formulations, they are ideally suited for visualization with MRI. This study aimed to evaluate the therapeutic efficacy of a novel vaccine platform, Depovax^TM and characterize the immune response with MRI in the in vivo longitudinal assessment of mice subject to tumor challenge. We have shown for the first time, direct evidence of complete tumor eradication and immune activation elicited by Depovax^TM. MRI is a valuable tool in effectively elucidating the mechanisms of action, clearance and characterization of immune response as these vaccine products proceed through clinical development.

Teresa Delgado-Goñi^1,2, Juana Martin-Sitjar^1,2, Rui V. Simões^1,2, Milena Acosta^1,2, Silvia Lope-Piedrafita^2,3, and Carles Arús^1,2
1Bioquimica i Biologia Molecular, Universitat Autònoma 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 Ressonància Magnètica Nuclear, Universitat Autònoma de Barcelona, Cerdanyola del Valles, Barcelona, Spain

Dimethyl Sulfoxide (DMSO) is a common administration vehicle for some drugs, such as temozolomide (TMZ). DMSO has been reported to cross the blood-brain-barrier and to produce contrast, detected by 1H MRS/MRSI, between GL261 mouse glioblastoma (GBM) and nearby/peritumoral brain parenchyma. This work evaluates the effect of TMZ in mouse GBM progression, which can be monitored by T2w MRI as a transient volume growth arrest. MRSI-detection of differential DMSO accumulation in GL261 treated with respect to non-treated tumours provides a non-invasive biomarker of this response.

Yue Zhang^1,2, Yang Guo², Jodi Nicolai², Rachel A Klein², Reed A Omary^2,3, and Andrew C. Larson^2,3
1Bioengineering, University of Illinois at Chicago, Chicago, IL, United States, ²Radiology, Northwestern University, Chicago, IL, United States, ³Robert H. Lurie Comprehensive Cancer Center, Northwestern University, Chicago, IL, United States

Irreversible electroporation (IRE) has been applied as a novel tissue ablation modality; IRE involves application of short-lived electrical fields across the cell membrane to permanently increase membrane permeability leading to cell death. MRI measurements have been used to intra-procedurally monitor tissue response immediately after IRE in normal hepatic parenchyma. The purpose of this study was to determine whether conventional T1- or T2-weighted MRI measurements are similarly effective for monitoring tissue response in liver tumors using the rodent N1-S1 hepatoma model. Our study showed conventional T1W, T2W MRI measurements are poorly reflective of IRE ablation zones in tumor tissues. Future studies are necessary to develop new imaging approaches better able to specifically detect signal alterations due cell membrane permeablization in targeted tumor tissues.

Ewelina Kluza¹, Marieke Heisen², Sophie Schmid¹, Daisy W.J. van der Schaft³, Raymond M Schiffelers⁴, Gert Storm⁴, Bart M ter Haar Romeny², Gustav J Strijkers¹, and Klaas Nicolay^1
1Biomedical NMR, Department of Biomedical Engineering, Eindhoven University of Technology, Eindhoven, Netherlands, ²Biomedical Image Analysis, Department of Biomedical Engineering, Eindhoven University of Technology, Eindhoven, Netherlands, ³Biomechanics and Tissue Engineering, Department of Biomedical Engineering, Eindhoven University of Technology, Eindhoven, Netherlands, ⁴Department of Pharmaceutical Sciences, Utrecht University, Utrecht, Netherlands

Liposome-incorporated glucocorticoids (GC) are known for their tumor growth inhibitory effects. The mechanism by which this occurs is largely unknown. The present study sought to test the hypothesis that liposomal GC activity involves inhibition of tumor angiogenesis. To that aim, dynamic contrast enhanced (DCE)-MRI was used to probe vascular function in B16F10 tumors, grown s.c. in C57BL/6 mice. DCE-MRI data were fitted with the two-compartment Tofts model. In addition, T2- and DW-MRI and histopathology were used. GC treatment caused a significant reduction in Ktrans from day 2 post-therapy and prevented the ADC reduction seen in non-treated tumors.

Aime T Franco¹, Ronald A Ghossein², H Carl Le³, Jason A Koutcher^4,5, and Jame Fagin^6
1Medicine and Human Human Oncology and Pathogenesis Program, MSKCC, New York, NY, United States, ²Pathology, MSKCC, New York, NY, United States, ³Medical Physics, MSKCC, New York, NY, United States, ⁴Medical Physics, MSKCC, ⁵Medicine, MSKCC, ⁶Medicine and Human Oncology and Pathogenesis Program, MSKCC

We have developed mice with a thyroid-specific knock-in of oncogenic BRAF (LSL-BRAFV600E/TPO-Cre) to explore the role of pharmacological treatments in a physiologically relevant mouse model of PTC. Murine Braf-induced PTCs were treated with the specific allosteric MEK1/2 inhibitor PD325901 or rapamycin for three weeks, beginning at 3 weeks. We have used non-invasively measured thyroid volume by high resolution MRI as the tumor biomarker

Anita Gianella^1,2, Peter A. Jarzyna¹, Venkatesh Mani¹, Sarayu Ramachandran¹, Claudia Calcagno¹, Gert Storm³, David P. Cormode¹, Victor L. Thijssen⁴, Arjan W. Griffioen⁴, Zahi A. Fayad¹, and Willem J.M. Mulder^1
1Translational and Molecular Imaging Institute, Mount Sinai School of Medicine, New York, NY, United States, ²Centro Cardiologico Monzino, Milano, Milano, Italy, ³Utrecht Institute for Pharmaceutical Science, Utrecht, Netherlands, ⁴Department of Medical Oncology, VU University Medical Center, Amsterdam, Netherlands

In this study we developed a new multifunctional nanoparticle platform to deliver therapeutic and diagnostic hydrophobic agents to tumors. The nanoparticles consist of 50 nm oil-in-water nanoemulsions that allow the inclusion of hydrophobic coated iron oxide nanocrystals in their core and the near infrared (NIR) fluorescent dye Cy7 coupled to lipids on the surface for multimodal imaging. The inclusion of an additional payload of a hydrophobic glucocorticoid resulted in nanoparticles that significantly inhibited tumor growth in a mouse model.

Benjamin Lemasson¹, Stefanie Galbán², Terence M Willians², Fei Li¹, Kevin A Heist¹, Timothy D Johnson³, Alnawaz Rehemtulla^1,2, Craig J Galbán¹, and Brian Dale Ross^1
1Radiology, University of Michigan, Ann Arbor, Michigan, United States, ²Radiation Oncology, University of Michigan, Center for Molecular Imaging, Ann Arbor, Michigan, United States,³Biostatistics, University of Michigan, Ann Arbor, Michigan, United States

We evaluated the efficacy of Gemcitabine+radiation to standard of care, Temozolomide+radiation, using a genetically engineered glioblastoma model in mice. We also tested the sensitivity of DW-MRI as a surrogate imaging biomarker of tumor response. We found that Gemcitabine+radiation resulted in a significant reduction in tumor volume and prolonged survival (same as temozolomide+radiation) in this clinically-relevant GBM mouse model. We also demonstrated the successful implementation of imaging biomarker surrogates (ADC) which correlated with therapeutic effectiveness. Based on these results, Gemcitabine+radiation appeared to be a suitable alternative to treat gliomas that have a poor response to traditional therapies (eg. unmethylated MGMT).

Peter Johnson¹, Simon Walker-Samuel², Vineeth Rajkumar³, Mathew Robson³, Mark F Lythgoe*², and Barbara Pedley*^3
1Institute of Cancer, University College London, London, United Kingdom, ²UCL Centre for Advanced Biomedical Imaging, Department of Medicine and Institute of Child Health, University College London, London, United Kingdom, ³Institute of Cancer, University College London, United Kingdom

Orthotopic tumour xenograft models, in which tumour cells are implanted and developed in the organ from which they were derived, have found increasing interest as it is thought that they provide a more representative model of tumour pathophysiology and tumour-stromal interaction. However, it is notoriously difficult ro evaluate therapeutic efficacy without sacrificing the host. In this study, the response to treatment with OXi4503, a vascular disrupting agent, is compared in subcutaneous and orthotopic liver tumour models using susceptibility and diffusion MRI. Marked differences in the rate and magnitude of response between the two models are identified and quantified.

Julio Cardenas¹, Yuguo Li², Christine A Howison³, Jean-Philippe Galons⁴, Amanda F Baker⁵, and Mark D Pagel^6
1Chemistry and Biochemistry, University of Arizona, Tucson, AZ, United States, ²Radiology, Case Western Reserve University, Cleveland, OH, United States, ³Arizona Research Laboratories, University of Arizona, Tucson, AZ, United States, ⁴Radiology, University of Arizona, Tucson, AZ, United States, ⁵Hematology/Oncology, Arizona Cancer Center, University of Arizona, Tucson, AZ, United States, ⁶Biomedical Engineering and Chemistry & Biochemistry, University of Arizona, Tucson, AZ, United States

We have used DCE-MRI and DW-MRI to investigate the biological response to TH-302, a hypoxia-activated bis-alkylating prodrug in a pre-clinical model of pancreatic cancer. As a consequence of TH-302 selective effects on poorly-vascularized tumor regions, we expected a change in DCE-MRI and/or DW-MRI. The results showed that TH-302 caused a change in tumor vasculature as measured with DCE-MRI, but did not change cell membrane integrity measured with DW-MRI. These effects were homogenous throughout the tumor. These results demonstrate advantages of combining DCE-MRI and DW-MRI for therapy studies.

Nkiruka C Atuegwu¹, Daniel C Colvin¹, Mary E Loveless^1,2, Lei Xu³, John C Gore^1,4, and Thomas E Yankeelov^1,4
1Institute of Imaging Science, Vanderbilt University, Nashville, Tennessee, United States, ²Biomedical Engineering, Vanderbilt University, Nashville, Tennessee, United States, ³Biostatistics, Vanderbilt University, Nashville, Tennessee, United States, ⁴Radiology, Vanderbilt University, Nashville, Tennessee, United States

Sequential diffusion weighted MRI (DW-MRI) data of 13 Fischer rats implanted with 9L tumors were incorporated into a mathematical model of tumor growth and treatment response. 8 rats were treated with the chemotherapy drug BCNU and 5 rats were tumor bearing controls. DW-MRI data at early time points were used to extract the proliferation rates of each tumor and was then used to calculate the number of cells at a later time point. The calculated and experimentally estimated number of cells were then compared.

Yann Jamin¹, Elizabeth R Cullis², Lynsey Vaughan², Hannah Webber², Jessica KR Boult¹, Lauren C Baker¹, Dow-Mu Kow¹, Louis Chesler², and Simon P Robinson^1
1CRUK and EPSRC Cancer Imaging Centre, The Institute of Cancer Research and Royal Marsden NHS Trust, Sutton, United Kingdom, ²Paediatric Oncology, The Institute of Cancer Research, Sutton, United Kingdom

In this study, we have challenged the TH-MYCN transgenic neuroblastoma murine model, which is a faithful representation of high-risk childhood neuroblastoma, with three different classes of anti-cancer agent: the conventional cytotoxic drugs cyclophosphamide (CPM) and methotrexate (MTX), the VEGF signalling inhibitor cediranib, and the tubulin-binding agent N-acetyl colchinol (ZD6126). We demonstrate a systematic reduction in T₁ upon successful treatment and that the accurate quantification of T₁ thus affords a generic, noninvasive and clinically translatable biomarker for chemotherapy-mediated cell death in the TH-MYCN neuroblastoma model.

Jake Samuel Burrell¹, Jane Halliday², Simon Walker-Samuel³, Jessica K R Boult¹, Yann Jamin¹, John C Waterton², and Simon P Robinson^1
1The Institute of Cancer Research, Sutton, Surrey, United Kingdom, ²AstraZeneca, Manchester, United Kingdom, ³Centre for Advanced Biomedical Imaging, UCL, London, United Kingdom

The combined carbogen USPIO (CUSPIO) imaging protocol uses two MRI biomarkers of vascular function, R₂* during carbogen breathing and R₂* after injection of USPIO particles, combined with a novel segmentation method, which provides information about their spatial co-localisation. CUSPIO imaging provides five response categories, which reflects tumour vascular function and architecture. In this study, CUSPIO imaging was applied to investigate the response of C6 rat xenografts to cediranib, an inhibitor of VEGF signalling. A decrease in fractional blood volume, and significant differences in CUSPIO response categories, were measured after treatment with cediranib.

Yanping Sun¹, Juan Wang¹, Amy M. Saur¹, Zandra Walton², Val Monrose¹, Kwok-Kin Wong², and Andrew L. Kung^1,3
1Lurie Family Imaging Center, Dana-Farber Cancer Institute and Harvard Medical School, Boston, MA, United States, ²Department of Medical Oncology, Dana-Farber Cancer Institute and Harvard Medical School, Boston, MA, United States, ³Pediatric Oncology, Children's Hospital Boston, Boston, MA, United States

For the purposes of pathology characterization and evaluation of new cancer treatments, a simple, fast and robust imaging protocol with high throughput is essential. We demonstrated that the IntraGate technique used in this study is a fast and simple way to image mouse lungs without the complication of cardiac and respiratory gating. The MRI lung imaging protocol established here showed increased SNR efficacy. We used this protocol to monitor the progression of adenocarcinomas and for evaluating the efficacy of the irreversible EGFR/HER2 inhibiter BIBW2992. Objective response to treatment with BIBW2992 was apparent even after a single treatment dose.

Carole Danielle Thomas^1,2, Florent Poyer^1,2, Philippe Maillard^2,3, Andreas Volk^1,2, Guillaume Garcia^2,3, Alain Croisy^1,2, Mihaela Lupu^1,2, and Joel Mispelter^1,2
1Research, Curie Institute/INSERM U759, Orsay, France, Metropolitan, ²University Paris XI, Orsay, France, Metropolitan, ³Research, Curie Institute/CNRS UMR176, Orsay, France, Metropolitan

Introduction: Two protocols of photodynamic therapy (PDT) were characterized via 1H and 23Na MRI. Methods: The treatment protocols were applied on human colorectal tumors implanted on nude mice: single targeting (antivascular) PDT and double targeting (anticellular and antivascular) PDT. Results: Extracellular sodium MRI characterized the early treatment efficacy, confirmed by the final tumoral development. Double targeting PDT prevents tumor progression in contrast to single target PDT. Conclusions: Double targeting PDT led to important cellular damage as shown by sodium MRI which is therefore a well suited, non-invasive imaging technique for longitudinal follow-up and treatment assessment.

Oussama Karroum¹, Julie Kengen¹, Pierre Danhier¹, Julien Verrax², Pedro Buc Calderon², Pierre Sonveaux³, Vincent Gregoire⁴, Bernard Gallez¹, and Benedicte F Jordan^1
1Louvrain Drug Research Institute, Biomedical Magnetic Resonance Group, Universite Catholique de Louvain, Brussels, Belgium, ²Louvrain Drug Research Institute,Pharmacokinetics, Metabolism, Nutrition, and Toxicology Group, Universite Catholique de Louvain, ³IREC, Pole of pharmacology and therapeutics, Universite Catholique de Louvain, ⁴IREC, molecular imaging and radiotherapy, Universite Catholique de Louvain

The effects of different MAPKinase inhibitors (Sorafenib, Gefitinib, FTS, and PD-0325901) were monitored in FSaII experimental mouse tumors. In vivo EPR oximetry (L-band) showed an increase in tumor pO2 for all MAPK inhibitors and allowed the identification of a window of reoxygenation, which was further exploited for improving radiation response. The increase in tumor oxygenation was shown to be the result of two major factors: (i) an increase in blood flow for sorafenib only and, (ii) a decrease in oxygen consumption for all MAPK inhibitors, as measured ex vivo using X-band EPR oximetry.

Basil Künnecke¹, Barbara Lenz¹, Markus Stephan-Güldner¹, Anna Maria Brändli-Baiocco¹, Jürgen Funk¹, Thomas Pfister¹, and Markus von Kienlin^1
1Pharmaceuticals Division, F. Hoffmann-La Roche Ltd, Basel, Switzerland

Thyroid hypertrophy, hyperplasia and neoplasia are spontaneous but also drug-induced findings observed in non-clinical safety studies in rats. In order to understand the time of onset, progression and potential regression of these abnormalities longitudinal (i.e., time course) assessment are required. In humans, MRI is commonly used to non-invasively diagnose thyroid neoplasia. The purpose of this study was to back-translate these modalities to the rodent situation. MRI yielded excellent tissue contrast in rat thyroid glands, thus providing the prerequisite for virtual histology in situ. Comparison of MR images to histopathological sections revealed good correlation for lesions larger than approximately 500um.

Jean-Christophe Brisset¹, Stefanie Galbán², Alnawaz Rehemtulla², Kenneth James Pienta³, Craig J Galbán¹, and Brian Dale Ross^1
1Radiology, University of Michigan, Ann Arbor, Michigan, United States, ²Department of Radiation Oncology, University of Michigan, Ann Arbor, Michigan, United States, ³Department of Internal Medicine and Urology, University of Michigan, Ann Arbor, Michigan, United States

Bone metastases occur in more than 90% of patients with advanced prostate cancer. The purpose of this study was to develop an imaging biomarker to monitoring a patients' individual therapeutic progress. Human prostate cancer cells were implanted by direct intra-tibial injection into mice. Control (N=6), docetaxel (qd7x3; N=5), Radiation therapy (2 Gy/day 5x per week, 2 weeks; N=7) and combined treatment (N=5) were performed. DW-MRI was performed the first 2 weeks. Our results demonstrate the potential of quantitative ADC map for assessing tumor response even at the early time post therapy.

Mirabelle Sajisevi^1,2, David A Bellnier³, Nestor Rigual², and Mukund Seshadri^1,4
1Preclinical Imaging Facility, Roswell Park Cancer Institute, Buffalo, NY, United States, ²Head and Neck Surgery, Roswell Park Cancer Institute, Buffalo, NY, United States, ³Cell Stress Biology, Roswell Park Cancer Institute, ⁴Pharmacology and Therapeutics, Roswell Park Cancer Institute, Buffalo, NY, United States

The biological response of tumors to photodynamic therapy (PDT) is complex and involves a combination of cell kill, vascular damage, and induction of immune/inflammatory responses, all of which result in varying changes in contrast on MR images. In this study, we examined the utility of MRI in monitoring tissue response to PDT in an invasive, intramuscular model of squamous cell carcinoma. Our results demonstrate the potential value of MRI as a tool to monitor PDT-induced tissue damage in vivo.

Alessandro Ricci¹, Serena Cecchetti¹, Maria Elena Pisanu¹, Luisa Paris¹, Luigi Portella², Stefania Scala², Egidio Iorio¹, and Franca Podo^1
1Cell Biology and Neurosciences, Istituto Superiore di Sanità, Rome, RM, Italy, ²Department of Clinical Immunology, National Cancer Institute, Naples, NA, Italy

CXCR4-CXCL12 axis controls the metastatic homing of tumor cells and may act as target for antitumor therapy. We investigated the 1H-MRS choline profile in relation to changes of CXCR4 expression induced by a selective inhibitor of phosphatidylcholine-specific phospholipase C (PC-plc), tricyclodecan-9-yl-potassium xanthate (D609) in human T-lymphoblastoid cells (CEM). We showed that PC-plc physically associates with CXCR4; inhibition of PC-PLC activity induces down-modulation of CXCR4 from the plasma membrane; 1H MRS analyses of cell extracts showed that an about 2-fold decrease in the phosphocholine signal area may act as marker of simultaneous PC-PLC inhibition and CXCR4 down-modulation in D609-treated CEM cells.

Anna Granata¹, Egidio Iorio², Maria Teresa Comito¹, Alessandro Ricci², Maria Elena Pisanu², Zaver M Bhujwalla³, Franca Podo², Silvana Canevari¹, Delia Mezzanzanica¹, and Marina Bagnoli^1
1Fondazione IRCCS Ist. Nazionale Tumori, Milano, Mi, Italy, ²Cell Biology and Neurosciences, Istituto Superiore di Sanità, Roma, RM, Italy, ³John Hopkins University, Baltimore, MA, United States

Detection and characterization by magnetic resonance spectroscopy of altered phosphatidylcholine metabolism in epithelial ovarian cancer (EOC) could provide choline-based imaging approaches as powerful tools to improve diagnosis and identify new therapeutic targets. We specifically silenced choline kinase alpha (ChoK-alpha) expression by transient RNA interference in two EOC cell lines and evaluated the effects induced on MRS choline profile, cell proliferation and global gene expression. Our observations supported the major role of ChoK-alpha in deregulated choline metabolism in EOC cells, warranting further investigations to further evaluate its possible use as new molecular target.

Noriko Mori¹, Mayur Gadiya¹, Flonne Wildes¹, Balaji Krishnamachary¹, and Zaver M Bhujwalla^1
1JHU ICMIC Program, The Russell H. Morgan Department of Radiology and Radiological Science, The Johns Hopkins University, School of Medicine, Baltimore, MD, United States

Elevated phosphocholine and high choline kinase (Chk) expression are typically observed in cancer. Chk is being evaluated as a novel target in cancer treatment using pharmacological and molecular inhibition. Since these agents are delivered systemically it is important to determine the effect of Chk on endothelial cells in normal and tumor tissue. We used siRNA against Chk (siRNA-chk) to down-regulate Chk expression in endothelial cells. Transient siRNA-chk transfection significantly reduced phosphocholine levels and proliferation in breast cancer cells, but not in endothelial cells. These data suggest that downregulation of Chk does not affect endothelial cell proliferation.

Ioannis Stasinopoulos¹, Tariq Shah¹, Yelena Mironchik¹, Balaji Krishnamachary¹, and Zaver M. Bhujwalla^1,2
1JHU ICMIC Program, Russell H. Morgan Department of Radiology and Radiological Science, The Johns Hopkins University School of Medicine, Baltimore, MD, United States, ²The Sidney Kimmel Comprehensive Cancer Center, The Johns Hopkins University School of Medicine, Baltimore, MD, United States

Increased phosphocholine and total choline are consistently observed in cancers, especially breast cancer, with 1H MRS, and it is important to uncover the complexity of factors that regulate choline metabolism in cancer, and the compensatory mechanisms in this pathway. Cyclooxygenase (COX)-2, the inducible cyclooxygenase that forms the inflammation-mediator prostaglandin E2 (PGE2), has an enormous impact on cell motility, invasion, vascular characteristics and metastatic dissemination. Here we have shown that COX-2 influences choline metabolism through choline kinase. The association between COX-2 and choline kinase may identify new biomarkers and new targets to use in combination with COX-2 targeting in cancer treatment.

Dominik Zietkowski¹, Eszter Nagy², Margaret A Mobberley³, Geoffrey S Payne¹, Timothy A Ryder³, and Nandita M deSouza^1
1CRUK and EPSRC Cancer Imaging Centre, The Institute of Cancer Research, Sutton, Surrey, United Kingdom, ²Section of Molecular Carcinogenesis, The Institute of Cancer Research, Sutton, Surrey, United Kingdom, ³Department of Cellular Pathology, Imperial College Healthcare NHS Trust, London, United Kingdom

This study compares mobile lipid resonances (MLR) in paclitaxel exposed cells undergoing apoptosis or mitotic arrest to explore their utility as a biomarker of drug resistance. Paclitaxel-induced apoptosis in HeLa cells is associated with an increase in MLR, especially in unsaturated fatty acids, which contain the diallyl group indicating poly-unsaturation and with an increase in number and size of lipid droplets. When resistant cells progress beyond mitosis, they accumulate less saturated and unsaturated lipids in smaller droplets than sensitive lines. Western blots of synthetic and catabolic enzymes showed that both processes contribute to observed increases in MLR.

Gigin Lin¹, Helen Troy¹, Lauren Elizabeth Jackson¹, Ian R Judson², John R Griffiths³, Dow-Mu Koh¹, Simon P Robinson¹, Martin O Leach¹, and Yuen-Li Chung^1
1CRUK and EPSRC Cancer Imaging Centre, Institute of Cancer Research and Royal Marsden Hospital, Sutton, Surrey, United Kingdom, ²CRUK Centre of Cancer Therapeutics Unit, Institute of Cancer Research and Royal Marsden Hospital, Sutton, Surrey, United Kingdom, ³CRUK Cambridge Research Institute, Li Ka Shing Centre, Cambridge, United Kingdom

Autophagy is a stress response whereby cellular organelles are sequestered for lysosomal degradation as an alternate energy source. Our study suggests a distinct metabolic profile of autophagy present across different cell lines and treatments. An increased level of several amino acids was observed that might result from degradation of cellular organelles and proteins. Reduced phosphocholine and increased glycerophosphocholine levels indicate changes in membrane metabolism, which might relate to the formation and/or degradation of the double-membrane autophagosomes. Metabolomic analysis of autophagy provides a distinct metabolic profile, which may potentially be used as a non-invasive surrogate biomarker.

Egidio Iorio¹, Chiara Alberti², Paola Alberti², Alessandro Ricci¹, Maria Elena Pisanu¹, Patrizia Pinciroli², Silvana Canevari², Franca Podo¹, and Antonella Tomasetti^2
1Cell Biology and Neurosciences, Istituto Superiore di Sanità, Roma, RM, Italy, ²Fondazione IRCCS Istituto Nazionale dei Tumori, Milano, Italy

Our previous MRS studies showed 3-to-8-fold increases in phosphocholine (PCho) in epithelial ovarian cancer (EOC) cells compared with non tumoral counterparts. The contribution of the PI3K/AKT pathway to PCho accumulation was investigated in three EOC cell lines exposed to the PI3K inhibitor Ly294002. PI3K inhibition was associated with strong decrease in AKT phosphorylation and block of cell proliferation in two LY294002-sensitive (IGROV1 and OAW42) but not in the resistant SKOV3 cells. No alterations were however detected in the PCho levels of any of these cell lines, indicating that multiple pathways likely contribute to the PCho accumulation detected in EOC cells.

Omkar B Ijare¹, Tedros Bezabeh¹, Nils Albiin², Annika Bergquist², Urban Arnelo², Matthias Lohr², and Ian C.P. Smith^1
1National Research Council Institute for Biodiagnostics, Winnipeg, Manitoba, Canada, ²Karolinska University Hospital, Karolinska Institutet, Huddinge, Stockholm, Sweden

Phosphatidylcholine is an important component of bile protecting bile ducts from harmful effects of bile acids. In some cholestatic patients, the hydrolysis of phosphatidylcholine has been observed. Since phosphatidylcholine and its hydrolysis products – glycerophosphocholine (GPC), and phosphocholine – have similar chemical shift values for their -N⁺(CH₃)₃ signals, it will be difficult to detect the biochemical changes occurring in the course of the disease using ¹H MRS. Although 2D ¹H-¹H COSY experiments are of help in this regard, they require longer experimental time. Hence, the potential of ³¹P MRS as an alternative approach is assessed in this study.

Dominik Zietkowski¹, Geoffrey Payne¹, and Nandita deSouza^1
1CRUK and EPSRC Cancer Imaging Centre, The Institute of Cancer Research, Sutton, Surrey, United Kingdom

This study compares mobile lipid resonances (MLR) in cervical tissue biopsies between normal epithelium & stroma, dysplastic epithelium (CIN) and cancer and tests for a correlation with the presence and number of cytoplasmic lipid droplets. Large heterogeneity in the lipid content among samples in all three classes resulted in no significant differences in MLR intensities between them, although CIN biopsies displayed more MLR (apart from the polyunsaturated peak at 2.8 and triglyceride at 4.1 and 4.3 ppm which were higher in the cancer class). The number of lipid droplets correlated significantly with most intense methyl and methylene peaks.

Kavindra Nath¹, David S Nelson¹, Andrew Ho¹, Brian P Weiser², Rong Zhou¹, Stephen Pickup¹, Lin ZJ Li¹, Deenis Leeper³, and Jerry D Glickson^1
1Radiology, University of Pennsylvania, Philadelphia, Pennsylvania, United States, ²Pharmacology, University of Pennsylvania, Philadelphia, Pennsylvania, United States, ³Radiation Oncology, Thomas Jefferson University, Philadelphia, Pennsylvania, United States

Synopsis: In vivo P-31MRS was used to monitor the effects of lonidamine on pH and bioenergetics of human melanoma xenografts and normal tissues. Intracellular pH was reduced through 180 minutes with a maximum decrease (0.6 units, p=0.003) occurring 80 min. after lonidamine injection, and the bioenergetics (βNTP/Pi ratio) of the tumor decreased relative to baseline (56%, p=0.004). In liver, a transient decrease in pH (~0.2 units, p=0.19) was observed after 20 mins, with a maximum decrease in βNTP/Pi (~21%, p=0.04) occurring after 40 min. Noninvasive monitoring of brain and muscle showed no significant changes. This selective decrease of tumor pH and bioenergetics may be exploited for sensitization to pH dependent chemotherapy.

Lori R. Arlinghaus^1,2, Richard D. Dortch^1,2, Adrienne N. Dula^1,2, Seth A. Smith^1,2, John C. Gore^1,2, and Thomas E. Yankeelov^1,2
1Institute of Imaging Science, Vanderbilt University, Nashville, TN, United States, ²Department of Radiology and Radiological Sciences, Vanderbilt University, Nashville, TN, United States

Magnetization transfer (MT) imaging is gaining increased attention in the field of cancer imaging. MT imaging of healthy controls was performed at 3T to explore the variability of the magnetization transfer ratio (MTR) of normal fibroglandular (FG) tissue. We were able to produce reliable MTR maps with good fat suppression, and mean MTR values for FG tissue ranged from 20% to 26%.

Jeon-Hor Chen^1,2, Siwa Chan³, Daniel H-E Chang¹, Muqing Lin¹, Orhan Nalcioglu¹, and Min-Ying L. Su^1
1Center for Functional Onco-Imaging and Department of Radiological Science, University of California Irvine, Irvine, California, United States, ²Department of Radiology, China Medical University Hospital, Taichung, Taiwan, ³Department of Radiology, Taichung Veterans General Hospital, Taichung, Taiwan

The variation of breast density measured from 4 different scanners (GE 1.5T, GE 3T, Siemens 1.5T, Philips 3T) was investigated. The breast volume, fibroglandular tissue volume, and the percent density were measured using computer algorithms. The image qualities of all 4 scanners are superb for density segmentation. Philips 3T has the highest SNR, and Siemens 1.5T has the best tissue contrast. The coefficient of variation is 4.0% for BV, 6.2% for FV; and 6.0% for PD. The results suggest that with a careful standardization in acquisition and analysis, density parameters obtained from multiple centers may be used for combined analysis.

Jeon-Hor Chen^1,2, Muqing Lin¹, Fu-Ju Lei², Jia-Pei Wu², Siwa Chan³, Orhan Nalcioglu¹, and Min-Ying L Su^1
1Center for Functional Onco-Imaging and Department of Radiological Science, University of California Irvine, Irvine, California, United States, ²Department of Radiology, China Medical University Hospital, Taichung, Taiwan, ³Department of Radiology, Taichung Veterans General Hospital, Taichung, Taiwan

In general computer-aided algorithms for segmentation of breast density on MRI work well, but some errors may remain in small regions of the breast. Two major problems were the strong intensity inhomogeneity within a large area, and the low contrast between fibroglandular tissue and fatty tissue. In this study the operator noted the errors and classified them into three types. Several correction strategies were developed based on the nature of these errors and demonstrated satisfactory correction results. These processes to identify potential errors and correction strategies are important for developing a fully automated procedure for quantitative analysis of MRI-based density.

Maria A Schmidt¹, Antonio de Stefano², Erica Scurr¹, James d'Arcy¹, and Martin O Leach^1
1Cancer Research UK & EPSRC Cancer Imaging Centre, Royal Marsden NHS Foundation Trust and Institute of Cancer Research, Sutton, England, United Kingdom, ²Medical Physics, Portsmouth NHS Hospitals Trust, Portsmouth, England, United Kingdom

Quantitative measurements of water and fat content are desirable for the assessment of breast density, a known risk factor in breast cancer. Dixon methods produce high-resolution separate fat and water images of high-resolution, but longitudinal changes in image intensity can be attributed to either T1 or proton density changes. We optimised data acquisition parameters to produce separate proton density maps for fat and water from Dixon images acquired with different flip angles. Monte-Carlo results were confirmed in test object studies, and implemented for breast examinations, providing a valuable tool for clinical studies.

Leah C Henze-Bancroft¹, Catherine J Moran², Scott B Reeder^3,4, Frederick Kelcz⁴, and Walter F Block^3,5
1Department of Biomedical Engineering, University of Wisconsin - Madison, Madison, WI, United States, ²Department of Radiology, Stanford University, Stanford, CA, ³Department of Medical Physics, University of Wisconsin - Madison, Madison, WI, ⁴Department of Radiology, University of Wisconsin - Madison, Madison, WI, ⁵Department of Biomedical Engineering, University of Wisconsin - Madison, Madison, WI

Bilateral breast MR is important for the screening of high risk women. Standard T2-weighted breast imaging must be performed for each breast separately and is limited in resolution due to long scan times. We present a 3D-radial balanced SSFP sequence with IDEAL fat/water separation capable of simultaneously acquiring data across both breasts to provide fat and water separated images with T2-like weighting and 0.6 mm isotropic resolution, in clinically acceptable scan times.

Martin D Pickles¹, Peter Gibbs¹, Martin Lowry¹, and Lindsay W Turnbull^1
1Centre for MR Investigations, University of Hull, Hull, East Yorkshire, United Kingdom

The aim of this work was to determine if there were any associations between pre-treatment MR derived vascular, texture and shape parameters with traditional histopathological based prognostic markers. The result of this work provides limited support to the findings of previous investigations that have demonstrated significant associations between MR derived vascular kinetics and histopathological markers of prognosis. Additionally, this work has revealed significant associations between MR texture and shape parameters with traditional prognostic indicators. Moreover texture parameters were retained by the logistic regression analysis model more often than both vascular and shape descriptors.

Marco Reisert¹, Matthias Weigel¹, Erez Eyal², Dov Grobgeld², Hadassa Degani², and Jürgen Hennig^1
1Medical Physics, University Medical Center Freiburg, Freiburg, Baden Württemberg, Germany, ²Biological Regulation Dept., Weizmann Institute of Science, Rehovot, Israel

The architecture of the ductal trees was first investigated by Sir Astley Cooper in 1840, using duct injection studies ex-vivo. Recently, computer derived tracking of whole-breast ductal trees has been achieved in few human breasts using mastectomy specimens. Studying the architecture of the entire ductal trees is very challenging and has not been achieved in vivo, yet. The functional breast tissue is composed of many lobes, which are highly variable in size and shape. Each lobe/system has one central duct with its peripheral branches forming a ductal tree and their associated glandular tissues. A new non-invasive MRI method for in vivo tracking of the mammary ductal trees using diffusion tensor imaging (DTI) is proposed.

Rachel Waichung Chan¹, Elizabeth Anne Ramsay², Edward Yiuwah Cheung³, and Donald Bruce Plewes^1,2
1Medical Biophysics, University of Toronto, Toronto, Ontario, Canada, ²Imaging Research, Sunnybrook Health Sciences Centre, Toronto, Ontario, Canada, ³University of Waterloo, Waterloo, Ontario, Canada

Flexible radial imaging allows multiple image sets, each having a different spatiotemporal balance, to be retrospectively reconstructed from the same dataset. Flexibility in dynamic contrast-enhanced breast imaging is desired because the optimal spatiotemporal balance for image diagnosis is unknown. Radially undersampled images have high temporal resolution, but suffer from undersampling streak artifacts that degrade image quality. Compressed sensing (CS) reconstruction has been shown to reduce such streak artifacts by enforcing sparsity in the image. Here, we implement and compare various flexible radial sampling schemes in an in vivo experiment, and reconstruct images with CS over a range of spatiotemporal resolutions.

Dmitri Artemov¹, Wenlian Zhu¹, and Yoshinori Kato^1
1Radiology, Johns Hopkins University, Baltimore, MD, United States

DCE-MRI is a powerful technique to detect, assess, and stage tumors. Unfortunately, quantitative measurement of tumor vascular parameters from DCE-MRI measurements is complicated by several problems including measuring arterial input function, determining contrast enhancement parameters, and modeling GdDTPA pharmacokinetics. We propose a new saturation recovery gradient echo imaging technique that can resolve some of these problems and provide a convenient way to determine quantitative vascular parameters Ktrans, kep, and ve using a standard two-compartment kinetic model. The method detected significant differences in vascular parameters measured in aggressive MDA-MB-231 and less aggressive MCF-7 human breast cancer tumor models.

Ferdinand Schweser^1,2, Karl-Heinz Herrmann¹, Andreas Deistung¹, Marie Atterbury^1,3, Pascal A Baltzer⁴, Hartmut Peter Burmeister⁴, Werner Alois Kaiser⁴, and Jürgen R. Reichenbach^1
1Medical Physics Group, Dept. of Diagnostic and Interventional Radiology 1, Jena University Hospital, Jena, Germany, ²School of Medicine, Friedrich Schiller University of Jena, Jena, Germany,³Dept. of Physics, Brown University, Providence, RI, United States, ⁴Dept. of Diagnostic and Interventional Radiology 1, Jena University Hospital, Jena, Germany

During the last 20 years MR mammography (MRM) has proven to be a powerful and highly sensitive modality to detect and study breast cancer. However, the diagnosis of breast carcinoma in early stage is still challenging. In this initial study, we demonstrate, for the first time, that clinical application of quantitative magnetic susceptibility mapping (QSM), a novel technique in the breast is feasible and may provide additional information about breast lesions potentially improving clinical diagnoses.

Laura Martincich¹, Matthieu Faivre-Pierret², Christian M. Zechmann³, Stefano Corcione⁴, Harrie CM van den Bosch⁵, Wei-Jun Peng⁶, Antonella Petrillo⁷, Katja Siegmann⁸, Johannes T. Heverhagen⁹, Pietro Panizza¹⁰, Hans-Björn Gehl¹¹, Felix Diekmann¹², Federica Pediconi¹³, Lin Ma¹⁴, Fiona J. Gilbert¹⁵, Francesco Sardanelli¹⁶, and Paolo Belli^17
1Department of Diagnostic Imaging, Institute for Cancer Research and Treatment (IRCC), Candiolo, Torino, Italy, ²Center Oscar Lambret, Lille, France, ³German Cancer Research Center, Heidelberg, Germany, ⁴University Hospital “S. Anna”, Ferrara, Italy, ⁵Catharina Hospital, Eindhoven, Netherlands, ⁶Cancer Hospital, Fudan University, Shanghai, China, People's Republic of,⁷National Cancer Institute, Fondazione G. Pascale, Napoli, Italy, ⁸University Hospital Tuebingen, Tuebingen, Germany, ⁹University Hospital, Philipps University, Marburg, Germany, ¹⁰Ospedale San Raffaele, Milano, Italy, ¹¹Klinikum Bielefeld, Bielefeld, Germany, ¹²Universitätsklinikum Charité, Berlin, Germany, ¹³University of Rome “La Sapienza”, Rome, Italy, ¹⁴Chinese People's Liberation Army (PLA) General Hospital, Beijing, China, People's Republic of, ¹⁵15. Aberdeen Biomedical Imaging Centre, Aberdeen, United Kingdom, ¹⁶Policlinico San Donato, San Donato Milanese, Italy, ¹⁷Institute of Radiology, "A. Gemelli", Rome, Italy

In an intra-individual crossover comparison of 0.1 mmol/kg doses of gadobenate dimeglumine and gadopentetate dimeglumine in 162 women, three unaffiliated, blinded readers each detected significantly more malignant breast lesions with gadobenate dimeglumine (132-136 [91.7-94.4%] vs. 115-120 [79.9-83.3%]; p¡Ü0.0003). Cancer misdiagnosis rates were roughly double with gadopentetate dimeglumine (4.9-11.9% vs. 2.6-4.0%). Significantly better sensitivity (91.1-95.2% vs. 81.2-84.6%; p¡Ü0.0011), specificity (96.9-99.0% vs. 93.8-97.8%; p¡Ü0.0094), accuracy (96.7-98.2% vs. 92.8-96.1%; p<0.0001), PPV (77.2-91.1% vs. 60.9-80.7%; p¡Ü0.0002) and NPV (99.0-99.4% vs. 97.8-98.1%; p¡Ü0.0003), was noted with gadobenate dimeglumine for cancer detection. Three-reader agreement was good (76.4%; ¦Ê=0.689) for gadobenate dimeglumine but moderate (66.2%; ¦Ê=0.574) for gadopentetate dimeglumine.

Ritse M. Mann¹, Roel D.M. Mus¹, Christian Geppert², Jelle O. Barentsz¹, and Henkjan Huisman^1
1Radiology, Radboud University Nijmegen Medical Centre, Nijmegen, Gld, Netherlands, ²MR Oncology, Siemens Healthcare, Erlangen, Germany

Purpose: To evaluate whether a TWIST sequence can be used to overcome the inherent choice between high spatial and high temporal resolution in breast-MRI. Methods: In 9 patients TWIST acquisitions were performed during the first two minutes after contrast administration. Spatial resolution was in accordance with current guidelines 1*0.9*2.5 mm. A temporal resolution of only 6.9 seconds per acquisition could be achieved. Results: Morphological analysis was comparable to conventional high spatial resolution acquisitions. In all patients the breast-MRI was interpretable by evaluation of the TWIST sequences alone. Conclusion: TWIST allows simultaneous high temporal and spatial resolution in breast-MRI.

Kristin L Granlund^1,2, Ernesto Staroswiecki^1,2, Catherine J Moran¹, Marcus T Alley¹, Bruce L Daniel¹, and Brian A Hargreaves^1
1Radiology, Stanford University, Stanford, CA, United States, ²Electrical Engineering, Stanford University, Stanford, CA, United States

Diffusion-weighted imaging is sensitive to all sources of motion and this is a challenge for breast imaging due to cardiac and respiratory motion. FADE is a steady-state sequence that can acquire T2- and diffusion-weighted images. We study the effectiveness of four different methods (breath holding, cardiac gating, respiratory gating, and non-sequential PE ordering) to reduce the motion artifacts in FADE images. Signal variations in the magnitude of projection data and image quality are evaluated. By correcting the motion artifact, we are able to generate high-quality T2- and diffusion-weighted images.

David Broadbent¹, Peter Wright¹, and Daniel Wilson^1
1Medical Physics, Leeds Teaching Hospitals, Leeds, United Kingdom

Longitudinal relaxation time (T₁)) mapping was performed on a phantom and patients with primary breast tumours using a 3D inversion recovery steady state free precession sequence. A spin-echo sequence was also used with the phantom for comparison. T₁ maps were produced by fitting signal intensities to an appropriate model using a least squares curve fitting algorithm. Mean T₁ values in the phantom T₁ maps from steady state free precession data agreed closely with the spin-echo sequence results. T₁ values of fat, lesion and arterial blood measured in vivo agreed well with published values.

Fang Liu^1,2, Anat Kornecki³, Olga Shmuilovich³, Yves Bureau^1,2, and Neil Gelman^1,2
1Imaging Division, Lawson Health Research Institute, London, Ontario, Canada, ²Department of Medical Biophysics, University of Western Ontario, London, Ontario, Canada, ³Department of Diagnostic Imaging, St. Joseph's Health Center, London, Ontario, Canada

This abstract presents an investigation of quantitative morphological features in hormone receptor positive and negative tumors obtained from dynamic contrast enhanced magnetic resonance images. In order to characterize the degree of tumor surface irregularity, quantitative morphological measures which describe the distribution of normalized radial lengths of each tumor were extracted and compared between breast tumors with different estrogen receptor (ER) and progesterone receptor (PR) status. Our results suggest that ER-/PR- tumors may have larger surface irregularities than ER+/PR+ tumors.

Sonia P Li¹, N. Jane Taylor², J. James Stirling², Mei-Lin W Ah-See¹, Mark J Beresford¹, David J Collins³, James A d'Arcy³, Andreas Makris¹, and Anwar R Padhani^2
1Mount Vernon Hospital, Northwood, Middlesex HA6 2RN, United Kingdom, ²Paul Strickland Scanner Centre, Mount Vernon Hospital, Northwood, Middlesex HA6 2RN, United Kingdom, ³CRUK-EPSRC Cancer Imaging Centre, Institute of Cancer Research and Royal Marsden Hospital, Sutton, Surrey SM2 5PT, United Kingdom

Triple negative breast carcinomas (TNBC) are oestrogen (ER), progesterone (PR) and HER2 receptor negative, with evidence suggesting that they are a distinct biological entity with aggressive features and limited treatment options. DCE-MRI characteristics of these tumours were compared to a more favourable prognostic group, ER+/ PR+/HER2- breast cancers. The increased cellularity and scant stromal content of TNBC was reflected by lower values for v_e, a parameter describing the extravascular extracellular space. Increased k_ep values consistent with the rapid return of contrast into the vasculature, suggest a higher capillary permeability in triple negative breast cancers.

Elita DeFeo¹, Elena Brachtel², Yannick Berker², Nathalie Strittmatter², Julia Hein², Dennis Sgroi², Barbara Smith³, and Leo L. Cheng^4
1Pathology, Massachusetts General Hospital, Charlestown, MA, United States, ²Pathology, Massachusetts General Hospital, ³Surgical Oncology, Massachusetts General Hospital, ⁴Radiology, Pathology, Massachusetts General Hospital

Diagnosis and prognosis of breast cancer is based on pathological evaluations that have been developed prior to the implementation of imaging based screening. Annual screenings have increased the number of patients with early stages of the disease, challenging the prognostic paradigms. Applying the HRMAS MRS technique, we can quantify individual metabolites from native tissue and correlate those metabolites with quantitative pathology measure from the same samples. In this study we seek to establish breast cancer metabolomic profiles that can improve the accuracy of breast cancer pathological evaluations and the assessment of patient outcome.

Martin D Pickles¹, Peter Gibbs¹, Martin Lowry¹, and Lindsay W Turnbull^1
1Centre for MR Investigations, University of Hull, Hull, East Yorkshire, United Kingdom

Biomarkers of treatment response are currently being sought. Lesion shape, enhancement and kinetic curve assessment have been used to aid in the classification of breast lesions. These imaging features may also help to highlight potential poorer responders. The aim of this work was to determine if there were any associations between pre-treatment MR derived descriptors and response to neoadjuvant chemotherapy (NAC). The final LRA model indicated that patients with baseline characteristics of positive ER status and elevated texture parameters f2 (contrast) and f5 (inverse difference moment) are more likely to have residual disease at the end of NAC.

Catherine Klifa¹, David Newitt¹, Catherine Park², Sachiko Suzuki¹, Lisa Wilmes¹, Ying Lu³, and Nola Hylton^1
1Radiology and Biomedical Imaging, University of California San Francisco, San Francisco, california, United States, ²Department of Radiation Oncology, University of California San Francisco, San Francisco, California, United States, ³Department of Health Research and Policy, Stanford University, Stanford, California, United States

There is growing interest in characterizing the role of stromal tissue surrounding breast tumors. We present a technique to quantitatively assess regions surrounding tumors on breast MRI, based on proximity to tumor tissue. We applied the “proximity mapping” method to MRI of 46 patients with breast cancer undergoing neoadjuvant chemotherapy. Enhancement levels were measured at radial distances from tumor edges. We found a trend between contrast enhancement level within 10mm of the tumor after one treatment cycle and disease-free survival. Our technique could also measure other MRI functional parameters around tumors which is of interest to help guide local therapy and monitor treatment response.

Aida Kuzucan¹, Jeon-Hor Chen^1,2, Rita S Mehta³, Shadfar Bahri¹, Philip M Carpenter⁴, Hon J Yu¹, Orhan Nalcioglu¹, and Min-Ying L. Su^1
1Center for Functional Onco-Imaging and Department of Radiological Science, University of California Irvine, Irvine, CA, United States, ²Department of Radiology, China Medical University Hospital, Taichung, Taiwan, ³Department of Medicine, University of California Irvine, CA, United States, ⁴Department of Pathology, University of California Irvine, CA, United States

In this study we evaluated the diagnostic performance of MRI for HER-2 negative breast cancer patients receiving neoadjuvant chemotherapy. The diagnostic performance between hormonal receptor (HR) positive and negative cancers was compared. Overall the pCR rate was higher in HR- than HR+ patients. The size discrepancy between MRI and pathologic diagnosis was higher in HR+ than HR- cancer, and also in low proliferating cancer with Ki-67 <40% than high proliferating cancers with Ki-67 > 40%. The results suggest that the diagnosis performance of MRI is better in tumors that are showing a good response than those showing a poor response.

Xia Li¹, Lori R Arlinghaus¹, E. Brian Welch¹, A. Bapsi Chakravarthy¹, Lei Xu¹, Jaime Farley¹, Ingrid Mayer¹, Mark Kelley¹, Ingrid Meszoely¹, Julie Means-Powell¹, Vandana Abramson¹, Ana Grau¹, Mia Levy¹, John C Gore¹, and Thomas E Yankeelov^1
1Vanderbilt University Institute of Imaging Science, Nashville, TN, United States

The response of breast tumors to chemotherapy is currently monitored by frank changes in tumor morphology. Unfortunately, it is difficult to quantify and often does not correlate with tumor activity. DCE-MRI can offer information related to tumor perfusion and permeability, vascular volume, and extravascular extracellular volume fraction. In this study, we use the extended Tofts model to return estimates on these parameters during neoadjuvant chemotherapy in breast cancer patients. We found a strong relationship among the early change in Ktrans and residual tumor burden, as well as the change in the enhancing volume of tumor tissue at conclusion of therapy.

Jyoti Parikh¹, and Geoff Charles-Edwards^2
1Clinical Radiology, Guys and St Thomas' Hospitals, London, England, United Kingdom, ²Medical Physics, Guys and St Thomas' Hospitals, London, England, United Kingdom

Functional imaging is being increasingly used in the assessment of tumour response to cancer treatments. Diffusion weighted imaging (DWI) was performed on 35 patients receiving neoadjuvant chemotherapy (NACT) for breast cancer. Apparent diffusion coefficient (ADC) values pre-, mid-, and post-treatment were measured and compared to tumour size changes. Pre-treatment ADC values were lower, and greater changes in ADC values were seen prior to size changes, in the responders over non responders. These results suggest DWI may have a significant role in the prediction of and early assessment of response to NACT in breast cancer.

Marco Borri¹, Maria A Schmidt¹, Matthew Blackledge¹, Erica Scurr¹, Elizabeth O'Flynn¹, David Collins¹, Matthew Orton¹, Veronica Morgan¹, Nandita de Souza¹, and Martin O Leach^1
1Cancer Research UK and EPSRC Cancer Imaging Centre, Royal Marsden NHS Foundation Trust and Institute of Cancer Research, Sutton, United Kingdom

Breast-DWI has been used in cancer diagnostic and monitoring of treatment response. Accurate calculation of Apparent Diffusion Coefficient(ADC) is essential and depends on the choice of b-values. In this work we utilise ADC data from large breast lesions to calculate optimal b-values, and compare protocols using different numbers of b-values. The calculated separation (1150s/mm2) is larger than the range of b-values used in many published clinical studies. Our data did not show significant advantage in the use of 2 b-values only, while a larger number of b-values could minimize artifacts and ensure correct sampling for a larger range of ADC-values.

Savannah C Partridge¹, Habib Rahbar¹, Revathi Murthy², Xiaoyu Chai³, Brenda Kurland³, Wendy DeMartini¹, and Constance Lehman^1
1Radiology, University of Washington, Seattle, WA, United States, ²Bioengineering, University of Washington, Seattle, WA, United States, ³Clinical Statistics, Fred Hutchinson Cancer Research Center, Seattle, WA, United States

In this study, we investigated the correlation between ADC and DCE kinetic characteristics of 100 breast lesions and evaluated the relative diagnostic value of each quantitative measure alone and in combination. We found that lower ADC values (higher cellularity) were associated with more suspicious kinetics and that both DWI and DCE parameters significantly discriminated between benign and malignant lesions. A multivariate model incorporating ADC and worst kinetic curve type provided the greatest discriminative ability, suggesting that DWI provides distinct and complementary information to DCE for characterizing breast lesions.

Mies A. Korteweg¹, Daniel L. Polders¹, Willem P.Th.M. Mali¹, Peter R. Luijten¹, Jaco J.M. Zwanenburg¹, and Wouter B. Veldhuis^1
1Radiology, University Medical Center Utrecht, Utrecht, Netherlands

This study explores the possibility of utilizing high resolution magnetization transfer (MT) and chemical exchange saturation transfer (CEST) measurements on a clinical 7T MR system as a tool to discriminate healthy from metastatic dissected lymph nodes of breast cancer patients. The results were verified by detailed pathology analyses. Similar quantitative values for both APT and MTR in normal lymph node tissue as compared to metastatic lymph node tissue were found. This could be the result of the ex vivo condition of the tissue or due to a lack of differences in cellular density in these dissected nodes.

Laura Johnson¹, Geoff Charles-Edwards², Jyoti Parikh³, Margaret Hall-Craggs⁴, Tobias Schaeffter⁵, Quentin Pankhurst⁶, and Michael Douek^1
1Research Oncology, Kings College London, London, London, United Kingdom, ²Medical Physics, Guy's and St Thomas' NHS Foundation Trust, ³Radiology, Guy's and St Thomas' NHS Foundation Trust, ⁴Department of Imaging and Medical Physics and Bioengineering, University College London, ⁵Imaging sciences, Kings College London, ⁶Royal Institution of Great Britain

We have previously demonstrated intraoperative identification of iron oxide containing SLNs in patients with early breast cancer following subcutaneous injection using an in-house developed magnetometer, the SentiMag. To optimize this technique we analysed the pharmacokinetics of SPIOs using dynamic MRI scanning to visualise loss of signal intensity at the SLN in 'real time'. The transit of tracer to node is variable and is not predictable based on primary tumour or patient demographics. 5 patients (45%) demonstrated loss of signal intensity of the SLN in the dynamic scan. 6 patients (55%) demonstrated loss of enhancement of the node on delayed T2* weighted scans after 120 minutes.

Laura Johnson¹, Sarah Pinder¹, Margaret Hall-Craggs², and Michael Douek^1
1Research Oncology, Kings College London, London, England, United Kingdom, ²Department of Imaging and Medical Physics and Bioengineering, University College London

Contrast enhanced MRI scans of the axilla may improve pre-operative identification of lymph node metastases in patients with breast cancer, identifying those patients who require formal axillary lymph node dissection and potentially sparing those with a clear axilla any axillary surgery at all. Understanding how superparamagnetic iron-oxide nanoparticles distribute within the normal and the metastasis containing lymph node may aid the identification of abnormal nodes. Histological examination of 85 sentinel lymph nodes following injection with SPIO demonstrates preferential uptake by macrophages within the sinuses of the node. Iron does not accumulate in areas of the node containing metastases.

Mies A. Korteweg¹, Suzanne C.E. Diepstraten¹, Willem P.Th.M. Mali¹, Peter R. Luijten¹, Paul J. van Diest², Ivan Dimitrov³, Wouter B. Veldhuis¹, and Dennis W.J. Klomp^1
1Radiology, University Medical Center Utrecht, Utrecht, Netherlands, ²Pathology, University Medical Center Utrecht, Utrecht, Netherlands, ³Advanced Imaging Research Center, UT Southwestern Medical Center, Dallas, TX, United States

Lipid composition of excised healthy and metastatic sentinel lymph nodes of breast cancer patients at 7 Tesla was explored by 3D multiple voxel 1H-MRS and verified by pathology analyses as lipids are a potential discriminatory marker for malignancy. 7T 1H -MRS in this study correlates strongly with pathology and is able to non-invasively detect saturated, mono unsaturated and PUFA’s, independent of the total lipid content Only the methylene and methine protons were significantly less present in metastatic than in benign nodes. There were no significant differences found regarding total lipid content or PUFA distribution between metastatic and benign lymph nodes.

Ralph Noeske¹, and Timo Schirmer^2
1Global Applied Science Laboratory, GE Healthcare, Berlin, Germany, ²Global Applied Science Laboratory, GE Healthcare, Munich, Germany

MR Spectroscopy (MRS) is used as a clinical tool in diagnosing and characterizing breast cancer. TE-averaged PRESS is a common technique for single voxel breast spectroscopy (SVS) to remove potential sideband artifacts due to oscillations of the B0 field during the beginning of the acquisition. This study will present an acquisition scheme where the number of acquisitions per echo time (averages) varies across the bandwidth of acquired echo times. This fractional NEX (number of excitations) averaging method gives increased flexibility to adapt acquisition time to lesion size with only negligible effect on sideband artifacts.

Caroline D Jordan^1,2, Bruce L Daniel¹, Kevin M Koch³, Huanzhou Yu⁴, Steve Conolly⁵, and Brian A Hargreaves^1
1Radiology, Stanford University, Stanford, CA, United States, ²Bioengineering, Stanford University, Stanford, CA, United States, ³Applied Science Laboratory, GE, Waukesha, WI, United States,⁴Global Applied Science Laboratory, GE Healthcare, Menlo Park, CA, United States, ⁵Bioengineering, UC Berkeley, Berkeley, CA, United States

MRI is an important tool for the detection, diagnosis and staging of breast cancer. The shape of the breast may result in subject-specific susceptibility-induced field changes due to the air-tissue boundary. We investigated a model-based B0 field map based on magnetic susceptibility to determine subject-specific field inhomogeneities in the breast. Quantitative results of 10 healthy female volunteers at 3T indicate that the computed field map closely models the inhomogeneities of the measured field map. This rapid computation may provide a field map estimate for improved breast shimming or for choosing scanning input parameters.

Mariska Petra Luttje¹, Jannie Petra Wijnen¹, Wybe J. M. van der Kemp¹, Peter R Luijten¹, and Dennis W. J. Klomp^1
1Radiology, University Medical Center Utrecht, Utrecht, Utrecht, Netherlands

The effects of susceptibility and breathing on the spectral resolution in the unilateral breast measurements of seven healthy female volunteers have been investigated. Using static 3^rd order B₀shimming, a spectral resolution of 0.18 ppm can be obtained from the entire breast. This allows using ³¹P MRS at 7T to investigate phospholipid metabolism in the human breast in e.g. the assessment of cancer treatments. Furthermore, substantial dynamic B₀ field alterations that strongly correlate to breathing are observed, that can be corrected for to improve spectral resolution and SNR even further.

Bertine Luus Stehouwer¹, Dennis WJ Klomp¹, Mies A Kortweg¹, Peter R Luijten¹, Willem P Th M Mali¹, Maurice A.AJ. van den Bosch¹, and Wouter B Veldhuis^1
1Radiology, UMCU, Utrecht, Netherlands

This abstract describes contrast-enhanced MR-imaging (DCE-MRI) of the breast at 7T, with 3T comparison and with 7T phantom B1-measuremnts, in a patient with ductulolobular carcinoma. 7T-DCE-MRI was done with a 3DT1-TFE sequence with selective water excitation (WATS), after injection of 0.1mmol/kg gadobutrol. 3T was performed according to standard protocol. The examinations were well tolerated. WATS resulted in good lesion-to-fat-and-fibroglandular-tissue contrast. Kinetic curves at both field strengths showed similar rapid initial-enhancement, followed by wash-out in the delayed phase for the most malignant curve. In conclusion, clinical 7T-DCE-MRI is technically feasible and may show enhancement kinetics amenable to standard BIRADS-interpretation.

Saadallah Ramadan¹, Hayden Nicholas Box¹, Pascal Baltzer², Alexander Lin¹, Peter Stanwell^3,4, Eva Gombos¹, Werner A Kaiser², and Carolyn Mountford^1
1Radiology, Brigham and Women's Hospital, Boston, MA, United States, ²Radiology, Institute of Diagnostic and Interventional Radiology, Jena, Germany, ³School of Health Sciences, The University of Newcastle, Newcastle, NSW, Australia, ⁴Radiology, Brigham and Women’s Hospital, Boston, MA, United States

This pilot study investigates changes to lipid chemistry in the breast that occur with invasive cancer, using Localized COrrelated SpectroscopY (L-COSY). Early results indicate that the COSY method distinguishes lobular from ductal carcinoma and from healthy breast tissue based on the chemical shift of the resonance in the 3.00-3.30ppm region and the ratio of lipid peak volumes .

Patrick J Bolan¹, and Navneeth Lakkadi^1
1Radiology/CMRR, University of Minnesota, Minneapolis, MN, United States

The clinical utility of ¹H MRS measurements in breast cancer is well-established. The question of whether or not a total choline (tCho) resonance is detectable in normal breast is relevant to the clinical application. This work aims to measure the total choline concentration in the normal (non-lactating, asymptomatic) human breast and determine if normal subjects can be used to developing new sequences. Using single-voxel spectroscopy and T₂-corrected water as an internal reference, a tCho peak was identified in 9/14 subjects, with a median [tCho] of 0.46 mmol/kg. tCho detection was predictable based on the unsuppressed water SNR.

Sergey Magnitsky¹, George Belka², Christopher Sterner², Lewis A Chodosh², and Jerry d Glickson^1
1Radiology, University of Pennsylvania, Philadelphia, PA, United States, ²Cancer Biology, University of Pennsylvania, Philadelphia

The steady state level of lactate was measured in orthotopic and inducible tumor models of breast cancer. A lower level of lactate was detected in vivo in the inducible model, while ex vivo detection in both models revealed similar lactate level. One of the reasons of this phenomenon may be due to higher amounts of fat tissue in the inducible model compared to the orthotopic model. The inducible breast cancer model not only better recapitulates biological aspects of human disease, but also provides additional characteristics related to the detection of lactate that are not available in orthotopic or xenograft models.

Catherine Judith Moran¹, Anderson N Nnewihe^1,2, Bruce L Daniel¹, Kristin L Granlund^1,3, and Brian A Hargreaves^1
1Radiology, Stanford University, Stanford, CA, United States, ²Bioengineering, Stanford University, Stanford, CA, United States, ³Electrical Engineering, Stanford University, Stanford, CA, United States

T2-weighted Imaging, as an adjunct to Dynamic Contrast Enhanced Imaging can improve the specificity of breast MRI; however, high-resolution T2-weighted acquisitions typically are limited by low through-plane resolution and long scan times. We investigate the combination of a 16-channel close-fitting breast coil and a 3D Fast Spin Echo sequence for high-resolution T2-weighted breast imaging. The performance of the 16-channel coil is compared to a commercially available 8-channel breast coil. High SNR and high-resolution T2-weighted data would allow for better depiction of fine lesion morphology while also facilitating alignment with DCE findings.

Marco Vicari^1,2, Ulrich Saueressig³, James W Wiskin^4,5, Paolo Pellegretti¹, Michele Zani¹, Vera Ivanovas³, Marisa Windfuhr-Blum³, Jonathan Kroschel³, Elmar Kotter³, and Mathias Langer^3
1Esaote S.p.A., Genova, Italy, ²Dept. of Radiology, Medical Physics, University Medical Center Freiburg, Freiburg, Germany, ³Dept. of Radiology, Clinical Radiology, University Medical Center Freiburg, Freiburg, Germany, ⁴Dept. of Bioengineering, University of Utah, Salt Lake City, Utah, United States, ⁵Techniscan, Inc., Salt Lake City, Utah, United States

Breast imaging is crucial in the diagnosis of breast cancer, the most common cancer in women. Mammography is currently the gold standard, sometimes complemented by handheld ultrasound and MRI. All of these techniques are affected by complementary, not negligible, drawbacks. Hence, exploration of new diagnostics techniques, which can ideally merge the strengths of the current ones without inheriting their worst disadvantages, does rise a large interest. In our study a prototype automated breast ultrasound computed tomography technique (USCT) has been performed on 70 patients to date, together with mammography, handheld ultrasound and sometimes MRI. Characteristic features of pathologic findings revealed intriguing similarities between USCT and MRI imaging.

Bing Wen Zheng¹, Dennis Lai-Hong Cheong^1,2, Christopher Au¹, Eileen Ng¹, Soo Chin Lee³, and Thian chor Ng^1,4
1Clinical Imaging Research Center, A*STAR & National University of Singapore, 117456, Singapore, ²Neuroradiology Department, National Neuroscience Institute, 308433, Singapore,³Departments of Hematology & Oncology, National University of Singapore, 119074, Singapore, ⁴Departments of Radiology, National University of Singapore, 119074, Singapore

The purpose of this work was to evaluate the feasibility of quantitative measurement of pharmacokinetic parameters in breast DCE-MRI with direct measurement of AIF, and investigate how the impact of unreliable AIF on the accuracy of these parameters obtained by fitting the Tofts' and adiabatic approximation to the tissue homogeneity (ATH) model. Results in this work showed FA errors on AIF have a more profound influence on the accuracy of pharmacokinetic parameter than that caused by T1 errors. By using an additional surface coil in the breast DCE-MRI, higher SNR on the aorta makes the direct measurement of AIF possible, but careful calibration of FA errors should be made. This work is part of a long term DCE-MRI project of breast tumor.

Lawrence Dougherty¹, Mark Alan Rosen¹, Hee Kwon Song¹, and Mitchell D. Schnall^1
1Radiology, Hospital of the University of Pennsylvania, Philadelphia, PA, United States

The effect of gadolimium contrast dose on the diagnostic performance of DCE-MR breast imaging was studied. A population of subjects with a suspicious breast mass received a MR examination using a high dose of contrast agent independent of weight (20 ml) and a second population received a low dose (0.1 mmol/kg). The diagnostic performance, assessed using tumor morphology as well as SER, was shown to increase with the higher dose.

Fang Liu^1,2, Anat Kornecki³, Olga Shmuilovich³, and Neil Gelman^1,2
1Department of Medical Biophysics, University of Western Ontario, London, Ontario, Canada, ²Imaging Division, Lawson Health Research Institute, London, Ontario, Canada, ³Department of Diagnostic Imaging, St. Joseph's Health Center, London, Ontario, Canada

Various empirical parameters have been employed for quantitative analysis of signal enhancement curves in dynamic contrast enhanced-MRI. Among them, the time-to-peak has shown promise as a diagnostic indicator. In this study, a measure to be referred to as the time-to-peak hot spot volume was analyzed and optimized for differentiating benign versus malignant breast lesion. An investigation of the relation of hot spot volume to the lesion volume was also investigated. Our results suggest that quantitative analysis of the time-to-peak hot spot volume can be optimized for diagnostic performance providing indicators for differentiating benign versus malignant breast lesions.

Stefan Alexandru Carp¹, Christy M Wanyo¹, and David Alan Boas^1
1Radiology, Massachusetts General Hospital, Charlestown, MA, United States

Measurement of breast tissue response to external stimulation has the potential to unveil additional dynamic biomarkers of disease states. Recently tumor contrast generated by breathing-gas induced vascular changes has been demonstrated using both BOLD contrast and MRI-guided near infrared diffuse optical tomography. At the same time, temporally resolved hemodynamic changes due to external compression are being investigated as breast cancer biomarkers. We have developed an integrated MRI-optical breast compression platform to enable simultaneous dynamic optical and MRI acquisition of the post-compression dynamics. We describe the instrumentation development and report initial dynamic MRI results from healthy compressed breast tissue.

Felipe Franco¹, Fiona Fennessy¹, Andriy Fedorov¹, Kemal Tuncali¹, Junichi Tokuda¹, Sandeep Gupta², and Clare Tempany^1
1Radiology, Brigham and Women's Hospital, Boston, MA, United States, ²Functional Imaging Lab, General Global Research Center

To evaluate the feasiblity by means of image registration techniques, we were able to identify the location of the specimen collected during the biopsy.Moreover,correlate the histological analysis outcome with mpMRI by analyzing ROC curves, in a retorspectively study.

Maysam Jafar¹, Veronica A Morgan¹, Sharon Giles¹, David J Collins¹, Maria A Schmidt¹, and Nandita M deSouza^1
1Clinical MR, Institute of Cancer Research & Royal Marsden NHS Foundation Trust, Belmont, Sutton SM2 5NG, United Kingdom

Diffusion weighted imaging (DWI) has been used in both clinical and research settings for detecting cancer-related disease. It is particularly beneficial in prostate cancer tumour detection. In this paper spatial geometric distortion arising from the air-tissue interface of an air-filled balloon endorectal coil vs a PFC filled one in DWI of the prostate is discussed. Regions of interest are drawn around a T2W image and a DW image to quantify the amount of spatial distortion. Our initial results indicate that inflating the endorectal coil with PFC in DWI prostate examinations results in a greater confidence interval for slices in the middle and in the base of the prostate but not at the prostate apex.

Guang Jia¹, Ronney Abaza², Joanna D Williams³, Debra L Zynger³, Jinyuan Zhou⁴, Zarine K Shah¹, Mitva Patel¹, Steffen Sammet¹, Lai Wei⁵, Robert R Bahnson², and Michael V Knopp^1
1Wright Center of Innovation in Biomedical Imaging and Department of Radiology, The Ohio State University, Columbus, OH, United States, ²Department of Urology, The Ohio State University, Columbus, OH, United States, ³Department of Pathology, The Ohio State University, Columbus, OH, United States, ⁴Department of Radiology, Johns Hopkins University, Baltimore, MD, United States, ⁵Center for Biostatistics, The Ohio State University, Columbus, OH, United States

APT-MR imaging was first applied to the prostate in this feasibility study and the initial experience from 12 prostate cancer patients reveals increased MR imaging-detectable mobile proteins in cancerous regions of the prostate.

Yi Xie^1,2, Yi Dang¹, Feiyu Li³, Bing Fan³, Ling Yang², Jue Zhang^1,4, Xiaoying Wang^1,3, and Jing Fang^1,4
1Academy for Advanced Interdisciplinary Studies, Peking University, Beijing, Beijing, China, People's Republic of, ²Dept. of Electronic Engineering, Chengdu University of Information Technology, Chengdu, Sichuan, China, People's Republic of, ³Dept. of Radiology, Peking University First Hospital, Beijing, Beijing, China, People's Republic of, ⁴College of Engineering, Peking University, Beijing, Beijing, China, People's Republic of

Early detection of prostate cancer is urgently needed to improve patient¡¯s survival rate. In recent years, limited studies have been reported to evaluate MR based suspected prostate lesions by utilizing the prostate T2WI features. In fact, bilateral symmetry exists in normal prostate, and as for a developing lesion, asymmetry gradually appear with symmetry breaking down. In this study, texture bilateral symmetry of prostate T2W Images is attempted to screen efficiently the suspected prostate tumor. The result shows a statistical significant difference between cancerous and the control group. The symmetric feature of prostate should be useful for improving prostate cancer screening.

Daniel Hausmann¹, Simon Konstandin², Stefan Haneder¹, Frank Zoellner², Friedrich Wetterling², Stefan O. Schönberg¹, Dietmar J. Dinter¹, and Lothar R. Schad^2
1Institute of Clinical Radiology and Nuclear Medicine, University Medical Center Mannheim, Mannheim, Baden-Württemberg, Germany, ²Computer Assisted Clinical Medicine, Heidelberg University, Mannheim, Germany

To evaluate the image quality of sodium MRI of the prostate in a clinical setting in at 3 T and to determine if quantification of sodium levels within the gland is feasible. Initial results of eight healthy volunteers. DWI- and T2- images with a H*-body-coil are acquired additionally for correlation of the findings. A mean sodium concentration of 63 mmol/l in the central and 70 mmol/l in the peripheral zone was measured, so far.This study suggests that sodium imaging on a clinical 3T scanner might be an appropriate, noninvasive method for imaging of the human prostate.

Chad R Haney¹, Xiaobing Fan¹, Garima Agrawal¹, Charles A Pelizzari², Gregory S Karczmar¹, Jonathan Bakst³, Tatjana Antic³, Scott E Eggener⁴, and Aytekin Oto^1
1Radiology, University of Chicago, Chicago, Illinois, United States, ²Radiation & Cellular Oncology, University of Chicago, Chicago, Illinois, United States, ³Pathology, University of Chicago, Chicago, Illinois, United States, ⁴Urology/Surgery, University of Chicago, Chicago, Illinois, United States

Histology is the gold standard for assessing the presence or absence of cancer in prostatectomy specimen. However, the entire prostate cannot be processed and the histological processing often damages the specimen. High-resolution ex vivo MRI of the prostate can not only serve as a bridge to registering in vivo MRI with histology but can provide additional diagnostic information. We present preliminary results from 8 patients that suggests date suggest that high resolution T2-weighted ex vivo MRI differentiates between carcinoma and benign prostatic hyperplasia.

Martijn Gerjan Schouten¹, N A Nagel¹, Thomas Hambrock¹, Caroline M Hoeks¹, Joyce Gerda Riek Bomers¹, Pieter C Vos¹, and Jurgen J Futterer^1
1Radiology, Radboud University Nijmegen Medical Centre, Nijmegen, Gelderland, Netherlands

Purpose of this study was to investigate the correlation between diffusion weighted imaging and histopathology obtained during the same MRI guided biopsy session. Apparent diffusion coefficient (ADC) maps were merged with images obtained after needle insertion to determine the ADC-value of the biopsy specimen. Two-tailed t-tests were performed to determine differences between groups. Significant differences were considered at p <.05. The mean ADC-values of men with malignant and non-malignant prostate cancer were significant different. Significant differences were seen between prostatitis and cancer and between healthy specimen and prostatitis. No significant difference was seen between high- and low- grade cancer.

Stephan E. Maier¹, Yi Tang¹, Lawrence P. Panych¹, Robert V. Mulkern², and Clare M Tempany^1
1Radiology, Brigham and Women's Hospital, Harvard Medical School, Boston, MA, United States, ²Radiology, Children's Hospital, Harvard Medical School, Boston, MA

MR diffusion imaging with b-factors up to 3500 s/mm2 was performed in patients with biopsy proven prostate cancer. The diffusion signal decay measured over this extended range was clearly non-monoexponential. Biexponential fitting of the diffusion signal decay in normal and tumorous peripheral zone regions revealed fast and slow diffusion coefficients that are not distinctly different. The difference in diffusion signal decay between normal and cancer tissue, which is also evident at lower b-factors, can rather be attributed to a shift in diffusion compartment size fractions, with cancer tissue exhibiting a much higher slow diffusion compartment size fraction than normal tissue. _¿¿¿¿

Yu-Jen Chen¹, Y-S Pu², Woei-Chyn Chu¹, and W-Y Isaac Tseng^3,4
1Institute of Biomedical Engineering, National Yang-Ming University, Taipei, Taipei, Taiwan, ²Department of Urology, ³Department of Medical Imaging, National Taiwan University Hospital,⁴Center for Optoelectronic Biomedicine, National Taiwan University College of Medicine, Taipei, Taipei, Taiwan

In this study, we perform a retrospective analysis of diffusion tensor imaging in regions containing prostate cancer. Tumor volume (TV), trace apparent diffusion coefficient tADC), and combined parameters were calculated to correlate with gleason score (GS) to determine an index with stronger correlation with GS. Our results are similar to previous reports showing moderate correlation between tADC value and GS (r=-0.61). TV and average radius (R) are also found having moderate correlation with GS (r=0.62, 0.64, respectively). Two combining parameters, tADC/TV and tADC/R, were calculated to correlate with GS and were found having stronger correlations with GS (tADC/TV: r=-0.69; tADC/R: r=-0.77). The parameter of tADC/R might reduce the weighting of tumor size and balance the weightings of tADC and TV to have a better correlation with GS. In conclusion, among the parameters of tADC and TV in pathologically confirmed PCA regions. We found that a combined parameter of tADC/R had a best correlation with GS.

Miriam Lagemaat¹, Thiele Kobus¹, Stephan Orzada², Andreas Bitz², Arend Heerschap¹, and Tom Scheenen^1,2
1Radiology, Radboud University Nijmegen Medical Centre, Nijmegen, Netherlands, ²Erwin L. Hahn Institute for Magnetic Resonance Imaging, Essen, Germany

The phosphorus metabolites present in the prostate and surrounding tissues of a healthy volunteer were studied by ³¹P magnetic resonance spectroscopic imaging at 7 Tesla with an endorectal transceiver coil. The spectral resolution was high enough to distinguish the various compounds in the phoshpomonoester (PME) and phosphodiester (PDE) region. Separate resonances in the range of inorganic phosphate probably originate from compartments with different pH, but may arise from other compounds as well.

Jan Weis¹, Francisco Ortiz-Nieto¹, and Håkan Ahlström^1
1Uppsala University Hospital, Uppsala, Sweden

Differences in metabolite-to-Cit ratio between peripheral zone and central gland as well as age dependent changes are important factors in analysis of the prostate spectra. So far, age dependent effects have been quantified only in 1.5 T scanners. In this work, zonal and age dependences of metabolite (Cho, PA, Cr) to Cit ratio of 51 healthy volunteers were measured in 3T using surface coil and 2D MRSI. The results of this study demonstrate that zonal and age differences in metabolite-to-Cit ratio are significant and have to be considered in the interpretation of prostate spectra.

Angel Moreno-Torres¹, Antonia Blanch², Cesar Arribas³, Jose-Maria Gil-Vernet Sedò⁴, and Isidro Bonet-Palau^5
1Research Department, Cetir Grup Medic, Esplugues de Llobregat, Spain, ²Unitat Clinica El Pilar, Cetir Grup Medic, Barcelona, ³Unitat Clinica El Pilar, Cetir Grup Medic, Barcelona, Spain, ⁴Centro Gil-Vernet de Urología, Centro Médico Teknon, Barcelona, Spain, ⁵Barcelona, Spain

Although cancer detection in the prostate’s central gland by means of proton MRS and MRI is a difficult issue, spectra with altered MRS are often found. In this ongoing study, we aimed to find a correlation between the observed alterations and the presence of cancer. Results showed that, in the studied group with MRS suggesting malignancy, there were both cancerous and non-cancerous prostates and, therefore, the seek correlation was only partial. Such results and the real percentages of cancer true and false positives are valuable data in a clinical setting for the interpretation of MRS in the prostate’s central gland.

Gregory John Metzger¹, Eddie J Auerbach¹, and Gregor Adriany^1
1Center for Magnetic Resonance Research, University of Minnesota, Minneapolis, MN, United States

High resolution three-dimensional spectroscopic imaging data were obtained at 7T with a receive only endorectal coil combined with a transcieve surface array.

Elita DeFeo¹, Johannes Kurth¹, Chin-Lee Wu², Shulin Wu², W. Scott McDougal³, and Leo L. Cheng^4
1Pathology, Massachusetts General Hospital, Charlestown, MA, United States, ²Pathology, Massachusetts General Hospital, ³Urology, Massachusetts General Hospital, ⁴Radiology, Pathology, Massachusetts General Hospital

Our 2005 publication on the investigations of human prostate cancer meabolomics led to two major conclusions that metabolomic profiles can differentiate tissue specimens with and without cancer glands, as well as patient pathological stages. However, this study only proposed the metabolomic structure from the analysis of a training cohort and did not evaluate another independent testing cohort. In this study we report our analysis of a testing cohort that was measured precisely following the exact experimental procedures that guided the 2005 study for the training cohort.

Martijn Gerjan Schouten¹, Joyce Gerda Riek Bomers¹, Derya Yakar¹, Henkjan Huisman¹, Tom W J Scheenen¹, Sarthak Misra², and Jurgen J Futterer^1
1Radiology, Radboud University Nijmegen Medical Centre, Nijmegen, Gelderland, Netherlands, ²Control Engineering Group, MIRA-Institute of Biomedical Technology and Technical Medicine, Enschede, Overijssel, Netherlands

The purpose of this study was to evaluate the accuracy and speed of a novel pneumatically controlled magnetic field compatible manipulator as an aid to perform magnetic resonance image (MRI)-guided biopsies on patients with cancer suspicious regions in the prostate. 3D MR images were made with the needle inserted to determine target displacement, targeting and biopsy error. Currently, the robotic technique for transrectal real-time MR-guided prostate biopsies did not outperform the manual technique. This study provided insight into reasons for target motion during a biopsy procedure. Our results suggest that most target displacement is caused by needle insertion.

Sabrina Doblas¹, Mathilde Wagner^1,2, Jean-Luc Daire^1,2, Nathalie Haddad^1,2, Helena Leitao^2,3, Ralph Sinkus¹, Bernard E Van Beers^1,2, and Valérie Vilgrain^1,2
1Centre de Recherche Biomédicale Bichat-Beaujon, INSERM U773, Clichy, France, ²Department of Radiology, Beaujon University Hospital, University Paris Diderot, Clichy, France, ³Department of Radiology, Hospitais de Universidade de Coimbra, Coimbra, Portugal

The inter-observer reliability and the inter-acquisition repeatability of a bi-exponential, multi-b diffusion-weighted imaging (DWI) approach were investigated in 13 hepatic tumors. Average measurements of perfusion fraction, pure diffusion fraction, perfusion coefficient, pure diffusion coefficient as well as conventional apparent diffusion coefficient (ADC) were performed by 2 independent observers for 2 successive DWI acquisitions. A maximal overall variation in DWI measurement of 32% was observed for the most reliable and repeatable parameters, whereas the perfusion coefficient appeared to be highly unreliable. Improved DWI acquisitions limiting respiratory motion artefacts could potentially lead to more reliable measurements.

Sabrina Doblas¹, Mathilde Wagner^1,2, Jean-Luc Daire^1,2, Nathalie Haddad^1,2, Helena Leitao^2,3, Ralph Sinkus¹, Valérie Vilgrain^1,2, and Bernard E Van Beers^1,2
1Centre de Recherche Biomédicale Bichat-Beaujon, INSERM U773, Clichy, France, ²Department of Radiology, Beaujon University Hospital, University Paris Diderot, Clichy, France, ³Department of Radiology, Hospitais de Universidade de Coimbra, Coimbra, Portugal

Conventional apparent diffusion coefficient measurements can identify tumor malignancy but fail to determine tumor types. This classical approach was compared to a bi-exponential scheme able to extract microcirculation perfusion and pure molecular diffusion information in a study of characterization of human hepatic tumors on a cohort of 89 patients. A more complete, bi-exponential approach did not offer any added value with respect to the distinction between benign and malignant lesions. However, perfusion and pure diffusion fractions helped to differentiate metastases as a specific tumor subgroup, showing the advantage of using such an approach in specific cases.

Robert F Hanna¹, Stephen M Lagana², Roger K Moreira², Jean C Emond³, Inna Postolov¹, and Martin R Prince^1,4
1Diagnostic Radiology, Columbia University - New York Presbyterian Hospital, New York, NY, United States, ²Pathology, Columbia University - New York Presbyterian Hospital, ³Surgery, Columbia University - New York Presbyterian Hospital, ⁴Diagnostic Radiology, Weill Cornell Medical Center, New York, NY, United States

Accuracy of MRI and triple phase CT for detecting hepatocellular carcinoma (HCC) in cirrhosis was investigated in 107 patients undergoing liver tranplantation utilizing histopathologic analysis of explant as the reference standard. MRI is significantly more accurate than CT, especially for small HCC with a per patient accuracy of 91% and a sensitivity for detecting < 1cm lesions of 80%.

E.G.W. ter Voert¹, L. Heijmen², W.J.G. Oyen³, J.H.W. de Wilt⁴, C.J.A. Punt², L.F. de Geus-Oei³, H.W.M. van Laarhoven², and A. Heerschap^1
1Radiology, Radboud University Nijmegen Medical Centre, Nijmegen, Netherlands, ²Medical Oncology, Radboud University Nijmegen Medical Centre, Nijmegen, Netherlands, ³Nuclear Medicine, Radboud University Nijmegen Medical Centre, Nijmegen, Netherlands, ⁴Surgery, Radboud University Nijmegen Medical Centre, Nijmegen, Netherlands

There is an increasing interest in magnetic susceptibility or T₂* MR imaging to replace or complement DCE-MRI for vascular assessment of therapeutic interventions in early treatment response monitoring, as this circumvents the need for a contrast agent. In this study we focused on the reproducibility of T₂* MR imaging in patients with liver metastases of colorectal cancer. Histogram analysis were performed and reproducibility was assessed for the mean T₂*, the 16^th and 84^th percentile values. On average T₂* showed good mean and histogram reproducibility. A trend towards a negative correlation with mean ADC values was observed.

Linda Heijmen¹, Arend Heerschap², Edwin ter Voert², Wim Oyen³, Hans de Wilt⁴, Cees Punt⁵, Lioe-Fee de Geus- Oei³, and Hanneke van Laarhoven^1
1Medical Oncology, Radboud University Nijmegen, Medical Centre, Nijmegen, Gelderland, Netherlands, ²Radiology, Radboud University Nijmegen, Medical Centre, Nijmegen, Netherlands, ³Nuclear Medicine, Radboud University Nijmegen, Medical Centre, ⁴Surgery, Radboud University Nijmegen, Medical Centre, ⁵Medical Oncology, Radboud University Nijmegen, Medical Centre

Diffusion weighted imaging (DWI) is a technique that shows potential for early response monitoring of cancer therapies. For the assessment of the value of DWI for early response monitoring in metastasized colorectal cancer knowledge of reproducibility study is crucial. In this study, the reproducibility of DWI and FDG-PET was investigated in liver metastases of colorectal cancer. A good reproducibility (coefficient of reproducibility 0.27) for the mean ADC was obtained. The distribution of ADC in a histogram also showed good reproducibility. A trend towards a negative correlation with mean standardized uptake values of FDG was found.

Renu M. Stephen¹, Denise J. Roe^1,2, Abhinav K. Jha³, Haiyan Cui¹, Georgette Frey¹, Scott Squire⁴, Ted P. Trouard^4,5, Jean P. Galons⁴, Jeff J. Rodriquez⁶, Mathew A. Kupinski^3,4, Eric Outwater⁷, Robert J. Gillies⁷, and Alison T. Stopeck^1
1Arizona Cancer Center, University of Arizona, Tucson, AZ, United States, ²Mel and Enid Zuckerman College of Public Health, University of Arizona, Tucson, ³College of Optical Sciences, University of Arizona, Tucson, ⁴Department of Radiology, University of Arizona, Tucson, ⁵Biomedical Engineering, University of Arizona, Tucson, ⁶Electrical and Computer Engineering, University of Arizona, Tucson, ⁷H.Lee Moffitt Cancer Center and Research Institute, Tampa, Florida

The purpose of this study was to optimize a protocol for obtaining reproducible results for DW-MRI of liver metastases. ADC values measured in metastatic liver lesion from breast cancer are affected by both imaging and tumor parameters. We investigated choice of b values, lesion segmentation, baseline ADC, lesion size, and the number of prior treatments as possible modulators of chemotherapy response. Our results indicate that prior treatment history and low pre-treatment ADC values analyzed using an appropriate choice of b value aid in predicting tumor response.

Gillian Macnaught¹, Fat-Wui Poon¹, S Viswanathan¹, Y-T Sim², and M Digby^2
1Radiology Department, Glasgow Royal Infirmary, NHS Greater Glasgow and Clyde, Glasgow, Scotland, United Kingdom, ²Radiology Department, Stobhill Ambulatory Care Hospital, NHS Greater Glasgow and Clyde, Glasgow, Scotland, United Kingdom

Two radiologists assessed the lymph nodes of rectal cancer patients on T2 Turbo Spin Echo (TSE) images. Nodes that had an irregular boundary and inhomogeneous signal were assigned as suspicious of malignancy. The nodes visualised on the TSE images were identified on the corresponding Diffusion Weighted (DW) images and Apparent Diffusion Coefficients (ADCs) were calculated. An unpaired t-test indicated a significant difference (p-value 0.04) in the mean ADC values of nodes agreed by both radiologists to be either benign (0.97±0.17mm2/s) or suspicious (0.77±0.08 mm2/s). This indicates the potential of DWI to determine the malignancy of nodes in rectal cancer patients.

Geoff Charles-Edwards^1,2, Jyoti Parikh¹, Nyree Griffin¹, Robert Johnstone¹, David Landau¹, and Andrew Gaya^1
1Guy's & St Thomas' NHS Foundation Trust, London, United Kingdom, ²King's College London, London, United Kingdom

This work reports on initial reproducibility results from dynamic contrast enhanced MRI of rectal cancers at 3T, and compares results from patient with and without the administration of a bowel relaxant. The significant effect of bowel motion on DCE-MRI data and subsequent PK modelling is demonstrated.

Olivia Moens¹, Julie Absil¹, Anne Demols², Thierry Metens¹, and Celso Matos^1
1Magnetic Resonance-Radiology, Université Libre de Bruxelles Hôpital Erasme, Brussels, Belgium, ²Gastroenterology, Université Libre de Bruxelles Hôpital Erasme, Brussels, Belgium

To determine if quantitative diffusion imaging can assess response to neoadjuvant therapy (NAT) in advanced rectal cancer. TSET2 and Diffusion images were obtained at 3T pre-and-post NAT, Response to NAT was given by Dworak pathologic tumor regression grade. Tumor Volume decreased and tumoral ADC increased in all patients, however Tumour Volume variation and ADC variation did not significantly differ between responders and non-responders, ADC and f were not significantly different between groups before-and-after NAT, these variables did not correlate with the Dworak grade. The pretreatment tumor volume correlated with the Dworak grade and was a good indicator of the response.

Anne-Catherine Fruytier¹, Julie Magat¹, Benedicte F Jordan¹, Gregory Cron², and Bernard Gallez^1
1Louvain Drug Research Institute, Biomedical Magnetic Resonance Research Group, University of Louvain, Brussels, Belgium, ²Ottawa Health Research Institute, Ottawa, Canada

DCE-MRI allows a non invasive assessment of the tumor hemodynamics. For the determination of quantitative parameters through pharmacokinetic modelling, the arterial input function (AIF) should be measured accurately. We observed that, at 11.7 Tesla, T2* relaxation has a major effect on arterial signal, causing magnitude imaging to fail to provide a proper measurement of the AIF. We therefore focused our work on characterization of the AIF using phase imaging. To validate our phase technique, we conducted an in vitro study with a phantom consisting of tubing mimicking the iliac artery of the mice (where the AIF is measured in vivo). We measured the phase shift (Ư) as a function of different concentrations of Gd-DOTA

Jacob U. Fluckiger¹, Xia Li¹, Jennifer Whisenant¹, Lei Xu¹, Junzhong Xu¹, Todd E. Peterson¹, John C. Gore¹, and Thomas Yankeelov^1
1Institute of Imaging Science, Vanderbilt University, Nashville, TN, United States

Quantitative PET modeling is made difficult by the inability to separate tissue time activity curves into their constituent compartments. DCE-MRI modeling could be used in conjunction with dynamic PET imaging to overcome this limitation. The fractional volume parameters measured from a DCE-MRI scan can be used to separate tissue time activity curves and simplify PET modeling.

Hee Kwon Song¹, Yiqun Xue¹, Jiangsheng Yu¹, Sarah Englander¹, Hyunseon C Kang¹, and Mark A Rosen^1
1Department of Radiology, University of Pennsylvania, Philadelphia, PA, United States

Pre-contrast signal levels in DCE-MRI datasets are often low due to the short TRs used to achieve high temporal resolution. Since this baseline signal is required to normalize the entire dynamic curve in computing tissue gadolinium concentration, low SNR can cause biases in the measurement of tissue perfusion. Recently, it was shown that the perfusion parameters could be computed more directly using absolute signal differences without the need for conversion to gadolinium concentration. In the current work, the performance of the signal difference method is compared to conventional methodologies for computing tissue perfusion in the presence of various noise levels.

Anita Banerji^1,2, Josephine H Naish^1,2, Giovanni A Buonaccorsi¹, and Geoff J M Parker^1,2
1Imaging Sciences and Biomedical Engineering, Manchester University, Manchester, United Kingdom, ²Biomedical Imaging Institute, Manchester, United Kingdom

Estimation of tracer kinetic model parameters from DCE-MRI data typically relies on the use of an optimisation algorithm for fitting the model to time course data and for estimating T₁ from variable flip angle images. We compare the precision and accuracy of model parameters estimated using different frequently employed algorithms and demonstrate that choice of algorithm has little influence on K^trans. We also show that estimating T₁ by linearising the pulse sequence equation does not lead to a loss of precision or accuracy of the estimated K^trans values.

Nadine El Tannir El Tayara^1,2, Nidhal Ben Achour^1,2, Christine Walczak^1,2, Fariba Nemati³, Joel Mispelter^1,2, Didier Decaudin³, Julien Vautier^1,2, and Andreas Volk^1,2
1Research Center, Institut Curie, Orsay, France, ²U759 INSERM, Orsay, France, ³Translational Research Department, Institut Curie, Paris, France

The purpose of this work was to evaluate a new quantitative DCE-MRI technique combining 3D tumor imaging with interleaved high time resolution 2D heart imaging (AIF) in a context of treatment follow-up. We investigated its ability to assess early change of microvascular parameters under anti-angiogenic therapy (Sunitinib®) and showed a histograms’ K^trans distribution shift towards lower values at day 4 after treatment initiation. Significant decrease in K^trans median and mode values was observed, as well as tumor growth inhibition over the 11 days treatment period. Results were compatible with a predictive value of these Ktrans parameters for anti-angiogenic treatment response.

Penny Louise Hubbard^1,2, Josephine H Naish^1,2, Caleb Roberts^1,2, Yvonne Watson^1,2, Karen Davies^1,2, John C Waterton^1,3, Helen Young³, John P Logue⁴, M Ben Taylor⁴, and Geoff J. M. Parker^1,2
1Imaging Sciences and Biomedical Engineering, The University of Manchester, Manchester, United Kingdom, ²The Biomedical Imaging Institute, The University of Manchester, Manchester, United Kingdom, ³AstraZeneca R & D, Alderley Park, Macclesfield, Cheshire, United Kingdom, ⁴Christie Hospital, Manchester, United Kingdom

The aims of this study were: 1) to assess the statistical justification for fitting complex time series models over simpler kinetic models (e.g. AATH versus extended Kety model) to DCE-MRI and DCE-CT data; and 2) to determine if spatial smoothing of DCE-MRI and DCE-CT data improves the robustness of AATH model fitting. Simple smoothing was not sufficient to allow the AATH model to be fit to the DCE-CT reliably. Smoothing of DCE-MRI data leads to a higher probability that the AATH model is effective in describing the data, when compared with the extended Kety, for a similar effective voxel size.

Jiangsheng Yu¹, Yiqun Xue¹, Mark A Rosen¹, Christina S Chu², and Hee Kwon Song^1
1Department of Radiology, University of Pennsylvania, Philadelphia, PA, United States, ²Obstetrics and Gynecology, University of Pennsylvania, Philadelphia, PA, United States

In current DCE-MRI protocols for tumor assessment, the excitation flip angle is typically assumed to be exactly known and homogeneous throughout the imaging FOV. However, it is known that the angle could vary substantially due to inhomogeneous RF fields and slice profile effects. In this work, we evaluate the degree to which flip angle deviations can occur in a large FOV DCE-MRI protocol and determine how those variations affect the measurement accuracy of perfusion parameters Ktrans and ve.

Éric Poulin¹, Étienne Croteau¹, Réjean Lebel¹, Luc Tremblay¹, Roger Lecomte¹, M'Hamed Bentourkia¹, and Martin Lepage^1
1Nuclear Medicine and Radiobiology, Université de Sherbrooke, Sherbrooke, Quebec, Canada

The purpose of this study was to evaluate the correspondence between the AIF derived from MRI and PET, and to determine whether the AIF obtained by one modality can be converted into the AIF for the other modality. We showed that Gd-DTPA and ¹⁸F-FDG AIF were not identical but that using a bi-exponential model, we were able to convert Gd-DTPA AIF to ¹⁸F-FDG AIF, andvice versa. No statistical difference was found between area under the curve of experimental and converted curves. Results suggest that it would be possible to perform dual kinetic modeling using MRI and PET.

Caleb Roberts^1,2, Sarah Hughes³, Josephine H Naish^1,2, Katherine Holliday^1,2, Yvonne Watson^1,2, Sue Cheung^1,2, Giovanni A Buonaccorsi^1,2, Helen Young⁴, Noel Clarke^3,5, and Geoff JM Parker^1,2
1Imaging Science and Biomedical Engineering, The University of Manchester, Manchester, Greater Manchester, United Kingdom, ²Biomedical Imaging Institute, The University of Manchester, Manchester, Greater Manchester, United Kingdom, ³Paterson Institute for Cancer Research, The University of Manchester, Manchester, Greater Manchester, United Kingdom, ⁴AstraZeneca, Macclesfield, Cheshire, United Kingdom, ⁵Department of Urology, Salford Royal Hospital NHS Foundation Trust, Salford, Greater Manchester, United Kingdom

An important influence on DCE-MRI tracer kinetic parameters that is commonly overlooked is the hematocrit (Hct), which relates the measured whole blood contrast agent concentration to the blood plasma contrast agent concentration, thereby influencing the arterial input function, K^trans and v_p. We investigate the errors in these parameters when using an assumed Hct rather than an individually measured Hct, which in this study was obtained at each DCE-MRI visit. Our findings have important implications for clinical trials where a treatment effect may be masked due to unaccounted for fluctuations in patients’ Hct throughout a course of therapy.

Ina Nora Kompan^1,2, Claudia Prieto³, Benjamin Richard Knowles⁴, Hendrik Laue¹, Geoff Charles-Edwards³, Matthias Guenther^1,2, and Tobias Schaeffter^3
1Fraunhofer MEVIS-Institute for Medical Image Computing, Bremen, Germany, ²Faculty of Physics and Electronics, University of Bremen, Bremen, Germany, ³Division of Imaging Sciences, Kings's College London, St. Thomas' Hospital, London, United Kingdom, ⁴Cardiovascular Division, Beth Israel Deaconess Medical Center, Harvard School of Medicine, Boston, MA, United States

For quantitative evaluation of dynamic contrast-enhanced MRI data, high temporal resolution is required, however high spatial resolution is needed to depict variations within a lesion. To cover both requirements, signal-adaptive schemes were investigated, which acquire images with high temporal resolution during uptake of contrast agent and high spatial resolution during washout. These schemes were tested using a numerical phantom. During uptake, different schemes were employed to speed up imaging. It could be shown that signal-adaptive schemes improve the standard deviation of the results, but produce systematic errors. The best result could be attained using a modified version of TRICKS.

Leonidas Georgiou^1,2, and Chris James Rose^1,2
1The University of Manchester Biomedical Imaging Institute, The University of Manchester, Manchester, Greater Manchester, United Kingdom, ²Manchester Academic Health Science Centre, The University of Manchester, Manchester, Greater Manchester, United Kingdom

Tracer kinetic model parameters typically have physiological interpretations, and constraints exist between them. While most fitting routines can impose constraints on parameters individually, this is not so for inter-variable constraints. Irrespective of the accuracy of the model itself, non-physiological parameter estimates will result in erroneous inferences being drawn about tumour microvascular characteristics. We study the frequency with which non-physiological fits occur in clinical data, propose a method with which those constraints can be imposed and, using simulations, compare the proposed solution to the conventional approach. We show that imposing inter-parameter constraints can result in more accurate and precise parameter estimates.

Stephanie Lynne Barnes^1,2, John C Gore^1,2, and Thomas E Yankeelov^1,2
1Radiology and Radiological Sciences, Vanderbilt University, Nashville, TN, United States, ²Institute of Imaging Sciences, Vanderbilt University, Nashville, TN, United States

DCE-MRI allows for the estimation of tumor tissue parameters by modeling contrast agent kinetics using the standard (Toft’s) model. However, cases exist in which the analysis results in parameter values that are not physically relevant. It has been hypothesized that the reason for this is diffusion that occurs in poorly perfused tissues. To investigate the possible effect of diffusion, we developed a 2D finite element model of CA concentration in the tissue as described by diffusion and CA delivery through a centralized vessel. Results indicate that diffusion effects can account for the physically irrelevant parameter estimations of the standard model.

Steven Sourbron¹, and David L Buckley^1
1Division of Medical Physics, University of Leeds, Leeds, United Kingdom

The Tofts Model (TM) and Extended Tofts Model (ETM) have become a standard for DCE-MRI analysis in oncology, but the scope of the models has never been identified rigourously. In this study, necessary and sufficient conditions are identified for which they can be applied. The TM should only be used if prior knowledge is available which guarantees that the tissue is weakly vascularised. The ETM may also be used in tissues that are known to be highly perfused. In all other conditions, TM and ETM do not produce accurate values - even if they offer a good fit to the data.

Reshmi Rajendran¹, Jie Ming Liang¹, Torsten Reese², Hannes Hentze², Susan van Boxtel², Brian Henry², and Kai-Hsiang Chuang^1
1Magnetic Resonance Imaging Group, Laboratory of Molecular Imaging, Singapore Bioimaging Consortium, Singapore, Singapore, Singapore, ²Translational Medicine Research Centre, MSD, Singapore

Quantitative perfusion MRI in mouse tumor models has been challenging due to low flow and movement/susceptibility artifacts in the abdomen. We optimized Arterial Spin Labeling MRI to map low perfusion in lumbar muscle of mice and applied the optimized sequence to map perfusion in a renal carcinoma xenograft mouse model. The results indicate that a perfusion rate of 34±7 ml/100g/min can be measured in muscle, whereas in xenograft tumors transplanted in the flank of nude mice, a heterogeneous perfusion rate was observed, with values ranging between 10-280 ml/100g/min. This optimized method could potentially be used to non-invasively investigate the heterogeneity of tumor perfusion in xenograft models or quantify alterations in perfusion in response to therapy. The continued development and application of such translational methods has the potential to enable further investigation into the complex biological processes underlying tumor biology in experimental animal models and the clinic.

Natenael B Semmineh¹, Junzhong Xu¹, John C Gore¹, and Christopher C Quarles^1
1Radiology and Radiological Sciences, Vanderbilt University Institute of Imaging Science, Nashville, TN, United States

The use of DSC-MRI in tumors can be confounded by extravascular T₂* effects that occur when contrast agent leaks out of the vasculature. The resulting DSC-MRI signals are consequently influenced by the extravascular compartmentalization of the contrast agent, but we propose that such effects can be analyzed to extract imaging metrics sensitive to the spatial distribution of tumor cells within tissue (e.g cell density, intercellular distance). Using simulations as well as in vitro and in vivo studies we demonstrate that DSC-MRI tumor data are highly sensitive to tumor cell density.

Jin Zhang¹, Lindsey Decarlo², Robert Schneider², and Sungheon Kim^1
1Center for Biomedical Imaging, Radiology, New York University School of Medicine, New York, New York, United States, ²Microbiology, New York University School of Medicine, New York, New York, United States

Exchange of water between intracellular and interstitial compartments is an important factor to be considered in estimating physiologically relevant parameters from dynamic contrast enhanced (DCE)-MRI data using a pharmacokinetic model. While it has been demonstrated that intracellular water lifetime (_i) can be estimated from DCE-MRI data, it has not been shown whether the estimated _i can be compared with any other in vivo measurement method. In this study, we measured _i using both DCE-MRI and diffusion weighted imaging (DWI) in BALB/c mice with 4T1 mammary carcinoma, and found a good agreement between their estimates of intracellular water lifetime.

Christopher Chad Quarles¹, John Christopher Gore¹, and Thomas Edison Yankeelov^1
1Radiology and Radiological Sciences, Vanderbilt University, Nashville, TN, United States

We have previously demonstrated that dual-echo DSC-MRI can be used to separate and quantify T₁ and T₂* based contrast agent leakage effects and, when combined with a pre-contrast T₁map, quantify the DCE-MRI parameters, K^trans and v_e. To further verify this approach we now compare DSC-MRI and DCE-MRI derived K^trans and v_e values in two brain tumor models with dissimilar perfusion characteristics and contrast agent dynamics. The DSC-MRI and DCE-MRI parameters were highly correlated and in good agreement as no significant differences were found in either of the tumor models.

Jakub Palowski^1,2, and Chris James Rose^1,2
1The University of Manchester Biomedical Imaging Institute, The University of Manchester, Manchester, Greater Manchester, United Kingdom, ²Manchester Academic Health Science Centre, The University of Manchester, Manchester, Greater Manchester, United Kingdom

Nonlinear regression is commonly used in MRI to estimate physiological quantities that cannot be measured directly, but can be modelled as a function of measureable phenomena: e.g., modelling the tracer kinetics of contrast-agents as in DCE-MRI. We present a novel nonlinear regression algorithm that estimates the parameters of the extended Tofts model, based on a machine learning method called “random forests”. Using simulated gadopentate dimeglumine (Gd-DTPA) concentration time series we show that, compared to conventional least squares, the proposed method can estimate all model parameters with greater accuracy and comparable precision, and in less time.

Yann Jamin¹, Jessica KR Boult¹, Lauren C Baker¹, Simon Walker-Samuel², Arne Östman³, and Simon P Robinson^1
1CRUK and EPSRC Cancer Imaging Centre, The Institute of Cancer Research and Royal Marsden NHS Trust, Sutton, United Kingdom, ²UCL Centre for Advanced Biomedical Imaging, Department of Medicine and Institute of Child Health, University College London, United Kingdom, ³Cancer Center Karolinska, Karolinska Institutet, Stockholm, Sweden

Mature, more patent vasculature is predicted to have more complete endothelial lining, basal membrane and smooth muscle and hence reduced permeability. The latter has not been yet unequivocally demonstrated in vivo. We have used DCE-MRI to interrogate tumours derived from B16 melanoma cells engineered to express PDGF-BB (B16/PDGF), which have previously been shown to exhibit a significantly more mature vasculature, compared to wild-type B16 melanomas (B16/ctr). Results show that PDGF/B16 tumours present a lower IAUCG₆₀ and T₁ than B16/ctr tumours, suggesting that B16/PDGF tumours present a reduced vascular permeability compared to B16/ctr tumours.

Jeff D Winter^1,2, Margarete K Akens³, and Hai-Ling Margaret Margaret Cheng^1,4
1Physiology and Experimental Medicine, The Hospital for Sick Children, Toronto, Ontario, Canada, ²Research and Development, IMRIS, Winnipeg, Manitoba, Canada, ³Orthopaedic Surgery, Sunnybrook Health Sciences, Toronto, Ontario, Canada, ⁴Medical Biophysics, University of Toronto, Toronto, Ontario, Canada

Hypoxic tumour regions exhibit reduced radiotherapy responses. One strategy to improve tumour O₂ levels is through breathing 100 % O₂ or carbogen gases. This study characterized T₁ and T₂* responses to 100 % O₂, and carbogen (with three CO₂ concentrations) in a VX2 tumour model in comparison to quantitative DCE and invasive pO₂ and perfusion measurements. We observed consistent T₁ decrease and T₂* increases (relative to air) for all gas challenges. Invasive measures demonstrated substantial pO₂ increases and moderate perfusion increases for all gas challenges. Study results suggest similar tumour oxygenation may be achieved with or without a CO₂ component.

Nilesh N Mistry¹, Jame Van Gambrell², and Christopher Chad Quarles^2
1Dept of Radiation Oncology, University of Maryland School of Medicine, Baltimore, MD, United States, ²Dept. of Radiology and Radiological Sciences, Vanderbilt University Institute of Imaging Sciences, Nashville, TN, United States

Nitrite undergoes reconversion to nitric oxide (NO) under conditions characteristic of the tumor microenvironment, such as hypoxia and low pH. This selective conversion of nitrite into NO in tumor tissue has led to the possibility of using nitrite to enhance drug delivery and radiation response. In this work we propose to serially characterize the vascular response of brain tumor bearing rats to nitrite using contrast-enhanced R2* mapping. Preliminary results indicate that nitrite induces tumor-specific vascular changes in C6 gliomas that are likely due to increased vascular permeability and iron oxide contrast-agent accumulation within the tumor tissue.

Giovanni Alessandro Buonaccorsi¹, Caleb Roberts¹, James PB O'Connor¹, Chris J Rose¹, Susan Cheung¹, Yvonne Watson¹, Alan Jackson², Gordon C Jayson³, and Geoff JM Parker^1
1ISBE, University of Manchester, Manchester, United Kingdom, ²WMIC, University of Manchester, Manchester, United Kingdom, ³Cancer Research UK Dept of Medical Oncology, Christie Hospital, Manchester, United Kingdom

Using DCE-MRI data from ten patients enrolled in a trial of the VEGF inhibitor bevacizumab we performed tracer-kinetic model-driven registration (TKMDR) followed by cross-visit tumour sub-segmentation. In 7 of the 9 evaluable post-registration data sets, TKMDR altered the number of retained principal components (PCs). We present an example in which a cluster in an unregistered segmentation result could have been misinterpreted as a treatment effect—we demonstrate the removal of the cluster by TKMDR and provide evidence that TKMDR improved the data structure in the PC space by eliminating patterns that were attributable to motion corruption.

Keiko Miyazaki¹, James A d'Arcy¹, Matthew R Orton¹, Dow-Mu Koh², David J Collins¹, and Martin O Leach^1
1CR-UK and EPSRC Cancer Imaging Centre, The Institute of Cancer Research and Royal Marsden Hospital, Sutton, Surrey, United Kingdom, ²Department of Radiology, Royal Marsden NHS Foundation Trust, Sutton, Surrey, United Kingdom

This study shows that initial T1 estimations of both lesions and blood can be improved using variable flip angle data acquired at the end of a DCE-MRI study where SNR is increased and long T1s are shortened as a result of contrast agent uptake. Improvements in the quantification of blood T1 values would enable more accurate vascular input function quantification. The reproducibility of initial lesion T1 estimates was also found to be better than for those obtained in the conventional manner. These improvements in the initial T1 estimations will positively affect the quantification of DCE-MRI parameters which are investigated in several clinical settings.

Yiqun Xue¹, Jiangsheng Yu¹, Mark A Rosen¹, Ramesh Rengan², Hyun Seon Kang¹, Sarah Englander¹, Rosemarie Mick³, and Hee Kwon Song^1
1Department of Radiology, University of Pennsylvania, Philadelphia, PA, United States, ²Department of Radiation Oncology, University of Pennsylvania, Philadelphia, PA, United States,³Department of Biostatistics and Epidemiology, University of Pennsylvania, Philadelphia, PA, United States

In DCE-MRI, perfusion parameters are particularly sensitive to the accuracy of the baseline (pre-contrast) signal of the AIF and tumor. However, the SNR of the pre-contrast signal can be low, since TR is typically much shorter than the T1 of blood and tissue. We describe a method to compensate for errors in the baseline DCE-MRI signal due to low SNR. The method involves utilization of a separate low flip angle image to better estimate the true baseline signal of the dynamic series. We demonstrate that the in vivo repeatability of Ktrans is improved with this technique.

Tammy Louise Kalber^1,2, Adrienne E Campbell¹, Katy L Parcell¹, Bernard M Siow¹, Anthony Neil Price^1,3, Simon Walker-Samuel¹, Quentin A Pankhurst⁴, Sam M Janes², and Mark F Lythgoe^1
1Centre for Advanced Biomedical Imaging, Division of Medicine and Institute of Child Health, University College London, London, United Kingdom, ²Centre for Respiratory Research, Department of Medicine, University College London, University College London, London, United Kingdom, ³Robert Steiner MRI Unit, Imaging Sciences Department, Hammersmith Hospital, Imperial College London, London, ⁴Davy-Faraday Research Laboratories, The Royal Institution of Great Britain, London, United Kingdom

A primed i.p. infusion of Gd.DTPA was used to assess enhancement in a MDA-MB-231 lung metastasis model using a double gated spin echo sequence. Only large tumours and diffuse areas of increased signal intensity were discernable on baseline scans. These tumours and diffuse small tumours not apparent on baseline scans showed clear enhancement after initial Gd.DTPA bolus/infusion. Small tumours become more delineated and extensive at the later time points when we would expect increased Gd.DTPA concentration and to reach equilibrium. These results show that primed infusions with a double gated spin echo sequences offer superior detection of metastases over conventional techniques.

Jori S. Carter¹, Navneeth Lakkadi², Jessica E. Kuehn-Hajder³, Isabelle V. Iltis², Levi S. Downs, Jr.⁴, and Patrick J. Bolan^2
1Obstetrics, Gynecology, and Women's Health, University of Minnesota, Minneapolis, Minnesota, United States, ²Radiology/CMRR, University of Minnesota, Minneapolis, MN, United States,³Radiology, University of Minnesota, Minneapolis, MN, United States, ⁴Obstetrics, Gynecology, and Women's Health, University of Minnesota, Minneapolis, MN, United States

Malignant ovarian masses can be differentiated from benign ovarian masses with the use of multiple, quantitative MR-based metrics at 3 Tesla, including T2, ADC, IAUC, and MRS. These preliminary results indicate that IAUC values from dynamic contrast-enhanced images and the detection of total choline-containing compounds by MRS can differentiate between malignant and benign masses. A larger population is needed to show if there is a difference for T2 values.

Mayumi Takeuchi¹, Kenji Matsuzaki¹, and Masafumi Harada^1
1Department of Radiology, University of Tokushima, Tokushima, Tokushima, Japan

We evaluated MR imaging of 30 small early stage cervical cancers at 3T by using diffusion-weighted imaging (DWI) and 3D-dynamic contrast-enhanced MRI. All 21 small visible cancers (1B1, 2A1) and 1 of 9 microscopic diseases (CIS-1A) showed early nodular or thickened mucosa-like enhancement exhibiting hypervascular cancerous foci, whereas 7 of 9 microscopic diseases showed early linear enhancement along the mucosa reflecting atypical vascular proliferation due to angiogenesis of cancer cells. All 21 small visible cancers and 6 of 9 microscopic diseases showed high intensity on DWI reflecting hypercellularity due to cancer cell proliferation.

Martine I Dujardin¹, Abdullah Aldosary¹, Peter Gibbs¹, Martin D Pickles¹, and Lindsay W Turnbull^1
1Centre for MR Investigations, University of Hull in association with Hull York Medical School, Hull, East Yorkshire, United Kingdom

This study investigates the ability of ADCs acquired at 3T to predict response to chemoradiotherapy in advanced cervical cancer (N=14). No significant difference between average ADC of responders (complete and partial response) and non-responders (stable and progressive disease) was found (p= 0.06). Mean group ADC values increased systematically with response improvement, supporting further study in a large cohort.