Alina Tudorica¹, David C Newitt², Karen Y Oh¹, Nicole Roy¹, Stephen Y-C Chui¹, Arpana Naik¹, Megan L Troxell¹, Yiyi Chen¹, Aneela Afzal¹, Megan L Holtorf¹, Nola M Hylton², and Wei Huang^1
1Oregon Health & Science University, Portland, OR, United States, ²University of California, San Francisco, CA, United States

High-temporal-resolution (17-20 s) DCE-MRI data were collected from 14 breast cancer patients (15 tumors) before and after one cycle of neoadjuvant chemotherapy. The original data were analyzed with pharmacokinetic (PK) modeling, while the resampled low-temporal-resolution (80-100 s) data were used to quantify functional tumor volumes (FTV) based on threshold values of early percent signal change and signal enhancement ratio. Empirical analysis based FTV metrics performed comparably for early prediction of pathologic response in comparison with quantitative PK parameters.

Xia Li¹, Lori R. Arlinghaus¹, Erin Rericha¹, and Thomas Yankeelov^1
1Vanderbilt University, Nashville, Tennessee, United States

Many pharmacokinetic models have been proposed for analyzing dynamic contrast enhanced MRI with different assumptions. It is difficult to select which model is most appropriate for a particular study. To address this limitation, we have taken an approach in which multiple models are weighted by a factor determined by how well they fit the data. We analyze each voxel in a DCE-MRI data set with an array of models and compute the Akaike Information Criteria for each fit. The AIC is then used to determine the statistically optimal model on a voxel-by-voxel basis.

Venkata Veerendra Nadh Chebrolu¹, Dattesh Shanbhag¹, Reem Bedair², Sandeep Gupta³, Patrice Hervo⁴, Scott Reid⁵, Fiona Gilbert², Andrew Patterson⁶, Martin Graves⁷, and Rakesh Mullick^8
1Medical Image Analysis Lab, GE Global Research, Bangalore, Karnataka, India, ²Radiology, University of Cambridge, Cambridge, United Kingdom,³Biomedical Image Analysis Lab, GE Global Research, NY, United States, ⁴GE Healthcare, Buc, France, ⁵GE Healthcare, Amersham, United Kingdom,⁶Cambridge University Hospitals Trust, Cambridge, United Kingdom, ⁷Radiology, Cambridge University Hospitals Trust, Cambridge, United Kingdom,⁸Diagnostics & Biomedical Technologies, GE Global Research, Bangalore, Karnataka, India

Dynamic contrast-enhanced (DCE)-MRI is increasingly being used in understanding the pharmaco-kinetic (pK) characteristics of breast tumors. The Ktrans and ve estimates used in tumor characterization depend on the DCE concentration time courses. Cardiac pulsation, breathing and voluntary patient motion introduce rigid and non-rigid motion, which may corrupt the DCE analysis. Previous work demonstrated symmetric diffeomorphic normalization (SyN) to be more effective than b-splines based approaches in correcting for motion in breast DCE MRI. In this work we demonstrate the impact of SyN based non-rigid registration on pK modelling for breast DCE-MRI.

Madhava P Aryal¹, Tavarekere N Nagaraja², Rasha Elmghribi^1,3, Kelly A Keenan², Swayamprava Panda¹, Glauber Cabral¹, Stephen L Brown⁴, and James R Ewing^1,3
1Dept. of Neurology, Henry Ford Hospital, Detroit, MI, United States, ²Dept. of Anesthesiology, Henry Ford Hospital, Detroit, MI, United States, ³Dept. of Physics, Oakland University, Rochester, MI, United States, ⁴Dept. of Radiation Oncology, Henry Ford Hospital, Detroit, MI, United States

In many solid tumors an elevated tumor interstitial fluid pressure (TIFP) is a central and potentially critical element for assessing therapeutic response. The present study examined the possibility of evaluating TIFP non-invasively using dynamic contrast enhanced MRI (DCE-MRI) data in a rat model of cerebral glioma. A significant correlation between DCE-MRI derived TIFPs with those of WIN estimates (n=9, R=0.77, p≤0.02) was established. These data strongly demonstrate the possibility of employing DCE-MRI for non-invasive estimates of TIFP in solid tumors, and also provides a means for estimating fluid conductivity in tumor surround.

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

Our aim was to investigate the correlation between in vivo perfusion measurement of renal masses by arterial spin labeling (ASL) and dynamic contrast-enhanced (DCE) MRI at 3T and validate these results against microvessel density (MVD) at histopathology. Quantitative ASL perfusion (P_ASL) and DCE parameters were obtained from the whole tumor, and from areas with high- and low- perfusion within the tumor. ASL measured blood flow in renal masses correlates with DCE-derived parameters.P_ASL, K^trans, and Kep correlate with MVD. ASL and DCE MRI at 3T provide a tool for assessing angiogenesis in renal cancer.

Jesper Folsted Kallehauge^1,2, Thomas Nielsen³, Markus Alber¹, Søren Haack^2,4, Erik Morre Pedersen⁵, Jacob Christian Lindegaard², Anne Ramlov², and Kari Tanderup^6,7
1Dept. of Medical Physics, Aarhus University Hospital, Aarhus, Denmark, ²Dept. of Oncology, Aarhus University Hospital, Aarhus, Denmark, ³CFIN/Mindlab, Aarhus University Hospital, Aarhus, Denmark, ⁴Dept. of Clinical Engineering, Aarhus University Hospital, Aarhus, Denmark, ⁵Dept. of Radiology, Aarhus University Hospital, Aarhus, Denmark, ⁶Dept. of Experimental Clinical Oncology, Aarhus University Hospital, Aarhus, Denmark, ⁷Dept. of Clinical Medicine, Aarhus University, Aarhus, Denmark

We applied a set of hierarchical tracer kinetic models to DCE-MRI data acquired prior to patient treatment in cervical cancer patients. This study identified contiguous regions where given kinetic models were optimal and identified that high volume fraction reflecting tissue enhancement best described by Tofts model correlated with tumor stage.

Chunling Liu¹, Changhong Liang¹, Yingjie Mei², Zaiyi Liu¹, and Jine Zhang^1
1Department of Radiology, Guangdong General Hospital/Guangdong Academy of Medical Sciences, Guangzhou, Guangdong, China, ²Philips Healthcare, Guangzhou, Guangdong, China

Breast cancer is the most commonly diagnosed type of cancers of women. In this study, we obtained the parameters (ADCs and ¦Á) information using a stretched-exponential model with high b-values for normal breast and breast lesions and compared them to apparent diffusion coefficient (ADC) with conventional DWI in their ability to discriminate benign lesions and malignant tumors. 77 patients underwent a standard bilateral breast MRI examination including DWI sequence and ADCs, ¦Á and ADC were obtained according to different diffusion models. The statistical analysis results suggest that ADCs of the stretched model can be used to differentiate benign and malignant lesions and has the highest specificity, comparing the conventional ADC, however, ¦Á is not different between benign and malignant tumors and may be further investigated.

Yachao Liu¹, Jiangping Gao², Jiajin Liu¹, Hui Liu³, Yong Xu², Baixuan Xu¹, and Jiahe Tian^1
1Nuclear Medicine Department, PLA 301 General Hospital, Beijing, Beijing, China, ²Urology Department, PLA 301 General Hospital, Beijing, China, ³NEA MR Collaboration, Siemens Ltd., China, Shanghai, China

Although 11C-Choline PET/CT has approved a valuable tools for recurrent prostate cancer detection and treatment evaluation. However, it is still a problem for primary prostate cancer. The combined MRI with 11C-Choline PET could be potentially useful for primary prostate cancer. We are evaluating the combined features of MRI and PET to improve the diagnostic performance of primary prostate.

Mattijs Elschot¹, Kirsten M. Selnæs^1,2, Brage Krüger-Stokke^1,3, Øystein Størkersen⁴, Helena Bertilsson^5,6, Siver A. Moestue^1,2, and Tone F. Bathen^1,2
1Department of Circulation and Medical Imaging, Norwegian University of Science and Technology, Trondheim, Sør-Trøndelag, Norway, ²St Olavs Hospital, Trondheim, Sør-Trøndelag, Norway, ³Department of Radiology, St Olavs Hospital, Trondheim, Sør-Trøndelag, Norway, ⁴Department of Pathology, St Olavs Hospital, Trondheim, Sør-Trøndelag, Norway, ⁵Department of Urology, St Olavs Hospital, Trondheim, Sør-Trøndelag, Norway, ⁶Department of Cancer Research and Molecular Medicine, Norwegian University of Science and Technology, Trondheim, Sør-Trøndelag, Norway

¹⁸F-FACBC PET/CT imaging has shown promise for the assessment of primary and metastatic prostate cancer, especially in conjoint use with MRI. Simultaneous PET/MRI may therefore be an interesting option, but designing the imaging protocol is complicated by the fast kinetics of ¹⁸F-FACBC and the long scan time required for multiparametric MRI. Based on the dynamic uptake of ¹⁸F-FACBC in 11 patients that underwent PET/MRI for pre-operative staging of high-risk prostate cancer, we propose a realistic protocol that includes 45 minutes of PET imaging in addition to all MRI sequences required for clinical evaluation of the prostate and pelvic lymph nodes.

Jennifer Mosebach¹, Christos Sachpekidis², Martin Freitag¹, Jens Hillengass³, Antonia Dimitrakopoulou-Strauss², Uwe Haberkorn⁴, Heinz-Peter Schlemmer¹, and Stefan Delorme^1
1Department of Radiology, German Cancer Research Center, Heidelberg, Germany, ²Clinical Cooperation Unit Nuclear Medicine, German Cancer Research Center, Heidelberg, Germany, ³Department of Medicine V, Multiple Myeloma Section, University of Heidelberg, Heidelberg, Germany, ⁴Division of Nuclear Medicine, University of Heidelberg, Heidelberg, Germany

Hybrid PET/MRI examinations offer an innovative approach to simultaneously depict metabolic as well as functional and anatomical MR parameters. Fifteen patients newly diagnosed with Multiple Myeloma received whole body PET/MRI including Diffusion-weighted Imaging (DWI). Lesion by lesion based analysis as well as comparison of standard-uptake-values (SUV) and Apparent-diffusion-coefficient (ADC)- values were conducted. For the detection of focal bone marrow infiltrates, PET and DWI can be complementary, with PET displaying more lesions located in ribs, whereas DWI seems to have advantages discovering lesions masked by the surrounding diffuse myeloma infiltrate. Those patients could therefore especially benefit from multimodality imaging.

Rajakumar Nagarajan¹, Neil Wilson¹, Nanette DeBruhl¹, Brian Burns¹, Melissa Joines¹, Maithili Gopalakrishnan¹, Fausto Rendon¹, Lawrence W Bassett¹, and M.Albert Thomas^1
1Radiological Sciences, UCLA School of Medicine, Los Angeles, CA, United States

Early detection and diagnosis are essential for successful treatment of breast cancer. Magnetic Resonance Spectroscopy (MRS) can provide high specificity for distinguishing benign from malignant lesions. The four dimensional (4D) echo-planar correlated spectroscopic imaging (EP-COSI) enables full slice coverage of the breast facilitating recording of multi-voxel based two-dimensional (2D) MRS than the single-voxel based localized correlated spectroscopy (L-COSY). The 4D EP-COSI offers improved spectral dispersion and sensitivity compared with 1D MRS data. Decreased ADC values derived from diffusion weighted imaging (DWI) can be correlated with increased choline and decreased lipids quantified by the EP-COSI technique.

Stefan Zbyn¹, Olgica Zaric¹, Vladimir Juras¹, Katja Pinker², Alex Farr³, Nadia Benkhedah⁴, Pascal Balzer², Vladimir Mlynarik¹, Armin Nagel⁴, Christian Singer³, Thomas Helbich², Wolfgang Bogner¹, and Siegfried Trattnig^1
1High Field MR Center, 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, ³Department of Gynecology and Obstetrics, Medical University of Vienna, Vienna, Austria, ⁴Department of Medical Physics in Radiology, German Cancer Research Center (DKFZ), Heidelberg, Germany

The aim of this study was to compare total sodium signal with relaxation-weighted sodium signal in breast lesions at 7T. While conventional sodium density images showed a signal proportional to the local sodium concentration, sodium ions with a longer T2*, such as in a cyst or edema, were well suppressed in relaxation-weighted images. Our data demonstrated improved differentiation between edema, benign lesion and breast cancer with relaxation-weighted sodium images than with conventional sodium density images. Combined information from conventional and from relaxation-weighted sodium images may improve noninvasive evaluation of breast lesions.

Manus Donahue^1,2, Paula CM Donahue^3,4, Swati Rane¹, Megan K Strother¹, Allison O Scott¹, and Seth A Smith^1
1Radiology and Radiological Sciences, Vanderbilt University Medical Center, Nashville, TN, United States, ²Physics and Astronomy, Vanderbilt University, Nashville, TN, United States, ³Physical Medicine and Rehabilitation, Vanderbilt University Medical Center, Nashville, TN, United States, ⁴Dayani Center for Health and Wellness, Nashville, TN, United States

The overall goal of this work is to apply a custom CEST imaging approach to patients with, or at risk for, breast cancer treatment-related lymphedema (BCRL) to assess interstitial protein accumulation and lymphatic compromise. Elevated interstitial protein accumulation is a hallmark for lymphatic failure; results of this study support the ability of APT CEST to provide contrast consistent with lateralizing disease in patients with breast cancer treatment-related lymphedema, thereby motivating the further investigation of these and similar methods for evaluating lymphatic dysfunction.

Tonje H. Haukaas^1,2, Leslie R. Euceda¹, Guro F. Giskeødegård¹, Marit Krohn^2,3, Ellen Schlichting⁴, Rolf Kåresen^2,4, Sandra Nyberg^2,3, Kristine Kleivi Sahlberg^2,3, Anne-Lise Børresen-Dale^2,3, and Tone F. Bathen^1,2
1Department of Circulation and Medical Imaging, Faculty of Medicine, NTNU, Trondheim, Norway, ²K.G. Jebsen Center for Breast Cancer Research, Institute of Clinical Medicine, Faculty of Medicine, University of Oslo, Oslo, Norway, ³Department of Genetics, Institute for Cancer Research Oslo University Hospital, The Norwegian Radium Hospital, Oslo, Norway, ⁴Department of Surgery, Oslo University Hospital, Ullevål, Oslo, Norway

Here we combine gene and protein expression subtypes with three metabolic subclasses found from hierarchical clustering of HR MAS MR spectra: c1, c2 and c3. The relative levels of 16 metabolites were significantly different between at least two of the subclasses. RPPA-subtype distribution was significantly different between the metabolic subclasses. INMEX uncovered metabolic and transcriptomic differences in glycerophospholipid metabolism between c1 and c2. GSEA revealed changes in pathways related to collagens and extracellular matrix in c1 and c3 when compared to c2. Thus, the combination of different molecular levels provides insight into the heterogeneity of breast cancer.

Radka Stoyanova¹, Alan Pollack¹, Nicholas Erho², Charles Lynne³, Lucia Lam², Christine Buerki², Sakhi Abraham¹, Merce Jorda⁴, Olexandr Kryvenko⁴, Matthew Abramowitz¹, Elai Davicioni², and Adrian Ishkanian^1
1Radiation Oncology, University of Miami, Miami, Florida, United States, ²GenomeDx Biosciences, Vancouver, British Columbia, Canada, ³Urology, University of Miami, Miami, Florida, United States, ⁴Pathology, University of Miami, Miami, Florida, United States

Current clinicopathological factors explain just a fraction of the observed heterogeneity in prostate cancer patient outcome. In this study we address the question whether quantitative imaging data from multiparametric (MP)-MRI are associated with gene expression characteristics (radiogenomics) and how these in turn relate to Gleason score (GS) and intra-patient versus inter-patient genomic heterogeneity. Quantitative MP-MRI-targeted diagnostic biopsies can potentially improve risk classification by directing pathological and genomic analysis to highest risk index lesions. This is the first demonstration of a link between quantitative imaging features (radiomics) with genomic features in MRI-directed prostate biopsies.