Cancer: Preclinical Studies of Animal Models

Tuesday 2 June 2015

Yu-Chun Lin¹, Gigin Lin¹, Chun-Chieh Wang², and Jiun-Jie Wang^3
1Department of Diagnostic Radiology, Chang Gung Memorial Hospital, Linkou, Taiwan, Taiwan, ²Department of Radiation Oncology, Chang Gung Memorial Hospital, Linkou, Taiwan, ³Department of Medical Imaging and Radiological Sciences, Chang Gung University, Yaoyuan, Taiwan

This study monitored the changes of microvasculature and hypoxia after radiotherapy in TRAMP tumor using carbogen challenging BOLD MRI. A single dose irradiation was delivered in peripheral half of the tumor. The BOLD response to carbogen in the irradiated portion decreased compared with the non-irradiated portion. The immunohistochemistry images of CD31 and pimonidazole showed less hypoxia in the irradiated portion, while a trend of acute hypoxia in the non-irradiated region was noticed to shift to chronic hypoxia in the irradiated region. Carbogen-challenging BOLD MRI can be used to detect the hypoxic change within a tumor in response to radiotherapy

Lu Jiang¹, Kamila Chughtai², Tiffany Greenwood¹, Zaver M. Bhujwalla¹, Venu Raman¹, Gert Eijkel², Ron Heeren², and Kristine Glunde^1
1Department of Radiology, Johns Hopkins University School of Medic, BALTIMORE, MD, United States, ²FOM-Institute AMOLF, Amsterdam, Netherlands

The total choline (tCho) signal in magnetic resonance spectroscopic imaging (MRSI) of tumors is a potential biomarker of breast cancer aggressiveness and can be visualized in vivo by H1 MRSI tCho maps. Mass spectrometry imaging (MSI) of histologic tumor sections is able to detect thousands of biomolecules, including tryptic peptides from on-tissue digested proteins, from the tissue surface. We have investigated differentially regulated proteins in high- versus low-tCho regions in a human breast cancer model by combining in vivo MRSI with ex vivo MSI, which identified specific protein species that are spatially correlated with tCho.

Ellen Ackerstaff¹, H. Carl LeKaye¹, Natalia Kruchevsky¹, Kristen L. Zakian¹, Nirilanto Ramamonjisoa¹, Ekaterina Moroz¹, Inna S. Serganova¹, Ronald G. Blasberg¹, and Jason A. Koutcher^1
1Memorial Sloan Kettering Cancer Center, New York, NY, United States

Mona Salehi Ravesh¹, Monika Huhndorf², Amir Moussavi¹, Kristin Koetz¹, Judith Becker¹, Kirsten Hattermann³, and Susann Boretius^1
1Clinic of Radiology and Neuroradiology, Section Biomedical Imaging, Kiel, Schleswig-Holstein, Germany, ²Department of Radiology and Neuroradiology, Schleswig-Holstein, Germany, ³Christian-Albrechts-University of Kiel, Anatomical Institute, Schleswig-Holstein, Germany

Amide proton transfer (APT) and the Nuclear Overhauser Enhancement (NOE) have emerged as potential new imaging contrasts. Here we applied these techniques on a glioma model in rats. Maps of the magnetization transfer rate, both at -3.5 ppm and +3.5 ppm, revealed a reduction of the MTR in the tumor. NOE-CEST however revealed additional information about the tumor boundaries. Thus, NOE-CEST MRI may confer specific benefits in the detection and characterization of glioblastoma in patients.

Hanna Maja Tunset¹, Eugene Kim¹, Jana Cebulla¹, Muhammad Riyas Vettukattil¹, Astrid Jullumstrø Feuerherm², Berit Johansen ², Tone Frost Bathen¹, and Siver Andreas Moestue^1
1MR Cancer Group, Department of Circulation and Medical Imaging, Norwegian University of Science and Technology, Trondheim, Norway, ²Avexxin AS, Department of Biology, Norwegian University of Science and Technology, Trondheim, Norway

The cPLA2IVA gene PLA2G4A is highly expressed in aggressive basal-like breast cancer (BLBC) and may be represent a new therapeutic target. 1H HR MAS MRS, µCT, and gene expression analysis were used to explore the effect of cPLA2IVA inhibition in BLBC xenografts. Treatment significantly inhibited tumor growth and induced early favorable metabolic changes (higher PCho, lower GPC), intermediate anti-angiogenic effects (smaller vessels, avascular regions), and late gene expression changes (up-regulated transcription and translation, possibly apoptosis). Thus, cPLA2IVA inhibition may represent a novel strategy for targeted treatment of BLBC

Richard Mair^1,2, Alan Wright¹, Kieren Allinson³, Tiago Rodrigues¹, Colin Watts², and Kevin Brindle^1
1CRUK Cambridge Institute, University of Cambridge, Cambridge, Cambridgeshire, United Kingdom, ²Division of Neurosurgery, University of Cambridge, Cambridge, Cambridgeshire, United Kingdom, ³Department of Pathology, Cambridge University Hospitals NHS Foundation Trust, Cambridgeshire, United Kingdom

Glioblastoma is the most common human-primary-intracranial neoplasm and has the worst prognosis. Novel orthotopic patient-derived xenograft models that recapitulate the pathophysiological phenotype of the disease allow development of novel imaging modalities for downstream clinical applications in treatment planning and monitoring. We demonstrate that ¹³C magnetic resonance spectroscopic imaging with hyperpolarized [1-¹³C]pyruvate shows the presence of tumor through an elevated lactate/pyruvate ratio in both contrast enhancing and non-enhancing brain tumor lesions including areas where tumor is undetectable on T₂weighted MRI but present on histology.

Amir Moussavi¹, Kristin Koetz¹, Sanjay Tiwari¹, and Susann Boretius^1
1Section Biomedical Imaging, Department of Radiology and Neuroradiology, Christian-Albrechts-University, Kiel, Germany

Pancreatic ductal adenocarcinoma (PDAC) is still among the most devastating types of cancer characterized by infiltrative growth often with extensive fibrosis, the latter being an important predictor of poor chemotherapy response. In an orthotopic mouse model of PDAC with variable degree of fibrosis, we established and evaluated in-vivo radial imaging with magnetization transfer weighting. MT saturation correcting the MT-weighted signal for signal amplitude and T1 relaxation allowed quantification of tumor associated desmoplasia. Transfer of these methods to humans may allow individualized adjustments of chemotherapy.

Zihan Zhu^1,2, Peter J. Shin^1,2, Christine Leon Swisher³, Peder E.Z. Larson^1,2, Hsin-Yu Chen^1,2, Hong Shang^1,2, Eugene Milshteyn^1,2, Robert A. Bok¹, Andrei Goga⁴, and Daniel B. Vigneron^1,2
1Department of Radiology and Biomedical Imaging, University of California, San Francsico, San Francisco, CA, United States, ²UC Berkeley-UCSF Graduate Program in Bioengineering, San Francisco, CA, United States, ³Massachusetts General Hospital and Harvard Medical School, MA, United States, ⁴Department of Cell and Tissue Biology, University of California, San Francsico, San Francisco, CA, United States

Molecular imaging with hyperpolarized C-13 substrates provides valuable metabolic exchange information non-invasively. In this work, the dynamic Metabolic Activity Decomposition Stimulated-echo Acquisition Mode (MAD-STEAM) sequence was applied to a switchable oncogene-driven breast cancer mouse model. The results showed the feasibility of using MAD-STEAM to detect enzymatic activity changes within tumor and demonstrated its potential to monitor tumor progression and regression.

Natalie Julie Serkova¹, Diana M Cittelly¹, Kendra M Huber¹, and Carol A Sartorius^1
1University of Colorado Anschutz Medical Center, Aurora, Colorado, United States

Non-invasive multiparametric mpMRI of multifocal metastatic breast cancer provides characteristic MRI signatures for brain versus bone lesions in a mouse model of disseminated breast cancer. Brain metastatic phenotype is characterized by low vascularity and cell density as revealed by low Ktrans and ADC values, respectively. In future, quantitative mpMRI end-points have potential to stratify patients into “smart” targeted treatment regimens based on a particular tumor radiologiocal profile.

Sandra Ortega-Martorell^1,2, Ivan Olier³, Teresa Delgado-Goñi⁴, Magdalena Ciezka^2,5, Ana Paula Candiota^2,5, Margarida Julià-Sapé^2,5, Martí Pumarola^2,5, Paulo Lisboa¹, and Carles Arús^2,5
1Liverpool John Moores University, Liverpool, Merseyside, United Kingdom, ²Networking Research Center on Bioengineering, Biomaterials and Nanomedicine, CIBER-BBN, Cerdanyola del Vallès, Spain, ³The University of Manchester, Manchester, United Kingdom, ⁴The Institute of Cancer Research, London, United Kingdom, ⁵Universitat Autònoma de Barcelona, Cerdanyola del Vallès, Spain

A source-based extraction method is shown to produce nosological images allowing convenient non-invasive tracking of preclinical GBM response to therapy. The source-based classifier was developed from a training set of 14 C57Bl/6 mice harbouring GL261 GBM and validated in a further independent set of 24 additional mice using comparison with post-sacrifice histopathology.

Melanie Freed^1,2, Kerryanne Winters^1,2, Jin Zhang^1,2, and Sungheon G. Kim^1,2
1Center for Advanced Imaging Innovation and Research (CAI2R), Dept. Radiology, NYU School of Medicine, New York, NY, United States, ²Bernard and Irene Schwartz Center for Biomedical Imaging, Dept. Radiology, NYU School of Medicine, New York, NY, United States

Metronomic chemotherapy is a promising method that delivers chemotherapy in low, frequent doses and has been shown to demonstrate superior response over traditional chemotherapy in some cases. There is a need to develop mouse models and techniques for evaluating dose and scheduling of metronomic chemotherapy to optimize its usage and benefits. In this study, we investigate the use of DCE-MRI and DWI to evaluate metronomic chemotherapy of BALB/c mice with implanted 4T1 mouse mammary tumor cells. Our data suggests this may be a good model for advanced breast cancer and that ADC values may be sensitive to early therapy response.

Zhuxian Zhou¹, Mohammed Qutaish¹, Zheng Han¹, Rebecca Schur¹, David Wilson¹, and Zheng-Rong Lu^1
1Department of Biomedical Engineering, Case Western Reserve University, Cleveland, Ohio, United States

We evaluated a targeted contrast agent CREKA-Tris(Gd-DOTA)3 for MR molecular imaging of breast cancer micrometastases.

Cancer: Clinical & Preclinical Studies on New Contrast Mechanisms

Tuesday 2 June 2015

Scott C. Beeman¹, Ying-Bo Shui², John A. Engelbach¹, Joseph J.H. Ackerman^1,3, and Joel R. Garbow^1
1Radiology, Washington University, Saint Louis, Missouri, United States, ²Ophthalmology, Washington University, Saint Louis, Missouri, United States,³Chemistry, Washington University, Saint Louis, Missouri, United States

Diatomic oxygen (O₂), as found dissolved in tissue, is a critical component in aerobic metabolism and a fundamental determinant of physiological functional status. Reliable, non-invasive methods for measuring tissue O₂ content are lacking. We quantify the relaxivity of brain-tissue O₂ and, subsequently, changes in brain-tissue oxygenation in healthy brain, radiation lesions, and tumor lesions. By quantifying the change in brain tissue oxygenation with MRI, we were able to differentiate tumor lesions from radiation necrosis lesions – two pathologies that have proven extremely difficult to differentiate with common radiological techniques.

Alexander E. Salmon¹, Brian J. Pellatt², Nikolai J. Mickevicius³, Elizabeth J. Cochran⁴, and Peter S. LaViolette^5
1Neuroscience, Medical College of Wisconsin, Milwaukee, WI, United States, ²Medical College of Wisconsin, Milwaukee, WI, United States, ³Biophysics, Medical College of Wisconsin, Milwaukee, WI, United States, ⁴Pathology, Medical College of Wisconsin, Milwaukee, WI, United States, ⁵Radiology, Medical College of Wisconsin, Milwaukee, WI, United States

Precise co-registration of brain tissue and medical imaging is critical for validation of novel imaging biomarkers. Brain tissue distortion during fixation and traditional brain sectioning techniques may complicate co-registration. To prevent tissue distortion, a clinical MRI-based mesh mold was 3D-printed for use during fixation. To obtain representative axial cuts, an individualized tissue sectioning apparatus was designed from a high resolution clinical scan and the slice profile of 6.5mm images. After sectioning, consistency, and co-registration error was analyzed by AFNI. Using a layer resolution of 0.40mm, printing cost for each brain was approximately $61 and printing duration was approximately 38 hours.

Joris Tchouala Nofiele¹, Inga E Haedicke², Yong Le Zhu², Xiao-an Zhang², and Hai-Ling Margaret Cheng^1,3
1Hospital for Sick Children, Toronto, Ontario, Canada, ²Chemistry, University of Toronto, Toronto, Ontario, Canada, ³Institute of Biomaterials & Biomedical Engineering, University of Toronto, Toronto, Ontario, Canada

Contrast administration is integral to tumor diagnostic imaging, and gadolinium (Gd)-based contrast agents are the current approved standard for clinical workflow. However, a number of patients, particularly those with impaired kidney function, are exempt from the benefits of a contrast-enhanced examination. In this study, we investigate the potential of a recently described novel extracellular and renally filtered manganese porphyrin T1 agent for tumor imaging. Results in tumor-bearing rats demonstrate that the new agent, MnTCP, exhibits similar pharmacokinetics but provides consistently greater enhancement than Gd-DTPA. MnTCP can potentially provide contrast-enhanced MRI for tumor diagnosis in a broader patient base.

Breast Cancer: Technical

Tuesday 2 June 2015

Tess Catherwood¹, Andrew Patterson¹, Martin Graves¹, Reem Bedair¹, Roie Manavaki¹, Mary McLean², John Griffiths², and Fiona Gilbert^1
1Radiology, Cambridge University Hospitals NHS Foundation Trust, Cambridge, Cambridgeshire, United Kingdom, ²Cancer Research UK Cambridge Institute, Cambridge, Cambridgeshire, United Kingdom

BOLD contrast MRI with hyperoxic gas challenge may provide a biomarker of breast tumour oxygenation and vascular function, with the potential to predict treatment response. This study measured the strength of BOLD response to ‘carbogen-light’ (2% CO₂) interleaved with medical air and oxygen, compared to an all-air control, in six healthy volunteers. ‘Carbogen-light’ was effective in inducing a measurable BOLD effect with respect to background physiological noise, without the respiratory discomfort often reported with 5% CO₂ mixtures. In general ‘carbogen-light’ versus oxygen induced a larger response, consistent with the opposing vasomodulatory effects of these two stimuli.

Xiaofeng Liu¹, E Morris², Robert Darrow¹, and Ileana Hancu^1
1GE Global Research, Niskayuna, NY, United States, ²Memorial Sloan Kettering Cancer Center, NY, United States

Many factors contribute to MR-guided breast biopsy procedure success or failure. Significantly, there is no immediate verification of specimen retrieval and no quantitative measures to indicate adequate removal of the enhancing lesion. This work demonstrates a tool enabling quantitative assessment of success of MR-guided breast biopsy. Results from 4 patients indicated that the ratio between volumes of the suspicious lesion and total tissue removed in a biopsy procedure averaged 52%, ranging from 4% to 99%. The availability of real time quantitative assessment of lesion removal, prior to procedure completion, should enable improved verification of lesion sampling.

Saeedeh Navaei Lavasani^1,2, Masoomeh Gity³, Mahnaz Nabil^1,4, Anahita Fathi Kazerooni^1,2, and Hamidreza Saligheh Rad^1,2
1Quantitative MR Imaging and Spectroscopy Group, Research Center for Molecular and Cellular Imaging, Tehran University of Medical Sciences, Tehran, Iran,²Department of Medical Physics and Biomedical Engineering, School of Medicine, Tehran University of Medical Sciences, Tehran, Iran, ³Department of Radiology, School of Medicine, Tehran University of Medical Sciences, Tehran, Iran, ⁴Department of Statistics, Tarbiat Modares University, Tehran, Iran

Dynamic Contrast Enhanced-Magnetic Resonance Imaging (DCE-MRI) is widely used as sensitive tool in breast tumor diagnosis. Interpretation of breast MRI requires focusing not only on morphologic changes, but also on the quantification of the areas with increased enhancement. In this setting, accurate selection of quantitative parameters and classification approach could result in reliable tumor differentiation. We propose an accurate approach, based on support vector machine classification of dynamic features of suspicious tumors within the breast to discriminate malignant or benign breast lesions.

Sunitha B Thakur¹, Jung Hun Oh², Milans Soledad², Harini Veeraraghavan², Merlin M Gnanasigamani², Elizabeth J Sutton², Joseph O Deasy², and Elizabeth A Morris^2
1Memorial Sloan Kettering Cancer Center, New York, NY, United States, ²Memorial Sloan Kettering Cancer Center, NY, United States

Lesions that are classified ‘high-risk’ typically require complete surgical excision; a procedure that is expensive and can cause anxiety and morbidity. Therefore, the ability to distinguish between lesions, especially to differentiate the high-risk benign group from other benign lesions would be extremely helpful in a clinical setting. We conducted this study in order to evaluate the imaging characteristics of non-malignant lesions and classify them using (i) apparent diffusion coefficient (ADC) values, (ii) morphological and (iii) texture-based image features derived from contrast enhanced magnetic resonance imaging. We used 3.0T MRI data from 111 women and found that lower ADC values appear to correlate with high-risk breast lesions. This study is useful because it probes the concept of distinguishing not just between malignant and benign lesions but goes further to classify benign lesions into subgroups that could be treated differently at the clinical level.

Saeedeh Navaei Lavasani^1,2, Masoomeh Gity³, Anahita Fathi Kazerooni^1,2, and Hamidreza Saligheh Rad^1,2
1Quantitative MR Imaging and Spectroscopy Group, Research Center for Molecular and Cellular Imaging, Tehran University of Medical Sciences, Tehran, Iran,²Department of Medical Physics and Biomedical Engineering, School of Medicine, Tehran University of Medical Sciences, Tehran, Iran, ³Department of Radiology, School of Medicine, Tehran University of Medical Sciences, Tehran, Iran

Breast cancer is a significant public health problem in the world. Automatic and objective analysis of DCE-MRI studies can greatly assist the radiologist to gain accurate evaluation of tumor size, malignancy and perfusion in the surrounding tissues, which is essential in diagnosis. In this work, we proposed breast lesion segmentation by means of fuzzy c-means clustering technique using wavelet kinetic and semi-quantitative features, extracted from the pixel-based time-signal intensity curves.

Milica Medved¹, William A Weiss¹, Hiroyuki Abe¹, Gillian M Newstead¹, Olufunmilayo I Olopade², Maryellen L Giger¹, and Gregory S Karczmar^1
1Department of Radiology, University of Chicago, Chicago, Illinois, United States, ²Department of Medicine, University of Chicago, Chicago, Illinois, United States

High spectral and spatial resolution (HiSS) MR imaging provides excellent diagnostic images of breast cancer even without the use of contrast agent. We report on a fast implementation of HiSS MRI using parallel imaging at 3T to achieve clinically feasible imaging times in < 2 mm³ voxels, with full bilateral coverage. Non-contrast HiSS MRI can be used to obtain images of lesions without artifacts resulting from contrast administration. In addition, spectral information obtained with HiSS MRI can be used to improve detection of small amounts of glandular tissue and improve breast density measurements in risk assessment studies.

Lenka Minarikova¹, Wolfgang Bogner¹, Katja Pinker-Domenig², Thomas Helbich², Siegfried Trattnig¹, and Stephan Gruber^1
1MRCE, Department of Biomedical Imaging and Image-guided Therapy, Medical University of Vienna, Vienna, Austria, ²Division of Molecular and Gender Imaging, Department of Biomedical Imaging and Image-guided Therapy, Medical University of Vienna, Vienna, Austria

Response prediction during neoadjuvant chemotherapy (NAC) should have an important role in therapy plan modifications, which would result in increase of disease-free survival. In our study we measured breast cancer patients in three time points: before, during and after the NAC, using DCE and DWI MRI at 3T to look for potentially better response prediction markers. Our results suggest that multiparametric data measured during the ongoing NAC correspond better with histopathologic outcome than data measured after NAC completion. However, mean ADC values alone didn’t show any outcome prediction ability.

Tijl A. van der Velden¹, Alexander M. Th. Schmitz¹, Kenneth G.A. Gilhuijs¹, Wouter B. Veldhuis¹, Peter R. Luijten¹, Vincent O. Boer¹, and Dennis W.J. Klomp^1
1Radiology, University Medical Center Utrecht, Utrecht, Utrecht, Netherlands

Fat suppression is essential in high-resolution DCE-MRI of the breast. While chemical shift selective excitation (i.e. ProSet) is sensitive to magnetic field distortions, Dixon techniques may be sensitive to T₂^* effects due to their multi echo acquisition (ME-Dixon), or to motion when applied as multi acquisition (MA-Dixon). In this study, we investigated the efficiency of these fat suppression techniques at ultra-high resolutions obtained at 7 tesla. In ten healthy volunteers ProSet revealed structural details that were not observed with Dixon. Despite a non-uniform fat suppression of ProSet, the technique remains superior in CNR and time efficiency over other techniques.

Jing Yuan¹, Gladys G. Lo², Oi Lei Wong¹, Helen H.L. Chan², Abby Y. Ding¹, Ting Ting Wong³, and Polly S.Y. Cheung^3
1Medical Physics and Research Department, Hong Kong Sanatorium & Hospital, Happy Valley, Hong Kong, China, ²Department of Diagnostic & Interventional Radiology, Hong Kong Sanatorium & Hospital, Happy Valley, Hong Kong, China, ³Breast Care Center, Hong Kong Sanatorium & Hospital, Happy Valley, Hong Kong, China

This study prompts to rigorously assess and statistically compare DWI signal fitting in malignant breast tumor by using mono-exponential ADC model and bi-exponential intra-voxel incoherent motion (IVIM) model via F-test. ROI-averaged analysis showed that more tumors (n=24) were significantly better fitted by bi-exponential than by mono-exponential models (n=14). In contrast, voxel-wise analysis showed that more voxels presented mono-exponential DWI signal decay in tumors. With clinically achievable SNR, IVIM-MRI should be helpful for ROI-averaged DWI fitting to characterize tumor microvascularity while voxel-wise IVIM values should be carefully interpreted with caution due to possible inaccuracy and high uncertainty.

Cancer: Prostate Cancer

Tuesday 2 June 2015

Nainesh Parikh¹, Justin Ream¹, Tobias Block², Weisheng Xu³, Hersh Chandarana¹, Li Feng², Samir Taneja⁴, and Andrew Rosenkrantz^1
1Radiology, NYU School of Medicine, New York, NY, United States, ²Radiology, Center for Advanced Imaging Innovation and Research NYU School of Medicine, New York, NY, United States, ³Pathology, NYU School of Medicine, New York, NY, United States, ⁴Urologic Oncology, NYU School of Medicine, New York, NY, United States

We used a novel DCE-MRI sequence [Golden-angle RAdial Sparse Parallel (GRASP)] that employees compressed sensing and continuous radial acquisition to retrospectively reconstruct prostate DCE-MRI data obtained after a single contrast injection at temporal resolutions varying between approximately 1 and 10 seconds. The temporal resolution did not impact diagnostic performance of quantitative DCE metrics for tumor detection.

Nassim Tayari¹, Anca R. Croitor Sava², Diana M. Sima², Sabine Van Huffel², and Arend Heerschap^1
1Department of Radiology and Nuclear Medicine, Radboud University Medical Center, Nijmegen, Netherlands, ²Department of Electrical Engineering, Katholieke Universiteit Leuven, Leuven, Belgium

An automated quality control algorithm plays an important role in automation of the analysis of MRSI data of the prostate cancer patients. In this work we present an automated unsupervised quality control algorithm for 3D 1H MRSI data sets. The method is based on feature extraction using Non Negative Matrix Factorization(NNMF). Consensus decisions of spectral quality judged by four experts is used as a Gold Standard for performance evaluation showing that the algorithm can separate good quality from bad quality spectra with 90% sensitivity and 90% specificity.

Gregory J. Metzger¹, Chaitanya Kalavagunta¹, Stephen C Schmechel², Patrick J. Bolan¹, Badrinath Konety³, Benjamin Spilseth⁴, Christopher A. Warlick³, and Joseph S. Koopmeiners^5
1Center for Magnetic Resonance Research, University of Minnesota, Minneapolis, MN, United States, ²Department of Pathology, University of Washington, Washington, United States, ³Department of Urologic Surgery, University of Minnesota, Minneapolis, MN, United States, ⁴Department of Radiology, University of Minnesota, Minneapolis, MN, United States, ⁵Division of Biostatistics, University of Minnesota, Minneapolis, MN, United States

A process is presented for generating critical correlative pathology for developing predictive models from voxel-wise mpMRI data based on mapping regions of disease from assembles pathology to in vivo MRI. The models generated from this novel data show improved performance over single quantitative MRI parameters for detection. The generation of composite biomarker maps has the potential to improve the use of mpMRI in the management of prostate cancer by providing a quantitative means to assess and monitor disease.

Kai Zhao¹, Chengyan Wang², Juan Hu¹, Xiaodong Zhang¹, Jue Zhang², and Xiaoying Wang^1
1Department of Radiology, Peking University First Hospital, Beijing, Beijing, China, ²College of Engineering, Peking University, Beijing, Beijing, China

We extract 12 quantitative features from the T2 weighted images of prostate cancer and non-cancer by CAD, within the central gland and peripheral zone of the prostate, respectively.You can find out which feature can differ cancer from non-cancer better and the different feature weight of each feature in our research. This could be widely used and help us in the prostate cancer diagnosis.

Xin Li¹, Jackilen Shannon¹, Mark G Garzotto^1,2, Chris Amling¹, William J Woodward¹, George Thomas¹, Elizabeth Dacey^1,2, Xiaohua Wang^1,2, Paige Farris¹, Wesley Stoller², Ann Martinez Acevedo¹, Amy Palma¹, Manoj K Sammi¹, William D Rooney¹, Fergus V Coakley¹, and Jonathan Q Purnell^1
1Oregon Health & Science University, Portland, Oregon, United States, ²Portland VA Medical Center, Portland, Oregon, United States

The lipogenic fatty acid synthase (FAS) is responsible for de novo fatty acid and triglyceride synthesis in tissue. Recently, FAS has been characterized as an oncogene and its overexpression is reported in several cancers which include that of the prostate. The altered lipid deposition of cancerous tissue thus can potentially be an independent biomarker for cancer aggressiveness monitoring. In this work, MR spectroscopic imaging (MRSI) is used to quantify prostatic lipid concentration for prostate cancer.

Naira Muradyan¹, Osama Elbuluk², Baris Turkbey², Sandeep Sankineni², Maria J Merino³, Senthil Periaswamy¹, Marcelino Bernardo², Francois Cornud⁴, and Peter L Choyke^2
1iCAD, Inc., Nashua, NH, United States, ²Molecular Imaging Program, NCI, NIH, Bethesda, MD, United States, ³Laboratory of Pathology, NCI, NIH, Bethesda, MD, United States, ⁴Tourville Imaging Centre, Paris, France

This study describes a combined analysis method of Diffusion Weighted and Dynamic Contrast Enhanced MRI for prostate imaging, which minimizes the need for separate evaluation of these sequences. Images of 105 patients from two sites were used retrospectively to develop and test the prostate zone-specific composite colormaps, showing better accuracy than standalone Apparent Diffusion Coefficient outcome. Such objective display of combined information, which could be easily added to routine imaging, may be useful for lesion visualization, evaluation and minimizing reader variability during multiparametric prostate imaging.

Francesca Mertan^1,2, Harsh K Agarwal^2,3, Sandeep Sankineni², Marcelino Bernardo^2,4, Dagane Daar^2,4, Maria Merino², Bradford Wood², Peter Pinto², Peter L Choyke², and Baris Turkbey^2
1Grove City College, Grove City, PA, United States, ²National Institutes of Health, Bethesda, MD, United States, ³Philips Research NA, Briarcliff Manor, New York, United States, ⁴Leidos Biomedical Research Inc., Frederic National Laboratory for Cancer Research, Frederick, MD, United States

ADC maps have been shown to detect and stage prostate cancer. However, at high b-value (b>800mm/s2) DWI signal from tumor tissue does not follow the mono-exponential diffusion decay. Therefore high b-value DWI is typically accessed as part of prostate MRI reading. In this abstract we have shown that the performance of high b-value DWI to detect high risk prostate cancer patients improves with the increase in b-value and its performance did not change significantly with the increase in data acquisition time from 23msec to 102msec for b≥0s/mm2 DWI. High b-value DWI also performed better than ADC from b≤750s/mm2.

Renuka Sriram¹, Mark Van Criekinge¹, Justin DeLos Santos¹, Daniel B Vigneron¹, Robert Bok¹, Donna Peehl², Kayvan Rahimi Keshari³, and John Kurhanewicz^1
1University of California, San Francisco, San Francisco, CA, United States, ²Stanford University, CA, United States, ³Memorial Sloan Kettering Cancer Center, New York, United States

The goal of the current study was to correlate hyperpolarized lactate with the pathologic grade of prostate cancer using living human prostate tissue slices (TSCs) obtained at surgery in a MR compatible 3D tissue culture bioreactor. TSCs obtained from 38 radical prostatectomy patients were studied after injection of hyperpolarized pyruvate and demonstrated that in addition to increased lactate production, high grade prostate cancer had significantly higher lactate efflux. Knowledge of significant extracellular HP lactate in aggressive prostate cancer will be useful for improving interpretation of the in vivo HP signals and interrogated using diffusion weighted HP ¹³C MR approaches.

Shivani Pahwa¹, Robert Abouassaly², Yun Jiang³, Karin Herrmann^4,5, Raj Paspulati^5,6, William Tabayoyong⁷, Soham Shah⁷, Brian Minnillo⁷, Gregory MacLennan⁷, Mark Griswold^1,8, Lee Ponsky^5,9, and Vikas Gulani^5,10
1Radiology, Case Western Reserve University, Cleveland, Ohio, United States, ²University Hospitals, Ohio, United States, ³Biomedical Engineering, Case Western Reserve University, Cleveland, Ohio, United States, ⁴Radiology, University Hospitals, Ohio, United States, ⁵CWRU School of Medicine, Ohio, United States, ⁶UH Case Medical Center, Ohio, United States, ⁷Urology, University Hospitals, Ohio, United States, ⁸Biomedical Engineering, Case Western Reserve University, Ohio, United States, ⁹Urology, UH Case Medical Center, Ohio, United States, ¹⁰Radiology, UH Case Medical Center, Ohio, United States

Serum PSA measurement and digital rectal examination (DRE) are the current standards of care for detection of prostate cancer. However, the performance of these methods as screening tools has been criticized because they over diagnose a large number of patients with clinically insignificant disease, adding to a significant morbidity for the patient without improving the outcome. We propose a novel, short non-contrast imaging protocol as a secondary screening tool that helps direct targeted biopsies on suspicious lesions. Our preliminary experience indicated that this strategy increases the accuracy of detecting clinically significant disease, and improves risk stratification for biopsy.

Daniel Hausmann¹, Nils Rathmann¹, Metin Sertdemir¹, Philipp Riffel¹, Anja Weidner¹, Stephan Kannengiesser², John N. Morelli³, Stefan O. Schoenberg¹, and Ulrike I. Attenberger^1
1Institute of Clinical Radiology and Nuclear Medicine, University Medical Center Mannheim, Medical Faculty Mannheim, University of Heidelberg, Germany, Mannheim, Baden-Württemberg, Germany, ²MR Applications Development, Siemens Healthcare, Erlangen, Germany, ³Department of Radiology, St. John’s Medical Center, Tulsa, Oklahoma, United States

The use of spatially-tailored, two-dimensional radiofrequency (RF) excitation pulses in single-shot echo-planar imaging (EPI), combined with a decreased FOV in the phase-encoding direction, results in a reduced amount of k-space acquisition lines, which shortens the EPI echo train length (ETL) and reduces susceptibility artifacts without an increase in acquisition time. Zoomed EPI leads to significant improvements in image quality and artifacts as well as image blur reduction improving prostate DWI and enabling accurate fusion with conventional sequences.

Tristan Barrett¹, Andrew N Priest², Edward M Lawrence¹, Debra Goldman³, Vincent J. Gnanapragasam⁴, Evis Sala⁵, and Ferdia A. Gallagher^1
1Radiology, Cambridge University Hospitals, Cambridge, Cambridgeshire, United Kingdom, ²Cambridge University Hospitals, Cambridge, Cambridgeshire, United Kingdom, ³Department of Epidemiology and Biostatistics, Memorial Sloan Kettering Cancer Center, New York, New York, United States, ⁴Urology, Cambridge University Hospitals, Cambridge, Cambridgeshire, United Kingdom, ⁵Radiology, Memorial Sloan Kettering Cancer Center, New York, New York, United States

We investigate the ability of apparent diffusion coefficient (ADC) ratios of tumour-to-normal prostate to improve concordance between ADC values obtained from different b-value combinations within the same patients. Whole gland histopathology was used as a gold-standard. Concordance correlation between different b-value combinations was higher for ADC ratios compared to absolute ADC values. ADC maps incorporating a b-0 value demonstrated greater variability for both ratio and absolute values, likely relating to pseudo-perfusion effects. ADC ratios showed stronger inverse relationships to Gleason grade than absolute values. ADC ratios may therefore provide a more robust means of assessing restricted diffusion in the prostate.

Chunmei Li¹, Bing Wu², and Min Chen^1
1Beijing Hospital, Beijing, Beijing, China, ²GE healthcare China, Beijing, Beijing, China

We investigate the use of a novel reduced FOV CEST method that allows us to zoom in the prostate region that helps delineating the tumor region from the benign region and improves accuracy in grading using CEST. The use of reduced FOV CEST in prostate was validated by comparing the APT map and MTRasym to those obtained using normal CEST, highly consistent results were obtained despite its lower intrinsic SNR. Reduced FOV CEST has the advantage of revealing the heterogeneity of the prostate, which may be difficult due to the constrained portion of the prostate region in normal FOV image.

Rebecca Rakow-Penner¹, Nathan White¹, Daniel Margolis², J. Kellogg Parsons³, Natalie Schenker-Ahmed¹, Joshua Kuperman¹, Hauke Bartsch¹, Hyung Choi², William Bradley¹, Ahmed Shabaik⁴, Jiaoti Huang⁵, Michael Liss⁶, Leonard Marks⁷, Christopher Kane³, Robert Reiter⁷, Steven Raman², David Karow¹, and Anders Dale^1
1Radiology, University of California San Diego, San Diego, CA, United States, ²Radiology, University of California Los Angeles, Los Angeles, California, United States, ³Urology, University of California San Diego, San Diego, CA, United States, ⁴Pathology, University of California San Diego, San Diego, CA, United States, ⁵Pathology, University of California Los Angeles, Los Angeles, CA, United States, ⁶Urology, University of Texas Health Science Center San Antonio, San Antonio, TX, United States, ⁷Urology, University of California Los Angeles, Los Angeles, CA, United States

Distortion maps were generated to reflect the offset of the collected data versus the corrected data.

Marnix C. Maas¹, Geert J.S. Litjens^1,2, Alan J. Wright³, Masoom A. Haider⁴, Katarzyna J. Macura⁵, Kirsten M. Selnæs⁶, Daniel J.A. Margolis⁷, Thomas Helbich⁸, Berthold Kiefer⁹, Jurgen J. Fütterer¹, and Tom W.J. Scheenen^1
1Radiology and Nuclear Medicine, Radboud University Medical Center, Nijmegen, GLD, Netherlands, ²Pathology, Radboud University Medical Center, Nijmegen, GLD, Netherlands, ³Cancer Research UK Cambridge Institute, University of Cambridge, Cambridge, United Kingdom, ⁴Sunnybrook Health Sciences Center, University of Toronto, Toronto, ON, Canada, ⁵Russel H. Morgan Department of Radiology and Radiological Science, Johns Hopkins University, Baltimore, MD, United States, ⁶Department of Circulation and Medical Imaging, Norwegian University of Science and Technology, Trondheim, Norway, ⁷Radiology, UCLA David Geffen School of Medicine, Los Angeles, CA, United States, ⁸Biomedical Imaging and Image-guided Therapy, Medical University Vienna - General Hospital Vienna, Vienna, Austria, ⁹Siemens AG Healthcare, Erlangen, Germany

This study investigates the ability of multiparametric MRI (mpMRI) including DWI, 1H-MRSI and DCE-MRI to discriminate low-grade from higher-grade peripheral zone (PZ) prostate cancer in a multi-center setting, using whole-mount section histopathology as the gold standard. Guided by histology and blinded to functional imaging, ROIs were defined on T2w imaging in PZ tumors in 39 patients from 5 centers, and transferred to functional parameter maps. ROC analysis resulted in areas under the curve of 0.80±0.13 for ADC and 0.83±0.15 for a multivariate model including both ADC and MRSI. Multicenter mpMRI can yield good separation between low and higher grade PZ tumors.

Chaitra Badve¹, Alice Yu², Shivani Pahwa³, Matthew Rogers², Yun Jiang⁴, Yiying Liu⁵, Mark Schluchter⁵, Lee Ponsky^6,7, Mark Griswold⁴, and Vikas Gulani^1,3
1Radiology, University Hospitals, Cleveland, Ohio, United States, ²School of Medicine, Case Western Reserve University, Cleveland, Ohio, United States,³Radiology, Case Western Reserve University, Cleveland, Ohio, United States, ⁴Biomedical Engineering, Case Western Reserve University, Cleveland, Ohio, United States, ⁵Biostatistics, Case Western Reserve University, Cleveland, Ohio, United States, ⁶Urology, University Hospitals, Cleveland, Ohio, United States,⁷Urology, Case Western Reserve University, Cleveland, Ohio, United States

Magnetic resonance fingerprinting (MRF) was acquired in patients with clinical suspicion of prostate cancer. After diagnoses were confirmed by pathology, T1 and T2 values from MRF were used with ADC values to retrospectively separate malignancy from normal peripheral zone (NPZ). All MR parameters showed significant difference between prostate cancer and NPZ. Furthermore, T2 and ADC used in conjunction had a high discriminatory power (AUC = 0.996). Further research is needed to determine the robustness of using T2 and ADC to identify prostate cancer prospectively, and to explore the possibly utility of T1 in identifying malignancy.

Meiyu Sun¹, Ailian Liu¹, Ye Li¹, Lihua Chen¹, Qingwei Song¹, Bin Xu¹, and Ziheng Zhang^2
1the first affiliated hospital of Dalian Medical University, Dalian, Liaoning, China, ²GE healthcare China, Beijing, China

The MR imaging of prostate cancer (PCa) was undistinguishable from that of benign prostatic hyperplasia (BPH) on conventional T2WI and dynamic contrast enhancement imaging. In this study, intravoxel incoherent motion (IVIM) DWI were performed to investigate the diffusion and perfusion features of PCa and BPH, and to evaluate the relationship of the parameters, standard ADC, slow ADC, fast ADC and fraction of fast ADC, as well as their sensitivity and specificity in diagnosis. It was demonstrated standard ADC and slow ADC from IVIM MRI maybe a significant MRI biomarkers to differentiate PCa from BPH with satisfying sensitivity and specificity.

Rui Wang¹, Juan Hu¹, Yuanyuan Jiang¹, and Xiaoying Wang^1
1Radiology, Peking university first hospital, Beijing, Beijing, China

In patients with PSA levels of 4-10 ng/mL, it is difficult to distinguish prostate cancer (PCa) from prostatitis. So our study is to evaluate the role of multiparametric MRI (mp-MRI) in detection of PCa and prediction of PCa incidence in patients with total PSA levels of 4¨C10 ng/mL by a prospective cohort study. Mp-MRI showed excellent area under the ROC curve, suggesting obviously clinically relevant predictive characteristics. Furthermore, with 137 months of follow-up, higher score of mp-MRI was proved to be significantly associated with a greater risk of detection rate of PCa.

Harsh K Agarwal^1,2, Jochen Keupp³, Marcelino Bernardo², Baris Turkbey², and Peter L Choyke^2
1Philips Research NA, Briarcliff Manor, New York, United States, ²National Institutes of Health, Bethesda, MD, United States, ³Philips Research Laboratories, Hamburg, Germany

Amide proton transfer (APT) MRI has been shown to localize and stage prostate. However there are multiple parameters such as saturation power and saturation duration that affect the APT MRI contrast. In this abstract we have shown that MTR asymmetry increased with the decrease in the pulse duration. Also the peak of MTR asymmetry shifted with the change in saturation power. However spurious MTR was observed over prostate especially in peripheral zone which could be due to the incoherent motion such as rectal spasm.

Pooria Khoshnoodi¹, Nelly Tan¹, Daniel J. A. Margolis¹, Wei-Chan Lin¹, Somrach Thamtorawat¹, David Y. Lu², Jiaoti Huang², Robert E. Reiter³, and Steven S. Raman^1
1Radiology, University of California, Los Angeles, Los Angeles, California, United States, ²Pathology, University of California, Los Angeles, Los Angeles, California, United States, ³Urology, University of California, Los Angeles, Los Angeles, California, United States

Multiparametric magnetic resonance imaging (MP-MRI) is used more frequently to detect and evaluate prostate cancer (CaP). We studied the size of the tumor prior to surgery on MP-MRI and its correlation with the actual size of tumor on pathology after prostatectomy. This correlation was stratified by overall PI-RADS suspecious score on multiparametric MRI and also Gleason Score at histopathology.

Taylor Stone¹ and Luis Beltran^2
1New York University, New York, NY, United States, ²New York University, NY, United States

NaF PET-MRI as a hybrid imaging study shows higher specificity, sensitivity, and overall accuracy than bone scintigraphy, PET or MRI in isolation in differentiating between metastatic and benign bone lesions. Accurately defining the number of lesions has important treatment implications as prostate cancer treatment is generally based on disease burden.

Stephanie Nougaret¹, Nicola L Robertson¹, Evis Sala¹, Hedvik Hricak¹, Behfar Ehdaie², and Hebert A Vargas^1
1Radiology department, Memorial Sloan Kettering Cancer Center, New York, NY, United States, ²Urology department, Memorial Sloan Kettering Cancer Center, New York, NY, United States

To evaluate the combined efficacy of MR Volumetry and Diffusion Weighted Imaging (DWI), in the detection of clinically significant prostate cancer using MRI-US fusion biopsy. This preliminary study demonstrates the potential added predictive value of using tumor volumetry in combination with DWI, in the diagnosis of clinically significant prostate cancer using MRI-US fusion biopsy. Furthermore, it evaluates the feasibility of quantitative assessment in tumor burden.

Christopher Nguyen^1,2, Ali-Reza Sharif-Afshar³, Zhaoyang Fan¹, Sidney Wilson², Xiaoming Bi⁴, Lucas Payor⁵, Rola Saouaf⁵, Hyung Kim³, and Debiao Li^1,2
1Biomedical Imaging Research Institute, Cedars Sinai Medical Center, Los Angeles, CA, United States, ²Bioengineering, University of California Los Angeles, Los Angeles, CA, United States, ³Surgery / Urology, Cedars-Sinai Medical Center, Los Angeles, CA, United States, ⁴Siemens Healthcare, Los Angeles, CA, United States, ⁵Radiology, Cedars-Sinai Medical Center, Los Angeles, CA, United States

In a small group of prostate cancer patients undergoing active surveillance (AS-PC) patients, a recently developed novel 3D diffusion-prepared multi-shot balanced steady-state free precession technique demonstrated improved suspicious low risk prostate cancer lesion detection compared with single-shot diffusion-weighted echo planar imaging (SS-DW-EPI). We conducted a follow-up study in a significantly larger cohort to further investigate the clinical impact of the proposed technique. The proposed technique continued to yield significantly better lesion detection of biopsy-confirmed lesions compared with SS-DW-EPI. By improving lesion detection, the proposed technique may allow DW MRI to potentially monitor low-risk prostate cancer lesions in AS-PC.

Tumor Therapy Responses: Preclinical & Clinical (except Brain Tumor)

Tuesday 2 June 2015

Igor Jacobs¹, Stefanie Hectors¹, Gustav Strijkers^1,2, Klaas Nicolay¹, and Matthias Schabel^3,4
1Biomedical NMR, Department of Biomedical Engineering, Eindhoven University of Technology, Eindhoven, Netherlands, ²Biomedical Engineering and Physics, Academic Medical Center, University of Amsterdam, Amsterdam, Netherlands, ³Advanced Imaging Research Center, Oregon Health and Science University, Portland, Oregon, United States, ⁴Utah Center for Advanced Imaging Research, University of Utah, Salt Lake City, Utah, United States

Accurate evaluation of antivascular tumor therapies requires detailed analysis of vascular alterations. We have recently proposed a novel multi-agent tracer-kinetic modeling approach for DCE-MRI, which allows simultaneous and self-consistent assessment of blood flow and permeability within one imaging session. In the present study, we have applied this approach to evaluate DMXAA treatment. An early increase in extraction fraction of the dendrimer-based contrast agents and reduced blood flow were simultaneously observed, thereby identifying increased permeability as the main mechanism of DMXAA-induced vascular shutdown, in agreement with literature. This demonstrates the potential of the multi-agent approach to assess therapy-induced vascular changes.

Tom Schreurs^1,2, Stefanie Hectors¹, Igor Jacobs¹, Holger Grüll^1,3, Gustav Strijkers^1,2, and Klaas Nicolay^1
1Biomedical Engineering, Eindhoven University of Technology, Eindhoven, Netherlands, ²Biomedical Engineering and Physics, Academic Medical Center, Amsterdam, Netherlands, ³Oncology Solutions, Philips Research, Eindhoven, Netherlands

Micro-vascular effects of photodynamic therapy (PDT) were assessed using Dynamic Contrast Enhanced (DCE) MRI in a mouse tumor model. PDT is an emerging cancer therapy based on a light-activated drug. DCE-MRI showed that tumors were almost completely perfused before PDT, but partly non-perfused after 1 hour and largely non-perfused after 24 hours. The results clearly indicated PDT-induced vascular occlusion. Remaining enhanced regions coincided with tumor recurrences. This work shows that DCE-MRI has large potential to monitor PDT efficacy, early after treatment.

Stefanie Hectors¹, Igor Jacobs¹, Edwin Heijman², Jochen Keupp³, Monique Berben², Gustav Strijkers^1,4, Holger Grüll^1,2, and Klaas Nicolay^1
1Biomedical NMR, Department of Biomedical Engineering, Eindhoven University of Technology, Eindhoven, Netherlands, ²Oncology Solutions, Philips Research Europe, Eindhoven, Netherlands, ³Tomographic Imaging Systems, Philips Research Europe, Hamburg, Germany, ⁴Biomedical Engineering and Physics, Academic Medical Center, University of Amsterdam, Amsterdam, Netherlands

A multiparametric MRI protocol, consisting of quantitative T₁, T₂, ADC, amide proton transfer (APT) and T_1ρ measurements, was used to evaluate MR-guided HIFU treatment of subcutaneous rat tumors. K-means clustering was performed on the multiparametric data using all different combinations (feature vectors) of MRI parameters. For feature vector {ADC, APTw signal} a strong correlation (r=0.92) between histology-derived and clustering-derived non-viable tumor fractions was observed, which was higher than correlations between histology-derived non-viable fractions and HIFU-treated fractions derived from MR thermometry and dynamic contrast-enhanced MRI. In conclusion, our data demonstrate that a multiparametric MR analysis can accurately identify HIFU-treated tumor tissue.

Kay Pepin¹, Steven Ansell², Richard L Ehman³, and Kiaran McGee^3
1Graduate School, Mayo Clinic, Rochester, Minnesota, United States, ²Hematology, Mayo Clinic, Minnesota, United States, ³Radiology, Mayo Clinic, Minnesota, United States

Change in tumor mechanical properties measured using MR Elastography may be a biomarker of response to chemotherapy. The ability to determine failure to therapy is a critical function of an imaging biomarker. The results of this animal study demonstrate that MRE is sensitive to an increase in tumor stiffness prior to an increase in tumor volume. Similarly, change in tumor shear stiffness is dose-dependent; the higher the chemotherapy dose the greater the subsequent decrease in tumor stiffness. These results suggest that MRE-derived shear stiffness measurements reflect the observed response to chemotherapy.

Charles S. Springer^1,2, Xin Li³, Mohan L. Jayatilake⁴, Martin M. Pike^2,3, William D. Rooney³, Rosalie C. Sears^2,5, and Wei Huang^2,3
1Advanced Imaging Research Center, Oregon Health & Science University, Portland, Or, United States, ²Knight Cancer Institute, Oregon Health & Science University, Portland, Oregon, United States, ³Advanced Imaging Research Center, Oregon Health & Science University, Portland, Oregon, United States,⁴Radiography and Radiotherapy, University of Peradeniya, Peradeniya, Peradeniya, Sri Lanka, ⁵Molecular and Medical Genetics, Oregon Health & Science University, Portland, Oregon, United States

Normally, DCE-MRI estimates vascular parameters, K^trans and v_p. However, shutter-speed analysis allows determination of cellular, metabolic parameters, k_io and v_i. The k_io quantity tracks cell membrane Na⁺,K⁺-ATPase activity. Exemplified with human and murine breast cancers, we show that these enable the discrimination of cytotoxic and targeted therapies in their early actions.

Naomi S. Sta Maria¹, Samuel R. Barnes¹, David Colcher², Andrew A. Raubitschek², and Russell E. Jacobs^1
1Biology and Biological Engineering, California Institute of Technology, Pasadena, CA, United States, ²Cancer Immunotherapeutics & Tumor Immunology, City of Hope, Duarte, CA, United States

We investigated whether the priming of tumors with immunocytokines can potentiate the accumulation of adoptive human natural killer cells labeled with ferumoxytol, an FDA-approved USPIO, in tumors. We used T2* weighted multi-gradient echo to generate parametric R2* maps of the tumors in NOD-scid-gamma-null mice with human colorectal adenocarcinoma cells, LS-174T.luc. Mice were differentiated into control or immunocytokine treated groups and were imaged pre and post (5hr and 10hr) injection of NK cells. Pretreated mice showed significant increases in tumor R2* values at the 10hr time point. MRI can be used to assess efficacy of adoptive NK cell transfer.

Arash Latifoltojar¹, Margaret Hall-Craggs², Alan Bainbridge², Stuart Taylor³, Kwee Yong³, Neil Rabin², Matthew Benger², Liam Watson², Michelle Siu², and Shonit Punwani^3
1University College London, London, London, United Kingdom, ²University College London Hospital, London, United Kingdom, ³University College London, London, United Kingdom

Whole body anatomical and functional MRI is being increasingly used for initial assessment and treatment response monitoring in patients with multiple myeloma. The replacement of the normal fatty marrow cells by neoplastic cells in multiple myeloma and its changes following therapy could be explored, using quantitative derive parameters from MR imaging. In this study we investigated the changes in quantitative derived parameters from whole body mDixon changes following therapy in patients with multiple myeloma.

Junyu Guo¹ and Wilburn E. Reddick^1
1Radiological Sciences, St Jude Children's Research Hospital, Memphis, TN, United States

This study investigated two DCE-MRI data analysis methods for evaluating response to treatment and prognosis of event-free-survival (EFS) and overall-survival in 42 patients with osteosarcoma. In two methods, eight parameters and their absolute-difference (ABD) between two neighboring time points were assessed for the role of predicting clinical results. Three ABD parameters before week5 could potentially predict response. The CPA parameter  on Day5 is the potential prognostic factor for EFS. Four ABD parameters before week5 are potential prognostic factors for overall-survival. All results show that two DCE-MRI analysis methods could provide valuable predictive information for clinical outcomes.

Yi Sui^1,2, Lei Tang³, Kejia Cai^2,4, Shun-Yu Gao³, Frederick C. Damen^2,4, Ying-Shi Sun³, and Xiaohong Joe Zhou^2,5
1Bioengineering, University of Illinois at Chicago, Chicago, IL, United States, ²Center for MR Research, University of Illinois Hospital & Health Sciences System, Chicago, IL, United States, ³Radiology, Peking University Cancer Hospital & Institute, Beijing, China, ⁴Radiology, University of Illinois Hospital & Health Sciences System, Chicago, IL, United States, ⁵Departments of Radiology, Neurosurgery and Bioengineering, University of Illinois Hospital & Health Sciences System, Chicago, IL, United States

Sunitinib offers a second line treatment to gastrointestinal stromal tumor (GIST) patients with imatinib-resistant progressive lesions. Prediction of the treatment response is crucial for achieving optimal outcomes. We used a non-Gaussian diffusion model, known as the fractional order calculus model (FROC), to assess the early response to sunitinib treatment in imatinib-resistant GIST lesions. Our results showed that the FROC model outperformed the conventional means of using tumor size in predicting tumor response at as early as 2 weeks after receiving sunitinib treatment. This study suggests that non-Gaussian diffusion imaging can provide timely information to evaluate sunitinib targeted therapy of GIST.

Wei Huang¹, Megan L Holtorf¹, Aneela Afzal¹, Yiyi Chen¹, Brooke R Beckett¹, and Christopher W Ryan^1
1Oregon Health & Science University, Portland, OR, United States

Twenty patients with extremity soft-tissue sarcoma who underwent preoperative chemoradiotherapy consented to research DCE-MRI, performed before, after one cycle of chemotherapy, and after completion of therapy before surgery. Ktrans and kep at baseline, after one cycle of therapy, and percent change were much better than RECIST tumor size measurement for early prediction of therapy response. Post-therapy Ktrans, ve, and kep were negatively correlated with necrosis percentage (NP) of the surgical specimen. No significant relationship was observed between post-therapy RECIST and NP.

Yonggang Lu¹, Nancy Lee¹, Vaois Hatzoglou¹, Nadeem Riaz¹, Joseph O. Deasy¹, and Amita Shukla-Dave^1
1Memorial Sloan-Kettering Cancer Center, NEW YORK, New York, United States

Our aim is to identify imaging biomarkers predictive of treatment response in nasopharyngeal cancer for personalizing radiotherapy. A total of 37 weekly intravoxel incoherent motion DW-MRI studies were performed on 5 patients to assess early chemoradiation treatment response. The results demonstrated that chemoradiation treatment led to a decrease of tumor volumes in all 5 patients and an increase of diffusion and perfusion metrics in 4 patients. One patient had tumor infiltration of the muscle and the diffusion metric remained unchanged in this patient during treatment. The study concludes that diffusion and perfusion metrics are potential imaging biomarkers of tumor response.

Alexander R Guimaraes^1,2, Wendy Atkinson³, Ephraim Hochberg⁴, Jeremy Abramson⁵, Onofrio Catalano², Bruce R Rosen², and Ciprian Catana^2
1Radiology, Oregon Health Sciences University, Portland, OR, United States, ²Radiology, Martinos Center for Biomedical imaging, Charlestown, MA, United States,³Radiology, Martinos Center for Biomedical imaging, Charlesown, MA, United States, ⁴Medicine, Massachusetts General Hospital, Boston, MA, United States,⁵Medicine, Massachusetts General Hospital, MA, United States

FDG-PET/CT has become an essential tool in the initial staging of Hodgkin lymphoma and aggressive non-Hodgkin lymphomas. In select histologies, early interim PET/CT has also shown to be a strong independent predictor of outcome. Hybrid PET/MR imaging is a novel imaging modality that offers improved soft tissue contrast, conspicuity of bone marrow infiltration, diffusion weighted imaging, and lower radiation dose: all factors that could have dramatic impact in the management of lymphoma. The overarching goal of this work was to test the hypothesis that PET/MRI offers an equivalent qualitative and quantitative imaging staging algorithm to PET/CT, but with decreased-risk of late radiation-associated toxicities.

Rafal Panek¹, Lauren C.J. Baker², Liam Welsh¹, Carol Box², Suzanne A. Eccles², Kate L. Newbold¹, Kevin J. Harrington¹, Maria A. Schmidt¹, Martin O. Leach¹, and Simon P. Robinson^2
1Royal Marsden NHS FT and Institute of Cancer Research, Sutton, Surrey, United Kingdom, ²Institute of Cancer Research, Sutton, Surrey, United Kingdom

The importance of acutely hypoxic tumour cells, resulting from intermittent tumour blood flow, on tumour progression and radioresistance is recognised. The potential of continuous intrinsic susceptibility MRI measurements of tumour R₂* to non-invasively detect acute, cyclic hypoxia through changes in the oxy/deoxyhaemoglobin ratio has been previously demonstrated. In this study we have tested the feasibility of measuring spontaneous fluctuations in tumour R₂* to identify regions with varying erythrocyte or plasma channels flow in head and neck squamous cell carcinoma xenografts on a clinical 3T platform. Additionally, hyperoxia-induced  R₂* and tissue histology were used to spatially confirm hypoxic and perfused tumour regions.

Andy J Kaempf¹, Yiyi Chen¹, Alina Tudorica¹, Stephen Y-C Chui¹, Arpana Naik¹, Karen Y Oh¹, Nicole Roy¹, Megan L Troxell¹, Aneela Afzal¹, Megan L Holtorf¹, Mohan Jayatilake¹, and Wei Huang^1
1Oregon Health & Science University, Portland, OR, United States

DCE-MRI data with ~10 min acquisition time were collected from 14 breast cancer patients (15 primary tumors) who received neoadjuvant chemotherapy (NACT). Data points were dropped incrementally from the far end of DCE time course to simulate ~9, 8, 7, and 6 min acquisition time. The original and simulated data were subjected to standard and shutter-speed model pharmacokinetic analyses. The results suggest 8 min acquisition time is needed to retain the Ktrans and kep abilities as early predictors of therapy response.

Aneela Afzal¹, Alina Tudorica¹, Yiyi Chen¹, Stephen Y-C Chui¹, Arpana Naik¹, Megan L Troxell¹, Kathleen A Kemmer¹, Karen Y Oh¹, Nicole Roy¹, Megan L Holtorf¹, Xin Li¹, and Wei Huang^1
1Oregon Health & Science University, Portland, OR, United States

Breast DCE-MRI data from 29 patients at baseline and after one cycle of neoadjuvant chemotherapy were analyzed using pharmacokinetic (PK) modeling and three different arterial input functions (AIFs). Despite considerable AIF-caused PK parameter variations (except for the Tau_i parameter), several PK parameters after 1 cycle of therapy and their percent changes remained good to excellent early predictors of pathologic response.

Cancer: Other, Original Research

Tuesday 2 June 2015

Weili Zheng¹, Joshua P. Kim¹, Indrin J. Chetty¹, and Carri K. Glide-Hurst^1
1Radiation Oncology, Henry Ford Health System, Detroit, Michigan, United States

Accurate bone segmentation is essential for the generation of (i) electron density maps needed for dose calculation in MR-only radiotherapy treatment planning (RTP) and (ii) MR-based attenuation correction maps in hybrid PET/MRI systems. While UTE sequences improve the visualization of tissues with short T2 such as cortical bone (1), efficient and robust segmentation of bone and air remains a significant challenge (2). To address this need, we introduce an application that uses a novel hybrid magnitude and phase UTE approach to differentiate tissue types, facilitate accurate segmentation, and generate synthetic CTs (synCTs) for the brain.

Tumor Perfusion & Permeability Applications

Tuesday 2 June 2015

Mihaela Rata¹, Matthew R Orton¹, Christina Messiou¹, Helen Young², Nandita de Souza¹, David J Collins¹, and Martin O Leach^1
1Radiotherapy and Imaging Department, CR-UK and EPSRC Cancer Imaging Centre, Institute of Cancer Research and Royal Marsden Hospital, Sutton, Surrey, United Kingdom, ²Early Clinical Development, AstraZeneca, Macclesfield, Cheshire, United Kingdom

Dynamic contrast-enhanced (DCE) MRI method is a technique employed in the assessment of tumour response to novel antiangiogenic therapy. Pharmacokinetic modelling of such data requires a reliable measure of the arterial input function (AIF) in order to obtain robust estimates of physiological parameters characterising tumour properties. This study explore the impact on Ktrans estimate of various sources of uncertainty affecting a MR-DCE acquisition. Three experiments were performed to separate sources of uncertainties. The best CoV of measured Ktrans is still obtained when using a population AIF. A perfect measure of AIF may yet improve a Ktrans repeatability up to 10%.

Hassan Bagher-Ebadian^1,2, Mohammadreza Mohammadian-Behbahani^3,4, Azimeh Noorizadeh Vahed Dehkordi^3,5, James R Ewing^2,6, Alireza Kamali-Asl³, Siamak P Nejad-Davarani⁷, Hamed Moradi⁸, Stephen Brown^2,9, Brent Griffith¹⁰, Ali S Arbab¹¹, Tom Mikkelsen¹², Lisa Scarpace¹², and Hamid Soltanian-Zadeh^1,13
1Radiology and Research Administration, Henry Ford Hospital, Detroit, Michigan, United States, ²Physics, Oakland University, Rochester, Michigan, United States,³Nuclear Engineering, Shahid Beheshti University, Tehran, Iran, ⁴Nuclear Engineering, Amir-Kabir University of Technology, Tehran, Iran, ⁵Nuclear Engineering, Najaf Abad Branch, Islamic Azad University, Isfahan, Iran, ⁶Neurology, Henry Ford Hospital, Detroit, Michigan, United States, ⁷Neurology, Henry Ford Hospital, Michigan, Iran, ⁸Nuclear Engineering, Shiraz University, Shiraz, Fars, Iran, ⁹Radiation Oncology, Henry Ford Hospital, Detroit, Michigan, United States,¹⁰Radiology, Henry Ford Hospital, Detroit, Michigan, United States, ¹¹GRU Cancer Center, Georgia Regents University, Atlanta, Georgia, United States,¹²Neurosurgery, Henry Ford Hospital, Detroit, Michigan, United States, ¹³CIPCE, School of Electrical and Computer Engineering, University of Tehran, Tehran, Iran

Given Dynamic-Contrast-Enhanced MRI data, accurate estimation of Pharmacokinetic (PK) parameters strongly relies on appropriate selection of the best PK model to fit the data. This study investigates the impact of different classical and adaptive MS technique such as F-Statistic (F-test), Akaike-Information-Criterion (AIC), Bayesian-Information-Criterion (BIC), Log-Likelihood-Ratio (LLR) and Artificial Neural Network (ANN) on estimation of vascular permeability parameters. Results imply that ANN generates significantly less-biased estimates of PK parameters compared to other techniques while both the LLR and BIC methods outperform the other classical MS techniques, the ANN, LLR and BIC are the best candidates for PK analysis of DCE-MRI data.

Hyunki Kim¹, Sharon Samuel¹, Marie Warren¹, Guihua Zhai¹, William Grizzle¹, Denise Oelschlager¹, Pedro Lopez-Casas², Manuel Hidalgo², Joy Kovar³, Kurt Zinn¹, and Donald Buchsbaum^1
1University of Alabama at Birmingham, Birmingham, AL, United States, ²Spanish National Cancer Research Center, Madrid, Spain, ³LI-COR Biosciences, Nebraska, United States

Dynamic contrast enhanced magnetic resonance imaging (DCE-MRI) was employed to identify the therapeutic mechanism of nab-paclitaxel in pancreatic cancer patient derived xenograft mouse models. DCE-MRI demonstrated that nab-paclitaxel did not significantly increase the microvascular perfusion in tumor xenografts for 3 weeks of nab-paclitaxel treatment, although gemcitabine did. Also, histologic analysis showed that tumor stroma was not significantly reduced by nab-paclitaxel, but high tumor retention of fluorophore labeled nab-paclitaxel was observed. Therefore the significant clinical benefits of nab-paclitaxel may result from its effective delivery into tumors in pancreatic cancer.

Yue Zhang¹, Payal Kapur^2,3, Qing Yuan¹, Ananth Madhuranthakam^1,4, Ingrid Carvo⁵, Sabina Signoretti⁵, Ivan Dimitrov⁶, Yin Xi¹, Katherine Wicks¹, Jeffrey Cadeddu^1,3, Vitaly Margulis³, James Brugarolas^7,8, and Ivan Pedrosa^1,4
1Radiology, University of Texas Southwestern Medical Center, Dallas, Texas, United States, ²Pathology, University of Texas Southwestern Medical Center, Dallas, Texas, United States, ³Urology, University of Texas Southwestern Medical Center, Dallas, Texas, United States, ⁴Advanced Imaging Research Center, University of Texas Southwestern Medical Center, Dallas, Texas, United States, ⁵Pathology, Brigham and Women’s Hospital, Boston, Massachusetts, United States, ⁶Philips Medical Systems, Cleveland, Ohio, United States, ⁷Internal Medicine, University of Texas Southwestern Medical Center, Dallas, Texas, United States,⁸Developmental Biology, University of Texas Southwestern Medical Center, Dallas, Texas, United States

The purpose was to investigate if the central non-enhancing defect seen commonly on contrast-enhanced MRI in clear cell renal cell carcinoma (ccRCC) correlates with coagulation necrosis at histopathology in the same tumor, a known poor prognostic feature for these tumors. We found that non-enhancing defects in ccRCC represent tumor scars instead of coagulative necrosis. Ki-67 staining revealed significantly lower cell proliferation rate for RCC with scar than that for those without scar. We hypothesize that this scar is the result of rapid enlargement of the tumor leading to a central infarct and may be formed during a subsequent reparative phase.

Rafal Panek¹, Liam Welsh¹, Maria A. Schmidt¹, Kate L. Newbold¹, Kee Wong¹, Angela M. Riddell¹, Dow-Mu Koh¹, Alex Dunlop¹, Dualta Mcquaid¹, Shreerang A. Bhide¹, Kevin J. Harrington², Christopher M. Nutting², Georgina Hopkinson³, Cheryl Richardson³, Simon P. Robinson⁴, and Martin O. Leach^1
1Royal Marsden NHS FT and Institute of Cancer Research, Sutton, Surrey, United Kingdom, ²Royal Marsden NHS FT and Institute of Cancer Research, London, United Kingdom, ³Royal Marsden NHS FT, London, United Kingdom, ⁴Institute of Cancer Research, Sutton, Surrey, United Kingdom

In this study we tested the feasibility of measuring spontaneous fluctuations in the transverse relaxation rate, R2*, to identify regions with intermittent erythrocyte and plasma flow, and hence oscillating oxygen delivery, in patients with advanced squamous cell carcinoma of the head and neck. Non-random fluctuations were detected in parts of lymph nodes with low Ktrans and R2* values, often in the vicinity, but not in the centre, of necrotic nodal core, suggesting the presence of fluctuating blood oxygen levels, which could be caused by an insufficient or intermittent blood delivery.

Cancer: Other Cancers

Tuesday 2 June 2015

Durgesh Kumar Dwivedi^1,2, Girdhar Singh Bora³, Rajeev Kumar³, Sanjay Sharma⁴, Sanjay Thulkar⁴, Siddhartha Datta Gupta⁵, and Naranamangalam Raghunathan Jagannathan^2
1Radiodiagnosis, King George's Medical University, Lucknow, U.P., India, ²NMR & MRI Facility, All India Institute of Medical Sciences, New Delhi, Delhi, India,³Urology, All India Institute of Medical Sciences, New Delhi, Delhi, India, ⁴Radiodiagnosis, All India Institute of Medical Sciences, New Delhi, Delhi, India,⁵Pathology, All India Institute of Medical Sciences, New Delhi, Delhi, India

RCC is a heterogeneous disease and therapy relies on accurate histologically subtype of tumors. mpMRI have shown its potential in diagnosing various malignancies. We studied RCC lesions using MRS and DW-MRI. MRS showed increased signal to noise ratio in mobile lipids in RCC and the resonance at 5.4 ppm was significantly higher in RCC. ADC was significantly decreased in RCC as compared to normal kidney. mpMRI could reliably differentiate malignant tissue from normal kidney parenchyma.

Arash Latifoltojar¹, Paul Humphries², Stuart Taylor³, Ananth Shankar², Stephen Daw², and Shonit Punwani^3
1University College London, London, London, United Kingdom, ²University College London Hospital, London, United Kingdom, ³University College London, London, United Kingdom

Positron emission tomography remains gold standard imaging for initial investigation and monitoring treatment response in paediatric lymphoma despite associated radiation exposure risk factors. Whole body MRI (WBMRI) provides a non-ionising alternative imaging modality for assessment of this group of patients. Anatomical and functional imaging such as diffusion weighted imaging can be implemented in WBMRI protocols for a comprehensive assessment of different aspects of suspected malignancy. In this work we investigated the diagnostic performance of whole body anatomical and functional imaging compared to reference standard PET-CT.

Breast Cancer Clinical

Tuesday 2 June 2015

Yuan Le¹, Hal D Kipfer¹, Shadie S Majidi¹, Brian Dale², Marcel Dominik Nickel³, Randall Kroeker², Elisabeth Weiland³, and Chen Lin^1
1Radiology and Imaging Science, Indiana University School of Medicine, Indianapolis, IN, United States, ²Siemens Medical Solutions, NC, United States, ³Siemens Healthcare, Erlangen, Bavaria, Germany

A modified Time-resolved angiography With Stochastic Trajectories (TWIST) sequence with flexible view sharing and dual-echo DIXON was evaluated in clinical breast DCE-MRI for its capability of measuring initial uptake phase with fast acquisitions in addition to providing high resolution peak- and post-contrast images. A good correlation between the results based on the maximum slope of the signal enhancement in the initial phase and that based on the standard protocol was found, indicating a potential of shortening the breast MRI scan time and in turn a lowering of the breast MRI cost with this new technique.

Sunitha B Thakur¹, Manuela Durando², Milans Soledad³, Elizabeth J Sutton², Dilip Giri², and Elizabeth A Morris^2
1Memorial Sloan Kettering Cancer Center, New York, NY, United States, ²Memorial Sloan Kettering Cancer Center, NY, United States, ³Memorial Sloan Kettering Cancer Center, New YORK, NY, United States

Preliminary studies have attempted to evaluate Apparent Diffusion Coefficient (ADC) values for its potential prognostic utility. However the impact of ADC on prognosis remains to be clearly defined. This study was conducted with the goal of determining whether there exists a correlation between ADC values and Oncotype Dx scores in estrogen-receptor positive and lymph node negative breast cancers. We found that mean ADC values were higher in Oncotype DX score stratified low risk cancers than in intermediate risk cancers. Our data at 3.0T MRI suggests that lower ADC values correlate with a higher Oncotype DX score and consequently, an increased risk of recurrence.

Naranamangalam R Jagannathan¹, Khushbu Agarwal¹, Uma Sharma¹, Smriti Hari², Vurthaluru Seenu³, and Rajinder Parshad^3
1Department of NMR & MRI Facility, All India Institute of Medical Sciences, New Delhi, Delhi, India, ²Department of Radiodiagnosis, All India Institute of Medical Sciences, New Delhi, Delhi, India, ³Department of Surgical Disciplines, All India Institute of Medical Sciences, New Delhi, Delhi, India

ChoSNR and ADC were measured from same voxels in different breast tumor regions (solid tumor, necrosis and margin) prior to and after III NACT in 11 LABC patients (6 clinical responders and 5 non-responders). In responders, increased ChoSNR with decreased ADC of solid tumor was observed after therapy. ChoSNR decreased significantly with no ADC change at tumor margin. At necrosis no change in ChoSNR or ADC was seen. In non-responders ChoSNR and ADC did not change in tumor regions. Positive Pearson correlation between ChoSNR and ADC change signifies combined utility of both techniques simultaneously in therapeutic management of breast cancer.

Katja Pinker¹, Pascal Baltzer¹, Wolfgang Bogner², Doris Leithner¹, Siegfried Trattnig², Olgica Zaric², Peter Dubsky³, Rupert Bartsch⁴, Zsuzsanna Bago-Horvath³, Stephan Gruber², Michael Weber¹, and Thomas H Helbich^1
1Dept. of Biomedical Imaging and Image-guided Therapy, Division of Molecular and Gender Imaging, Medical University of Vienna, Vienna, Austria, ²Dept. of Biomedical Imaging and Image-guided Therapy, MR Centre of Excellence, Medical University of Vienna, Vienna, Vienna, Austria, ³Dept. of Surgery, Medical University of Vienna, Vienna, Austria, ⁴Dept. of Internal Medicine, Division of Oncology, Medical University of Vienna, Vienna, Austria

The clinical use of multiparametric MR imaging at 7T is feasible with good or excellent image quality. Multiparametric MR imaging of the breast with DCE MR imaging and DWI at 7T seems to have the potential to diagnose breast cancer with high diagnostic accuracy (AUC 0.941). Multiparametric MR imaging of the breast with DCE MR imaging and DWI at 7T improves specificity as compared to interpretation of DCE MRI alone (p=0.031). Multiparametric MR imaging of the breast with DCE MR imaging and DWI at 7T can avoid unnecessary breast biopsies recommended with DCE MR imaging, in our series in 6/8 (75%, p=0.031) of benign breast tumors.

Lawrence Kenning¹, Martin Pickles¹, and Lindsay Turnbull^1
1Centre for MR Investigations, Hull York Medical School at University of Hull, Hull, United Kingdom

Motion during breast MR examination can degrade DCE data and cause spatial mismatches between sequences. Data acquired from 19 patients was retrospectively corrected using rigid and non-rigid schemes. For the DCE series, mean coefficients of variation were found to be 45.2±3.9%, 40.0±3.1% and 35.7±2.5% for the original data, rigid motion corrected data and the non-rigid motion corrected data. Non-rigid registration between MR series was successfully applied to all patients and showed good visual agreement. The global 4D registration methodology enables researchers and clinicians to contour single volume of interests that can be used to simultaneously interrogate multiple breast MR series.

Elizabeth O'Flynn¹, David Collins¹, James D'Arcy¹, Maria Schmidt¹, and Nandita deSouza^1
1CRUK Cancer Imaging Centre, The Institute of Cancer Research, Sutton, Surrey, United Kingdom

The purpose of this study was to compare multiparametric MRI parameters of ER positive tumours in pre- and postmenopausal women to assess any differences in tumour functional characteristics and to establish which parameter correlated significantly with response as defined by tumor volume reduction.In premenopausal women with ER positive breast cancer, tumors were larger and more vascular than in postmenopausal women. A rise in the ve correlated significantly with volume reduction in premenopausal women; no specific parameter correlated with response in postmenopausal women.

Yuan Jiang¹, Naishan Qin¹, Xiaoying Wang¹, and Li Guo^1
1Radiology Department, Peking University First Hospital, Beijing, Beijing, China

Different molecular subtypes of breast cancer respond differently to neoadjuvant chemotherapy. Patients who achieve pathologic complete response (pCR) have better prognosis than non-pCR. In our study, semi-quantitative analysis of dynamic contrast-enhanced (DCE) MR Imaging between pCR and non-pCR was performed in four types of breast cancer (Luminal A, Luminal B, HER2+ and Triples negative), and our results showed that DCE-MR Imaging could become the potential predictor of pCR to NAC in breast cancer with Luminal B and Triples negative types. Maybe further study should be performed based on molecular subtypes of breast cancer.

Cheng-Liang Liu¹, Savannah C Partridge¹, Diana L Lam¹, Constance D Lehman¹, and Habib Rahbar^1
1Department of Radiology, University of Washington, Seattle, Washington, United States

Background parenchymal enhancement (BPE) on MRI has been proposed to be a biomarker of breast cancer risk. We sought to develop an optimal method to measure BPE quantitatively for breast cancer risk assessment. By measuring various BPE metrics at enhancement thresholds ranging from 5-100% in a case-control (n=36), we found that quantitative BPE measures are higher in women who developed breast cancer than in controls, with a 70% enhancement threshold for BPE area providing the highest accuracy for predicting risk. Our findings suggest quantitative BPE measures can assess breast cancer risk, potentially allowing individualized screening and prevention strategies.

Naranamangalam R Jagannathan¹, Khushbu Agarwal¹, Rani G Sah¹, Uma Sharma¹, Smriti Hari², Vurthaluru Seenu³, and Rajinder Parshad^3
1Department of NMR & MRI Facility, All India Institute of Medical Sciences, New Delhi, Delhi, India, ²Department of Radiodiagnosis, All India Institute of Medical Sciences, New Delhi, Delhi, India, ³Department of Surgical Disciplines, All India Institute of Medical Sciences, New Delhi, Delhi, India

ADC were measured at whole tumor (WT), and from five geographical zones from inside tumor margin (IM) in a layerwise manner to outside tumor margin III (OM1, OM2, OM3) and intra-tumoral necrotic domains (Nec) sequentially i.e., prior to (Tp0), after I (Tp1) and III (Tp3) neoadjuvant chemotherapy (NACT) in 36 breast cancer patients. A significant increase in ADC was seen for WT and IM, while it decreased at OM1 compared to OM2 and OM3 at Tp3. The change in ADC of WT and nec after therapy showed positive Pearson correlation. Results indicated the utility of DWI in breast conservation surgery.

Edna Furman-Haran¹, Dov Grobgeld², Noam Nissan², Myra Feinberg-Shapiro³, Tania Zehavi³, Zvi Kaufman³, and Hadassa Degani^2
1Department of Biological Services, The Weizmann Institute of Science, Rehovot, Israel, ²Department of Biological Regulation, The Weizmann Institute of Science, Rehovot, Israel, ³Meir Medical Center, Kfar Saba, Israel

DTI tracks the water diffusion in different directions and thereby allows characterizing breast microstructures with restricted diffusion, revealing their organization in vector/ellipsoid-maps and parametric-maps of anisotropy-indices, derived from the diffusion tensor eigenvectors and eigenvalues, respectively. The study evaluates of the various anisotropy-indices investigating their ability to detect breast cancer. The results indicate that the commonly used normalized anisotropy-indices fail to differentiate cancer from normal tissue, while the un-normalized maximal anisotropy index is significantly lower in cancer as compared to normal tissue and can be efficiently used for detecting breast cancer in conjunction with the reduction in the diffusion coefficients.

Shiteng Suo¹, Fang Cheng¹, He Wang², Jia Hua¹, and Jianrong Xu^1
1Department of Radiology, Ren Ji Hospital, School of Medicine, Shanghai Jiao Tong University, Shanghai, China, China, ²Philips Research Chinia, Shanghai, China, China

In the study, the preliminary results showed that the water diffusion followed a non-Gaussian behavior by using diffusion kurtosis imaging (DKI) in breast lesions. The malignant tumors tended to have a restricted diffusion pattern and an increased K value compared to fibroadenoma. With the capability to quantify the deviation from Gaussian distribution, the K value may give insight into the microstructure complexity in vivo.

Niloufar Fozouni¹, Cheng-Liang Liu¹, Habib Rahbar¹, Constance D Lehman¹, and Savannah C Partridge^1
1Department of Radiology, University of Washington, Seattle, Washington, United States

Diffusion-weighted imaging (DWI) holds potential for reducing false positives on conventional breast MRI based on differences in apparent diffusion coefficient (ADC) values between benign and malignant lesions. However, standard DWI approaches can result in ADC values strongly influenced by perfusion, potentially increasing overlap in ADC between pathologies. In this study, we found that separation of perfusion and diffusion-weighted components of lesion ADC measures using a simplified approach can improve diagnostic performance. Furthermore, we found that a multiparametric combination of DWI and DCE features can provide more accurate assessment of suspicious MRI-detected breast lesions.

Jing Yuan¹, Gladys G. Lo², Oi Lei Wong¹, Helen H.L. Chan², Abby Y. Ding¹, Ting Ting Wong³, and Polly S.Y. Cheung^3
1Medical Physics and Research Department, Hong Kong Sanatorium & Hospital, Happy Valley, Hong Kong, China, ²Department of Diagnostic & Interventional Radiology, Hong Kong Sanatorium & Hospital, Happy Valley, Hong Kong, China, ³Breast Care Center, Hong Kong Sanatorium & Hospital, Happy Valley, Hong Kong, China

This study aims to investigate the micro-vascularity of fibroglandular tissues (FGTs) in breast cancer (BC) using intravoxel incoherent motion (IVIM) MRI. 21 histology-confirmed patients received IVIM-MRI at 3T. True diffusion D, pseudo-diffusion D*, pseudo-diffusion fraction f and ADC in ipsilateral and contralateral breast FGTs were compared. Results show that FGT D* and f in the ipsilateral tumor-bearing breast were higher than those in the contralateral breast (D*: 24.96¡Ó9.37¡Ñ10^-3mm²/s v.s. 19.98¡Ó9.37¡Ñ10^-3mm²/s; f: 0.123¡Ó0.034 v.s. 0.109¡Ó0.030), with D* being significant (p=0.033). FGT microvascular variation revealed by IVIM-MRI may potentially be used for breast cancer risk assessment, characterization and/or prognosis

Jason Ostenson¹, Linda Moy¹, Sungheon G. Kim¹, Amy Melsaether¹, Komal Jhaveri², Christian Geppert³, David Faul³, Francisco Esteva², Sylvia Adams², Freya Schnabel⁴, Kimberly Jackson¹, Joon Lee¹, Christopher Glielmi³, Gene Young Cho^1,5, Thorsten Feiweier⁶, and Eric E. Sigmund^1
1Department of Radiology, NYU Langone Medical Center, New York, NY, United States, ²Perlmutter Cancer Center, NYU Langone Medical Center, New York, NY, United States, ³Siemens Medical Solutions, New York, NY, United States, ⁴Department of Surgery, NYU Langone Medical Center, New York, NY, United States,⁵Sackler Institute of Graduate Biomedical Sciences, NYU School of Medicine, New York, NY, United States, ⁶Siemens AG, Erlangen, Germany

Simultaneously acquired diffusion weighted imaging and 18-fluorodeoxyglucose(FDG)-PET are employed to explore the relationship between glucose avidity, diffusion, and perfusion in breast cancer. Intravoxel incoherent motion metrics are calculated and combined with correlated FDG-PET standard uptake values (SUV) for all voxels in 16 histopathologically proven breast cancer lesions. Two-dimensional histograms and Pearson’s correlations are used to define three lesion types. Each lesion’s voxels are separated into three groups according to their SUV, diffusion, and perfusion fraction. The results yield insight into the heterogeneity of the tumor microenvironment and show significant correlations with estrogen receptor expression and a cellular proliferation marker (Ki-67).

Yuan Le¹, Hal D Kipfer¹, Marcel Dominik Nickel², Randall Kroeker³, Stephanie P Holz¹, Elisabeth Weiland², and Chen Lin^1
1Radiology and Imaging Science, Indiana University School of Medicine, Indianapolis, IN, United States, ²Siemens Healthcare, Erlangen, Bavaria, Germany,³Siemens Medical Solutions, NC, United States

Time-resolved angiography With Stochastic Trajectories (TWIST) Sequence was modified with flexible TWIST view sharing and flexible echo time Dixon for the acquisition of images with high and flexible spatial and temporal resolution as well as uniform fat suppression. Our initial experience shows that it provides potentially valuable information on early tumor enhancement characteristics while maintaining excellent image quality at post-contrast enhancement in breast DCE-MRI, allowing more optimized spatial and temporal resolution in clinical breast imaging.

Cancer: Others

Tuesday 2 June 2015

Hyeon-Man Baek^1,2, Youngjae Jeon¹, Jooyun Kim¹, and Mirim Bang^1
1Center for MR Research, Korea Basic Science Institute, Ochang, Chungbuk, Korea, ²Department of Bio-Analytical Science, University of Science & Technology, Daejeon, Chungnam, Korea

The aim of this study was two-fold: first, to investigate the metabolic profiling analysis based on 13C-NMR spectroscopy with stable 13C-labeled isotope, and second, to demonstrate whether 2HG labeling from [U-13C]glucose substrate feeding could be detected in IDH1 or 2 mutated cells.The analysis of 1H- and 13C-NMR spectra of the cell extracts showed a significant increase in the concentration of the 2HG in IDH mutated cells, but not in IDH wild type and mutant IDH cells with low transfection efficiency.

Chiara Giraudo¹, Michael Weber¹, Markus Raderer², Georgios Karanikas¹, and Marius Erik Mayerhoefer^1
1Departments of Biomedical Imaging and Image-guided Therapy, Medical University of Vienna, Vienna, Austria, ²Internal Medicine I, Medical University of Vienna, Vienna, Austria

This study was performed to assess the diagnostic value of PET/MR and DWI in the evaluation of nodal and extranodal lymphoma. PET/MR is an emerging hybrid technique that has already demonstrated very good results for the diagnosis of malignancies in different anatomical districts but up to now, only few data are available in the literature about its application on patients affected by lymphoma. Our initial results demonstrated that, for both nodal and extranodal disease, PET/MR is a highly promising technique and DWI confirmed to be an essential tool to improve the diagnostic work-up of this heterogeneous cancer group.

Peter Mazurkewitz¹, Daniel Bystrov¹, Peter Koken¹, Torbjoern Vik¹, and Julien Sénégas^1
1Philips Research Laboratories, Hamburg, Germany

In this work, we present a robust and accurate method for the automated planning of scan geometry in follow-up prostate exams. In order to better cope with the specific challenges of prostate MR exams, where high-resolution scans with small FOV generally need to be planned, the rigid registration method presented recently was adapted by using an image based organ detection as initialization of the registration algorithm. This step ensures a robust and stable registration. This new approach was evaluated retrospectively on volunteer data and tested on a commercial MRI system.

Cecilia Besa¹, Guido Jajamovich², Adnan Ali³, Wei Huang⁴, Kenneth Haines⁵, Ash Tewari³, and Bachir Taouli^6
1Radiology, Icahn School of Medicine at Mount Sinai, New York, NY, United States, ²icahn School of Medicine at Mount Sinai, NY, United States, ³Urology, Icahn School of Medicine at Mount Sinai, NY, United States, ⁴Radiology, Oregon Health & Science University, Portland, OR, United States, ⁵Pathology, Icahn School of Medicine at Mount Sinai, NY, United States, ⁶Icahn School of Medicine at Mount Sinai, New York, NY, United States

Synopsis: The purpose of this study was to evaluate the value of quantitative diffusion weighted imaging (DWI) and dynamic contrast enhanced (CE) MRI in predicting histopathologic characteristics of liver metastatic neuroendocrine tumors (NET). We found that neuroendocrine carcinoma (G3) liver metastases had significantly lower ADC values and higher arterial enhancement rate (ER) than Grade 1 and 2 NETs. In addition, significant negative correlation was observed between ADC and ER and mitotic count and Ki-67% labeling index. DWI with ADC quantification and CE-MRI may be useful for predicting tumor grade in metastatic hepatic NET.

Elaheh Kia^1,2, Anahita Fathi Kazerooni^1,2, Saeedeh Navaei Lavasani^1,2, Alireza Ahmadian², and Hamidreza Saligheh Rad^1,2
1Quantitative MR Imaging and Spectroscopy Group, Research Center for Molecular and Cellular Imaging, Tehran University of Medical Sciences, Tehran, Iran,²Department of Medical Physics and Biomedical Engineering, School of Medicine, Tehran University of Medical Sciences, Tehran, Iran

Dynamic contrast enhanced MRI (DCE-MRI) has shown to be promising for quantitative assessment of complex ovarian cancers. Quantification of DCE-MR images could be affected by motion artifacts and intensity inhomogeneity induced by bias fields. Proper selection of a registration algorithm could impact the outcome of this problem. In this work, we proposed an efficient non-rigid registration method in a group-wise setting to non-rigidly align DCE-MR images to reliable quantification of ovarian masses.