David C Newitt¹, Savannah C Partridge², Belinda Chang¹, Bonnie N Joe¹, and Nola Hylton^1
1Radiology and Biomedical Imaging, University of California, San Francisco, CA, United States, ²Radiology, University of Washington, Seattle, WA

Tumor volumes derived from Breast DCE-MRI are effective for measuring treatment response, but volumetric measurement typically requires setting a percent enhancement threshold for segmenting malignant tissue regions. Optimization of this threshold for outcomes prediction may depend on many factors including imaging parameters, biomarker definition, and outcome chosen. In a neoadjuvant chemotherapy breast cancer cohort with extended recurrence free survival (RFS) data we performed systematic optimization of the threshold based on risk ratios for RFS demonstrating significant but varying risk predictions for MRI volumetric measurements as a function of PE threshold.

Sunitha Thakur^1,2, Sharp Malak², Sanjay Annarao¹, Dilip Giri³, Jason Koutcher^1,2, and Elizabeth Morris^2
1Medical Physics, Memorial Sloan Kettering Cancer Center, New York, NY, United States, ²Radiology, Memorial Sloan Kettering Cancer Center, New York, NY, United States, ³Pathology, Memorial Sloan Kettering Cancer Center

DWI provides information about early changes in morphology and physiology of tissues by measuring apparent diffusion coefficient (ADC) of water molecules. Here we present the clinical usefulness of DWI at 3T-field strength for detecting malignancy. We also studied if there is any correlation exists between ADC values and traditional markers such as histology, molecular markers such as ER, PR and HER-2, and lymph node status. Results showed that ADC values of malignant lesions were significantly lower than benign lesions. There is also significant correlation between ADC and histology grade or lymph node status. Histopathology was used as the reference standard.

Habib Rahbar^1,2, Savannah Partridge^1,2, Wendy DeMartini^1,2, Franklin Liu^1,2, Robert Gutierrez^1,2, Kimberly Allison³, Sue Peacock^1,2, and Constance Lehman^1,2
1Radiology, University of Washington, Seattle, WA, United States, ²Radiology, Seattle Cancer Care Alliance, Seattle, WA, United States, ³Pathology, University of Washington, Seattle, WA

DCIS is a heterogeneous malignancy with variable invasive potential. There are currently no imaging models to assess in vivo DCIS grade. We investigated the feasibility to assess DCIS grade in 55 consecutive pure DCIS lesions using both univariate and multivariate logistic regression modeling based on DCE and DWI MRI characteristics. We found that DCIS grade could be accurately differentiated by both DWI and DCE features and that a model incorporating maximum lesion size and DWI CNR best discriminated DCIS by grade. Our findings suggest that MRI has the capability to contribute to risk stratification of DCIS.

Mami Iima¹, Denis Le Bihan^2,3, Tomohisa Okada¹, Koji Fujimoto¹, Shotaro Kanao¹, Shiro Tanaka⁴, and Kaori Togashi^1
1Dept. of Diagnostic Radiology, Kyoto University Graduate School of Medicine, Kyoto, Japan, ²Human Brain Research Center, Kyoto University Graduate School of Medicine, Kyoto, Japan, ³Neurospin, CEA-Saclay Center, Gif-sur-Yvette Cedex, France, ⁴The Translational Research Center, Kyoto University Hospital, Kyoto, Japan

This study included 18 patients who underwent breast MRI examinations at 1.5T. The MRI protocol contained diffusion-weighted images with b values of 0 and 1,000 s/mm2. The ADC value of high-grade DCIS was significantly lower than that of low-grade DCIS (p =0.01), and there was a significant linear positive trend between ADC and lesion stage (p <0.01). Above ADC threshold of 1.3 x 10-3 mm²/s for the diagnosis of high-grade DCIS, all DCIS lesions were identified as low-grade DCIS with a 100% specificity. Quantitative DWI could be used to identify patients with low-stage DCIS with very high specificity.

Paul E. Sijens^1,2, Monique D. Dorrius¹, Ruud M. Pijnappel¹, Martine C. Jansen-van der Weide¹, Peter Kappert¹, and Matthijs Oudkerk^1,2
1UMCG, Groningen, Netherlands, ²CMI, Groningen, Netherlands

Most of the studies published to date use single-voxel MRS methodology suffering from limitations inherent to the method used. The remainder are multivoxel studies using qualitative or semiquantitative measurements for the detection of Cho (i.e. detectability or Cho signal-to-noise ratio. Here a recently published multivoxel MRS method based on quantitative measurement was used. The significant difference between the Cho concentrations in benign and malignant breast lesions, strongly indicates that the lesion Cho concentration in mM offers a diagnostically meaningful test, i.e. a cut-off point of 1.5 mM. Quantitative multivoxel MRS may thus be applied to reliably exclude benign breast lesions such as fibroadenomas from biopsy or other invasive procedures.

Chenguang Zhao¹, Patrick Bolan², Melanie Royce³, Lavneeth Lakkadi², Sang-Joon Lee⁴, Steve Eberhard⁵, and Stefan Posse^1,6
1Neurology, University of New Mexico, Albuquerque, NM, United States, ²CMRR, University of Minnesota, Minneapolis, MN, United States, ³Medical Oncology, University of New Mexico, Albuquerque, NM, United States, ⁴Internal Medicine, University of New Mexico, Albuquerque, NM, United States, ⁵Radiology, University of New Mexico, Albuquerque, NM, United States, ⁶Electrical and Computer Engineering, University of New Mexico, Albuquerque, NM, United States

This study investigates the feasibility of using high-speed Proton-Echo-Planar-Spectroscopic-Imaging (PEPSI) to map a key biomarker, total Choline (tCho) in breast cancer. In a patient with locally advanced infiltrating ductal carcinoma, the tCho concentration measured in 9 voxels was 14.6±5.4 mmol/kg using PEPSI vs. 12.7±8.0 mmol/kg using PRESS CSI. In a group of healthy controls, the slice averaged tCho concentration measured using PEPSI was 0.48±0.2 mmol/kg. This was consistent with measurement using PRESS CSI (0.56 ±0.44mmol/kg) and PRESS single voxel Spectroscopy (0.69 ± 0.6 mmol/kg).

Samata M Kakkad¹, Jiangyang Zhang¹, Lu Jiang¹, Zaver M Bhujwalla¹, and Kristine Glunde^1
1JHU ICMIC Program, Russell H. Morgan Department of Radiology and Radiological Science, Johns Hopkins University School of Medicine, Baltimore, MD, United States

Collagen I (Col1) fiber density is increased in breast cancers. Increased mammary Col1 fiber density can cause tumor initiation, progression, and metastasis. Our goal is to determine the influence of the Col1 fiber architecture in human beast cancer on molecular diffusion. We performed ultra-high-resolution diffusion tensor imaging (DTI) of a human breast cancer specimen, followed by second harmonic generation microscopy detection of intrinsic signal from Col1 fibers. High Col1 fiber density positively correlated with the apparent diffusion coefficient, indicating that Col1 fibers can enhance water diffusion in breast tumors, and diffusion anisotropy may reflect the organization of Col1 fibers.

Edna Furman-Haran¹, Erez Eyal¹, Myra Shapiro-Feinberg², Dov Grobgeld¹, talia Golan³, Yaacov Itzchak³, Raphael Catane³, Moshe Papa³, and Hadassa Degani^1
1The Weizmann Institute of Science, Rehovot, Israel, ²Meir Medical Center, Israel, ³Sheba Medical Center, Israel

Diffusion tensor imaging (DTI) was applied to the breast in order to evaluate it's feasibility for tracking the ductal tree in vivo and for identifying the distinct tensor parameters for delineating and diagnosing breast cancer. Pixel by pixel analysis of high spatial resolution DTI of the two breasts yielded parametric maps of the vector maps of the main diffusion direction, diffusion coefficients in the three orthogonal directions, and diffusion anisotropy. Statistical analysis demonstrated the high potential of this approach for cancer detection and diagnosis with 96% sensitivity and specificity for cancer detection, and 92% sensitivity and specificity for cancer diagnosis.

Siwa Chan¹, Jeon-Hor Chen^2,3, Jia-Pei Wu³, Fu-Ju Lei³, Muqing Lin², Orhan Nalcioglu², and Min-Ying L Su^2
1Department of Radiology, Taichung Veterans General Hospital, Taichung, Taiwan, ²Center for Functional Onco-Imaging and Department of Radiological Science, University of California Irvine, Irvine, California, United States, ³Department of Radiology, China Medical University Hospital, Taichung, Taiwan

The fluctuation of breast density measurements was investigated by comparing the breast volume, fibroglandular tissue volume and the percent density measured from 4 MRI’s performed weekly during a menstrual cycle. Thirty subjects, 25 pre-menopausal and 5 post-menopausal, were studied. The coefficient of variation for each parameter over the 4 MRI’s was calculated to assess the measurement fluctuation. Overall, the mean CV was in the range of 4% to 8%, which was approximately twice that of operator uncertainly. As expected, a high fluctuation of > 10% was seen in 5 pre-menopausal women, but in none of the post-menopausal women.

Laura Johnson¹, Geoff Charles-Edwards², Jyoti Parikh³, Margaret Hall-Craggs⁴, Tobias Schaeffter⁵, and Michael Douek^1
1Research Oncology, Kings College London, London, England, United Kingdom, ²Medical physics, Guy's and St Thomas' NHS Foundation Trust, ³Radiology, Guys and St Thomas' NHS Foundation Trust, ⁴Imaging and Medical Physics and Bio−Engineering, University College London, ⁵Imaging sciences, Kings College London

Magnetic nanoparticles administered subcutaneously can differentiate normal from abnormal lymph nodes when comparing T2 values in preclinical studies. We demonstrate the identification and characterization on MRI of the sentinel lymph node (SLN) in breast cancer following subcutaneous injection of iron-oxide nanoparticles into the affected breast. Normal (uninvolved) SLNs undergo a significant decrease in the mean T2 value post injection of Endorem. This decrease is more pronounced over time when comparing early post-injection scans (mean delay 12 minutes) with delayed scans (mean delay 119 minutes). Further work must be done to evaluate how this differs in metastatic SLNs.