Apparent Diffusion Coefficient Using Diffusion Weighted MRI and Biochemical Correlates inĀ  Human Breast Cancer
Kavya Umachandran1, Andres Saucedo1, Stephanie Lee-Felker1, Melissa M Joines1, Manoj Kumar Sarma1, Sumit Kumar1, Maggie DiNome2, Nanette DeBruhl1, and Michael Albert Thomas1

1Radiological Sciences, UCLA Geffen School of Medicine, Los Angeles, CA, United States, 2Surgery, UCLA Geffen School of Medicine, Los Angeles, CA, United States


Multiparametric MRI has been investigated in breast and prostate cancer, and other tumors. We evaluated diffusion weighted imaging in a pilot cohort of 20 malignant and 11 benign breast cancer patients, and 7 healthy women. MR spectra were recorded using an accelerated version of five dimensional echo-planar correlated spectroscopic imaging (5D EP-COSI). Significant decline in ADC values of malignant breast cancer compared to benign and healthy women. There was a negative correlation of choline with ADC values in malignant cancer patients. Our findings suggest that choline and lipids can be reliable biomarkers in addition to widely used ADC.


During the last three decades, multiparametric MR imaging has been widely investigated in breast cancer1-3. Diffusion-weighted imaging (DWI) is one of MRI modalities to differentiate malignant from benign tumors using the DWI metric, namely apparent diffusion coefficient (ADC)4-8. The DWI has been widely investigated in breast cancer before and after neoadjuvant chemotherapy. Non-invasive biochemical characterization of metabolites such as choline and lipids has been accomplished in breast cancer using water suppressed MR Spectroscopy (MRS) and changes of choline and lipids in breast cancer has been investigated also using MRS9-12. In this project, we investigated DWI and multi-dimensional MR Spectroscopic Imaging (MRSI) in malignant and benign breast cancer patients and also in healthy women.

Materials and Methods

We recruited 20 malignant (mean age of 53.2 years) and 11 benign (mean age of 36.7 years) breast cancer patients and 7 healthy women (mean age of 38.7 years). A 3T MRI scanner equipped with a 15 channel breast phased-array coil was used for this investigation. The DWI acquisition protocol included the following: 2D spin-echo echo-planar imaging (EPI) sequence (TR/TE of 3800/93ms; data matrix, 192 × 192; signal average, 3; slice thickness, 3 mm; distance factor, 20%) in the axial plane. Sensitizing diffusion gradients in three orthogonal directions with b values of 50 and 800 s/mm were applied. The ADC maps were created automatically by the system from the trace-weighted images with b values of 50 and 800 s/mm. The 5D EP-COSI acquisition parameters were as follows: The 5D EP-COSI sequence was home-developed 13 and 8X acceleration was used along ky, kz and t1 dimensions. TR/TE=1.5s/30ms; Field of view (FOV) along the 3 spatial dimensions were 160mmx160mmx120mm with 16-32 along kx, 16 along ky and 8 along kz; A voxel size of approximately 60x60x50mm3 was localized by three slice-selective RF pulses (900-1800-900). 2D COSY spectra were extracted from a voxel resolution of 1.5ml. The DWI data was analyzed using the manufacturer supplied software and ADC values were quantified for all groups of subjects. A home-developed Matlab code was used to reconstruct the undersampled 5D EP-COSI data using group sparsity14.


Shown in Fig.1 are ADC maps derived from the DWI data acquired in a 45 y.o. malignant and 35 y.o. benign breast cancer patients and a 45 y.o. healthy subject. Fig.2 (A) shows mean±SD of ADC values in 15 malignant and 8 benign breast cancer patients, and 7 healthy women. The remaining data in 5 malignant and 3 benign breast cancer patients were excluded due to inferior DWI quality. Fig.2 (B) shows the distribution of ADC values in the three groups. We observed significant reduction of ADC in malignant patients compared to benign (1.1 vs 1.66) and healthy controls (1.1 vs 1.95) with p<0.001. Figure 3 shows ADC values versus A) tumor grades and B) MRI tumor sizes in malignant patients. Shown in Fig.4 are multi-voxel COSY spectra recorded in a 40 y.o. malignant breast cancer patient (grade 3, size of 63mm, KI-67 of 20-30%) with the axial T1-weighted MRI showing the MRSI grids. Table 1 shows excellent correlation between the two metrics, namely ADC and choline.


Both DWI and 5D EP-COSI data were acquired using the echo-planar read-outs which are sensitive to B0 inhomogeneity leading to artifacts. Our results are in agreement with previous studies on invasive breast cancer15. The ADC values of malignant patients were significantly lower than benign and healthy subjects. 2D COSY spectra were extractable from a voxel size of 1.5ml from the 5D EP-COSI data. Strong correlation of choline measured through 5D EP-COSI shows the reliability of spectroscopic data. 5D EP-COSI has the advantage of better coverage of breast tumors compared to other known spectroscopic sequences enhancing the accuracy. Increased ADC and decreased choline in malignant group reflect the increased cellularity16 of the malignant lesions without the need for the administration of contrast medium. DWI and 5D EP-COSI shows strong potential as an adjunct technique to reduce breast biopsies, and could increase the overall specificity of DCE-MRI.


The pilot findings of this study using DWI and 5D WP-COSI are encouraging and provide positive evidence to support 5D EP-COSI and DWI as useful adjuncts to standard breast MRI protocols in assisting with the diagnosis of breast cancer. However, further validation using a larger cohort of breast cancer patients is needed.


This research was supported by a CDMRP Breakthrough Step I award # W81XWH-16-1-0524.


1) Zhang M, Horvat JV, Bernard-Devila B, et al.. Multiparametric MRI model with dynamic contrast-enhanced and diffusion weighted imaging enables breast cancer diagnosis with high accuracy. J Magn Reson Imaging 2018, Epub Ahead Oct.30.

2) Penzeri MM, Losio C, Della Corte A, et al. Prediction of Chemoresistance in Women Undergoing Neo-Adjuvant Chemotherapy for Locally Advanced Breast Cancer: Volumetric Analysis of First-Order Textural Features Extracted from Multiparametric MRI.. Contrast Media Mol Imaging. 2018 May 3;2018:8329041. eCollection 2018.

3) Sharma U, Agarwal K, Sah RG,,, et al. Can Multi-Parametric MR Based Approach Improve the Predictive Value of Pathological and Clinical Therapeutic Response in Breast Cancer Patients?. Front Oncol. 2018 Aug 15;8:319.. eCollection 2018. Epub Ahead.

4) Pinker K, Moy L, Sutton EJ,, et al. Diffusion-Weighted Imaging With Apparent Diffusion Coefficient Mapping for Breast Cancer Detection as a Stand-Alone Parameter: Comparison With Dynamic Contrast-Enhanced and Multiparametric Magnetic Resonance Imaging.. Invest Radiol. 2018 Oct;53(10):587-595.

5) Liao XL, Wei JB, Li YQ, et al. Functional Magnetic Resonance Imaging in the Diagnosis of Locally Recurrent Prostate Cancer: Are All Pulse Sequences Helpful?. Korean J Radiol. 2018 Nov-Dec;19(6):1110-1118. Epub 2018 Oct 18..

6) Newitt DC, Zhang Z, Gibbs JE,, et al. Test-retest repeatability and reproducibility of ADC measures by breast DWI: Results from the ACRIN 6698 trial.. J Magn Reson Imaging. 2018 Oct 22, Epub AHead.

7) deSouza NM. Diffusion-weighted MRI in Multicenter Trials of Breast Cancer: A Useful Measure of Tumor Response?. Radiology. 2018 Sep 4:181717, Epub Ahead.

8) Amornsiripanitch N, Nguyen VT, Rahbar H, et al. Diffusion-weighted MRI characteristics associated with prognostic pathological factors and recurrence risk in invasive ER+/HER2- breast cancers.. J Magn Reson Imaging. 2018 Jul;48(1):226-236. Epub 2017 Nov 27.

9) Mirka H, Tupy R, Narsanska A, Hes O, Ferda J. Pre-surgical Multiparametric Assessment of Breast Lesions Using 3-Tesla Magnetic Resonance. Anticancer Res. 2017 Dec;37(12):6965-6970.

10) Jagannathan NR, Sharma U.. Characterization of malignant breast tissue of breast cancer patients and the normal breast tissue of healthy lactating women volunteers using diffusion MRI and in vivo 1H MR spectroscopy.. J Magn Reson Imaging. 2015 Jan;41(1):169-74. Epub 2013 Nov 22.

11) Bolan PJ, Kim E, Herman BA,, et al. MR spectroscopy of breast cancer for assessing early treatment response: Results from the ACRIN 6657 MRS trial.. J Magn Reson Imaging. 2017 Jul;46(1):290-302. Epub 2016 Dec 16.

12) Nelson MT, Everson LI, Garwood M, Emory T, Bolan PJ. MR Spectroscopy in the diagnosis and treatment of breast cancer.. Semin Breast Dis. 2008 Jun 1;11(2):100-105.

13 )Wilson NE, Burns BL, Iqbal Z, Thomas MA. Correlated spectroscopic imaging of calf muscle in three spatial dimensions using group sparse reconstruction of undersampled single and multichannel data.. Magn Reson Med2015;74:1199–208.

14) Burns BL, Wilson NE, Thomas MA. Group Sparse reconstruction of multidimensional spectroscopic imaging in human brain in vivo. Algorithms 2014;7:276–294.

15) Boulogianni G, Chryssogonidis , Drevelegas A. Diffusion weighted MRI and spectroscopy in invasive carcinoma of the breast at 3Tesla. Correlation with dynamic contrast enhancement and pathologic findings. Hippokratia. 2016;20(3):192-197.

16) Cen D, Xu L. Differential diagnosis between malignant and benign breast lesions using single-voxel proton MRS: a meta-analysis. J Cancer Res Clin Oncol. 2014;140:993–100


Figure.1: ADC maps of the malignant and benign breast cancer, and healthy subjects. The regions of interest chosen for the analysis are also marked with a circle in malignant and benign breast cancer subjects.

Figure.2: A) ADC values quantified from DWI of malignant and breast cancer patients and healthy women. B) Scatter plot showing the individual values of ADCs.

Figure.3: Plots showing ADC values versus A) tumor grades and B) tumor sizes.

Figure 4. A) Axial MRI slice showing the field of view (FOV) with a yellow box and grids and the white box representing the volume of interest (VOI) localized by the 5D EP-COSI sequence(900ss-1800ss-900ss); B) Extracted multi-voxel COSY spectra from one of the slices; C) an extracted COSY spectrum marked in red color in B.

Figure 5. Correlation between the ADC values and Choline

Proc. Intl. Soc. Mag. Reson. Med. 27 (2019)