Wo m e n ’s I m a g i n g • O r i g i n a l R e s e a r c h McDonald et al. 3-T DWI of Normal Breast
Downloaded from www.ajronline.org by East Carolina University on 06/09/14 from IP address 150.216.68.200. Copyright ARRS. For personal use only; all rights reserved
Women’s Imaging Original Research
Diffusion-Weighted MRI: Association Between Patient Characteristics and Apparent Diffusion Coefficients of Normal Breast Fibroglandular Tissue at 3 T Elizabeth S. McDonald1, 2 Jennifer G. Schopp1 Sue Peacock1 Wendy D. DeMartini 3 Habib Rahbar 1 Constance D. Lehman1 Savannah C. Partridge1 McDonald ES, Schopp JG, Peacock S, et al.
Keywords: 3-T MRI, apparent diffusion coefficient, breast MRI, diffusion-weighted imaging, normal breast DOI:10.2214/AJR.13.11159 Received April 30, 2013; accepted after revision August 17, 2013. C. D. Lehman has received honoraria from and serves as an advisory board member and scientific expert for GE Healthcare and Bayer Healthcare. Presented at the 2012 annual meeting of the ARRS, Vancouver, Canada. This study was funded by National Cancer Institute grant R01CA151326. 1 Department of Radiology, University of Washington School of Medicine, Seattle Cancer Care Alliance, 825 Eastlake Ave E, G3-200, Seattle, WA 98109-1023. Address correspondence to S. C. Partridge ([email protected]). 2
Present address: Department of Radiology, University of Pennsylvania, Philadelphia, PA.
3 Department of Radiology, University of Wisconsin School of Medicine and Public Health, Madison, WI.
WEB This is a web exclusive article. AJR 2014; 202:W496–W502 0361–803X/14/2025–W496 © American Roentgen Ray Society
W496
OBJECTIVE. The purpose of this study is to assess associations between patient characteristics and apparent diffusion coefficient (ADC) values of normal breast fibroglandular tissue on diffusion-weighted imaging (DWI) at 3 T. MATERIALS AND METHODS. The retrospective study included 103 women with negative bilateral findings on 3-T breast MRI examinations (BI-RADS category 1). DWI was acquired during clinical breast MRI scans using b = 0 and b = 800 s/mm2. Mean ADC of normal breast fibroglandular tissue was calculated for each breast using a semiautomated software tool in which parenchyma pixels were selected by interactive thresholding of the b = 0 s/ mm2 image to exclude fat. Intrasubject right- and left-breast ADC values were compared and averaged together to evaluate the association of mean breast ADC with age, mammographic breast density, and background parenchymal enhancement. RESULTS. Overall mean ± SD breast ADC was 1.62 ± 0.30 × 10 –3 mm2 /s. Intrasubject right- and left-breast ADC measurements were highly correlated (R2 = 0.89; p < 0.0001). Increased breast density was strongly associated with increased ADC (p ≤ 0.0001). Age and background parenchymal enhancement were not associated with ADC. CONCLUSION. Normal breast parenchymal ADC values increase with mammographic density but are independent of age and background parenchymal enhancement. Because breast malignancies have been shown to have low ADC values, DWI may be particularly valuable in women with dense breasts owing to greater contrast between lesion and normal tissue.
D
iffusion-weighted imaging (DWI) is a short (2–3 minutes) unenhanced MRI technique that has shown promise for the detection and characterization of breast cancer. DWI measures the 3D mobility of water in tissues and is sensitive to microstructural properties including cell density, cellular organization, and cell membrane integrity. Diffusion can be quantified by using the apparent diffusion coefficient (ADC) to increase the specificity of this technique [1]. Cellular density is increased in malignancies, which is reflected by lower ADC [2–6]. Within the past 15 years, there have been approximately 100 publications investigating the role of DWI in breast imaging. These studies have largely focused on using DWI to distinguish between benign and malignant lesions or monitor response of a known cancer to treatment and have shown potential for DWI to detect and characterize breast malignancies [7]. Despite a large amount of scientific research and published data, clinical use of
DWI is still not widespread. This is most likely because of the lack of standardization of DWI parameters and a uniform method of interpretation. The development of standardized interpretation criteria requires a solid understanding of the appearance and normative range of ADC values of the normal breast on diffusion imaging. However, only a small fraction of the DWI publications have specifically focused on characterizing the normal breast [8–13]. Breast ADC can vary widely between individuals and different baseline ADC values may affect the conspicuity of malignancy. Improved understanding of factors influencing ADC in normal breast tissue is needed to simplify and standardize interpretation of diffusion images. The purpose of this study was to measure ADC values in normal breast tissue at 3 T and assess whether they correlate with any or all of the following patient characteristics: age, breast density, and background parenchymal enhancement.
AJR:202, May 2014
Fig. 1—Example of semiautomated diffusionweighted imaging analysis. Mean apparent diffusion coefficient of normal breast fibroglandular tissue was calculated for each breast using semiautomated software tool in which fibroglandular tissue pixels were selected by interactive thresholding of b = 0 s/ mm2 image to exclude fat. On threshold image, ROI was defined for each breast to include entire breast parenchyma, excluding fat. A, Diffusion-weighted image (b = 0 s/mm2) at level of nipple used for analysis, before thresholding. B, Thresholded b = 0 s/mm2 image. Circular outline over right breast defines ROI for this breast.
A Materials and Methods Subjects This HIPAA-compliant study was approved by the institutional review board. Consecutive screening breast MRI examinations performed from February 2010 to March 2011 were retrospectively reviewed. Informed consent was waived owing to the retrospective nature of the study. DWI was performed as part of the standard clinical protocol over this time frame. Eligible subjects were those with either negative findings on a bilateral screening breast MRI (BI-RADS category 1) or negative or benign findings on a screening mammogram (BI-RADS 1 or 2) [14]. There were 130 consecutive women identified over the study time frame who met the inclusion criteria. DWI was not performed as part of the clinical MRI examination for 21 of 130 women (16%) owing to time constraints during imaging. Of the remaining 109 women, five who underwent DWI were excluded from the study because mammograms were not performed or not available for
B review, and one was excluded who had a congenital abnormality leading to markedly asymmetric breast size. Therefore, 103 women were included in the final study cohort (Table 1). The mean age ± SD of the subjects was 47 ± 11 years.
MR Image Acquisition In our high-risk screening MRI population, DWI was performed as part of the standard clinical breast MRI examination using a 3-T scanner (Achieva, Philips Healthcare) with a 16-channel breast coil. The clinical breast MRI protocol also included a T2-weighted fast spin-echo sequence, a T1-weighted non–fat-suppressed sequence, and a T1-weighted fat-suppressed dynamic contrastenhanced (DCE) MRI sequence. DWI was performed after DCE-MRI, which is considered to be the most essential part of the examination, in case a patient was not able to tolerate the entire examination owing to claustrophobia or other discomfort. DWI was performed using a diffusion-
weighted echo-planar imaging sequence with parallel imaging array spatial sensitivity encoding technique and the following parameters: reduction factor, 3; TR/TE, 5336/61; two averages; matrix, 240 × 240; FOV, 36 cm; slice thickness, 5 mm; and gap, 0. Diffusion gradients were applied in six directions with b = 0 and b = 800 s/mm 2.
Image Analysis The MRI examinations were interpreted prospectively by one of four fellowship-trained radiologists specializing in breast imaging. Each case was reviewed for suspicious features and characterized using the American College of Radiology BI-RADS breast MRI lexicon [15]. Additional features of each examination were recorded at the time of interpretation including assessment of background parenchymal enhancement and mammographic breast density. This information was entered into our institutional breast imaging database [16] and extracted for this study.
TABLE 1: Patient Characteristics
2.5
Group Left-Breast ADC (x 10 -3mm2/s)
Downloaded from www.ajronline.org by East Carolina University on 06/09/14 from IP address 150.216.68.200. Copyright ARRS. For personal use only; all rights reserved
3-T DWI of Normal Breast
No. (%)
Breast density
2
Fatty
1.5
3 (2.9)
Scattered heterogeneous
26 (25.2)
Heterogeneously dense
52 (50.5)
Extremely dense
22 (21.4)
Background parenchymal enhancement
1 y = 0.93277x + 0.00162 R 2 = 0.8914
Minimal
0.5
0 0
0.5
1 1.5 Right-Breast ADC (x 10 -3mm2/s)
2
Fig. 2—Graph shows intrasubject comparison of left- and right-breast apparent diffusion coefficient (ADC) measures in 103 subjects with negative findings on breast MRI examinations (BI-RADS category 1).
2.5
39 (37.9)
Mild
32 (31.1)
Moderate
17 (16.5)
Marked
15 (14.6)
Age (y) < 50
56 (54.4)
≥ 50
47 (45.6)
AJR:202, May 2014 W497
Downloaded from www.ajronline.org by East Carolina University on 06/09/14 from IP address 150.216.68.200. Copyright ARRS. For personal use only; all rights reserved
McDonald et al.
A
B
C Fig. 3—52-year-old woman with known BRCA2 mutation who underwent breast MRI for high-risk screening. A, Mediolateral oblique view on mammography shows scattered fibroglandular densities in normal breast. B, Contrast-enhanced subtraction maximum intensity projection shows minimal background parenchymal enhancement. C–E, Axial MR images are shown at level of nipple: contrast-enhanced T1-weighted image (C); diffusionweighted (b = 0 s/mm2) T2-weighted image (D); and apparent diffusion coefficient (ADC) map (mean ADC, 1.23 × 10 –3 mm2 /s) (E).
D
E
The diffusion-weighted MR images were retrospectively analyzed by a radiologist who was blinded to patient characteristics in the medical record (age, mammographic breast density, background parenchymal enhancement classification) but was not blinded to the DCE and other MR images acquired during the same examination. Diffusion maps were created using Java-based image processing software incorporating ImageJ (National Institutes of Health, public domain). ADC maps were calculated from the spatially registered diffusion-weighted MR images using the following equation [1]: ADC = (–1/b)ln(S DWI /S 0 ), where S DWI is the combined DWI (geometric average of individual b = 800 s/mm 2 diffusion-weighted MR images) and S 0 is the T2-weighted b = 0 s/mm 2 reference image [17]. Mean ADC of normal breast fibroglandular tissue was calculated for each breast using a semiautomated software tool in which parenchyma
pixels were selected by interactive thresholding of the b = 0 image to exclude fat. A region of interest (ROI) was defined on the threshold image at the level of the nipple for each breast to include the entire breast fibroglandular tissue, excluding the fat (Fig. 1). In the case of a cyst identified in the slice of interest, the ROI was drawn to avoid the cyst or an alternative nearby slice outside the cyst was chosen. The breast parenchymal ADC was calculated as the mean of the voxels in the ROI.
parenchymal enhancement on DCE-MRI was categorized as minimal, mild, moderate, or marked (in increasing order). Each breast was given a background parenchymal enhancement assessment, and the highest assessment was used. Of the 103 women, 102 were given the same background parenchymal enhancement level for both breasts. The instances of differing bilateral density and background parenchymal enhancement assessments were in two different women.
Patient Characteristics
Statistical Analyses
Patient characteristics were recorded from the medical record. In the radiology reports, mammographic density was categorized at the time of interpretation as either fatty, scattered fibroglandular tissue, heterogeneously dense, or extremely dense (in increasing order). Each breast was given a density assessment, and the highest assessment was used. Of the 103 women, 102 were given the same density level for both breasts. Background
Intrasubject right- and left-breast ADC values were compared by Pearson correlation coefficient and were then averaged together for comparison of mean breast ADC with subject age (
Downloaded from www.ajronline.org by East Carolina University on 06/09/14 from IP address 150.216.68.200. Copyright ARRS. For personal use only; all rights reserved
Women’s Imaging Original Research
Diffusion-Weighted MRI: Association Between Patient Characteristics and Apparent Diffusion Coefficients of Normal Breast Fibroglandular Tissue at 3 T Elizabeth S. McDonald1, 2 Jennifer G. Schopp1 Sue Peacock1 Wendy D. DeMartini 3 Habib Rahbar 1 Constance D. Lehman1 Savannah C. Partridge1 McDonald ES, Schopp JG, Peacock S, et al.
Keywords: 3-T MRI, apparent diffusion coefficient, breast MRI, diffusion-weighted imaging, normal breast DOI:10.2214/AJR.13.11159 Received April 30, 2013; accepted after revision August 17, 2013. C. D. Lehman has received honoraria from and serves as an advisory board member and scientific expert for GE Healthcare and Bayer Healthcare. Presented at the 2012 annual meeting of the ARRS, Vancouver, Canada. This study was funded by National Cancer Institute grant R01CA151326. 1 Department of Radiology, University of Washington School of Medicine, Seattle Cancer Care Alliance, 825 Eastlake Ave E, G3-200, Seattle, WA 98109-1023. Address correspondence to S. C. Partridge ([email protected]). 2
Present address: Department of Radiology, University of Pennsylvania, Philadelphia, PA.
3 Department of Radiology, University of Wisconsin School of Medicine and Public Health, Madison, WI.
WEB This is a web exclusive article. AJR 2014; 202:W496–W502 0361–803X/14/2025–W496 © American Roentgen Ray Society
W496
OBJECTIVE. The purpose of this study is to assess associations between patient characteristics and apparent diffusion coefficient (ADC) values of normal breast fibroglandular tissue on diffusion-weighted imaging (DWI) at 3 T. MATERIALS AND METHODS. The retrospective study included 103 women with negative bilateral findings on 3-T breast MRI examinations (BI-RADS category 1). DWI was acquired during clinical breast MRI scans using b = 0 and b = 800 s/mm2. Mean ADC of normal breast fibroglandular tissue was calculated for each breast using a semiautomated software tool in which parenchyma pixels were selected by interactive thresholding of the b = 0 s/ mm2 image to exclude fat. Intrasubject right- and left-breast ADC values were compared and averaged together to evaluate the association of mean breast ADC with age, mammographic breast density, and background parenchymal enhancement. RESULTS. Overall mean ± SD breast ADC was 1.62 ± 0.30 × 10 –3 mm2 /s. Intrasubject right- and left-breast ADC measurements were highly correlated (R2 = 0.89; p < 0.0001). Increased breast density was strongly associated with increased ADC (p ≤ 0.0001). Age and background parenchymal enhancement were not associated with ADC. CONCLUSION. Normal breast parenchymal ADC values increase with mammographic density but are independent of age and background parenchymal enhancement. Because breast malignancies have been shown to have low ADC values, DWI may be particularly valuable in women with dense breasts owing to greater contrast between lesion and normal tissue.
D
iffusion-weighted imaging (DWI) is a short (2–3 minutes) unenhanced MRI technique that has shown promise for the detection and characterization of breast cancer. DWI measures the 3D mobility of water in tissues and is sensitive to microstructural properties including cell density, cellular organization, and cell membrane integrity. Diffusion can be quantified by using the apparent diffusion coefficient (ADC) to increase the specificity of this technique [1]. Cellular density is increased in malignancies, which is reflected by lower ADC [2–6]. Within the past 15 years, there have been approximately 100 publications investigating the role of DWI in breast imaging. These studies have largely focused on using DWI to distinguish between benign and malignant lesions or monitor response of a known cancer to treatment and have shown potential for DWI to detect and characterize breast malignancies [7]. Despite a large amount of scientific research and published data, clinical use of
DWI is still not widespread. This is most likely because of the lack of standardization of DWI parameters and a uniform method of interpretation. The development of standardized interpretation criteria requires a solid understanding of the appearance and normative range of ADC values of the normal breast on diffusion imaging. However, only a small fraction of the DWI publications have specifically focused on characterizing the normal breast [8–13]. Breast ADC can vary widely between individuals and different baseline ADC values may affect the conspicuity of malignancy. Improved understanding of factors influencing ADC in normal breast tissue is needed to simplify and standardize interpretation of diffusion images. The purpose of this study was to measure ADC values in normal breast tissue at 3 T and assess whether they correlate with any or all of the following patient characteristics: age, breast density, and background parenchymal enhancement.
AJR:202, May 2014
Fig. 1—Example of semiautomated diffusionweighted imaging analysis. Mean apparent diffusion coefficient of normal breast fibroglandular tissue was calculated for each breast using semiautomated software tool in which fibroglandular tissue pixels were selected by interactive thresholding of b = 0 s/ mm2 image to exclude fat. On threshold image, ROI was defined for each breast to include entire breast parenchyma, excluding fat. A, Diffusion-weighted image (b = 0 s/mm2) at level of nipple used for analysis, before thresholding. B, Thresholded b = 0 s/mm2 image. Circular outline over right breast defines ROI for this breast.
A Materials and Methods Subjects This HIPAA-compliant study was approved by the institutional review board. Consecutive screening breast MRI examinations performed from February 2010 to March 2011 were retrospectively reviewed. Informed consent was waived owing to the retrospective nature of the study. DWI was performed as part of the standard clinical protocol over this time frame. Eligible subjects were those with either negative findings on a bilateral screening breast MRI (BI-RADS category 1) or negative or benign findings on a screening mammogram (BI-RADS 1 or 2) [14]. There were 130 consecutive women identified over the study time frame who met the inclusion criteria. DWI was not performed as part of the clinical MRI examination for 21 of 130 women (16%) owing to time constraints during imaging. Of the remaining 109 women, five who underwent DWI were excluded from the study because mammograms were not performed or not available for
B review, and one was excluded who had a congenital abnormality leading to markedly asymmetric breast size. Therefore, 103 women were included in the final study cohort (Table 1). The mean age ± SD of the subjects was 47 ± 11 years.
MR Image Acquisition In our high-risk screening MRI population, DWI was performed as part of the standard clinical breast MRI examination using a 3-T scanner (Achieva, Philips Healthcare) with a 16-channel breast coil. The clinical breast MRI protocol also included a T2-weighted fast spin-echo sequence, a T1-weighted non–fat-suppressed sequence, and a T1-weighted fat-suppressed dynamic contrastenhanced (DCE) MRI sequence. DWI was performed after DCE-MRI, which is considered to be the most essential part of the examination, in case a patient was not able to tolerate the entire examination owing to claustrophobia or other discomfort. DWI was performed using a diffusion-
weighted echo-planar imaging sequence with parallel imaging array spatial sensitivity encoding technique and the following parameters: reduction factor, 3; TR/TE, 5336/61; two averages; matrix, 240 × 240; FOV, 36 cm; slice thickness, 5 mm; and gap, 0. Diffusion gradients were applied in six directions with b = 0 and b = 800 s/mm 2.
Image Analysis The MRI examinations were interpreted prospectively by one of four fellowship-trained radiologists specializing in breast imaging. Each case was reviewed for suspicious features and characterized using the American College of Radiology BI-RADS breast MRI lexicon [15]. Additional features of each examination were recorded at the time of interpretation including assessment of background parenchymal enhancement and mammographic breast density. This information was entered into our institutional breast imaging database [16] and extracted for this study.
TABLE 1: Patient Characteristics
2.5
Group Left-Breast ADC (x 10 -3mm2/s)
Downloaded from www.ajronline.org by East Carolina University on 06/09/14 from IP address 150.216.68.200. Copyright ARRS. For personal use only; all rights reserved
3-T DWI of Normal Breast
No. (%)
Breast density
2
Fatty
1.5
3 (2.9)
Scattered heterogeneous
26 (25.2)
Heterogeneously dense
52 (50.5)
Extremely dense
22 (21.4)
Background parenchymal enhancement
1 y = 0.93277x + 0.00162 R 2 = 0.8914
Minimal
0.5
0 0
0.5
1 1.5 Right-Breast ADC (x 10 -3mm2/s)
2
Fig. 2—Graph shows intrasubject comparison of left- and right-breast apparent diffusion coefficient (ADC) measures in 103 subjects with negative findings on breast MRI examinations (BI-RADS category 1).
2.5
39 (37.9)
Mild
32 (31.1)
Moderate
17 (16.5)
Marked
15 (14.6)
Age (y) < 50
56 (54.4)
≥ 50
47 (45.6)
AJR:202, May 2014 W497
Downloaded from www.ajronline.org by East Carolina University on 06/09/14 from IP address 150.216.68.200. Copyright ARRS. For personal use only; all rights reserved
McDonald et al.
A
B
C Fig. 3—52-year-old woman with known BRCA2 mutation who underwent breast MRI for high-risk screening. A, Mediolateral oblique view on mammography shows scattered fibroglandular densities in normal breast. B, Contrast-enhanced subtraction maximum intensity projection shows minimal background parenchymal enhancement. C–E, Axial MR images are shown at level of nipple: contrast-enhanced T1-weighted image (C); diffusionweighted (b = 0 s/mm2) T2-weighted image (D); and apparent diffusion coefficient (ADC) map (mean ADC, 1.23 × 10 –3 mm2 /s) (E).
D
E
The diffusion-weighted MR images were retrospectively analyzed by a radiologist who was blinded to patient characteristics in the medical record (age, mammographic breast density, background parenchymal enhancement classification) but was not blinded to the DCE and other MR images acquired during the same examination. Diffusion maps were created using Java-based image processing software incorporating ImageJ (National Institutes of Health, public domain). ADC maps were calculated from the spatially registered diffusion-weighted MR images using the following equation [1]: ADC = (–1/b)ln(S DWI /S 0 ), where S DWI is the combined DWI (geometric average of individual b = 800 s/mm 2 diffusion-weighted MR images) and S 0 is the T2-weighted b = 0 s/mm 2 reference image [17]. Mean ADC of normal breast fibroglandular tissue was calculated for each breast using a semiautomated software tool in which parenchyma
pixels were selected by interactive thresholding of the b = 0 image to exclude fat. A region of interest (ROI) was defined on the threshold image at the level of the nipple for each breast to include the entire breast fibroglandular tissue, excluding the fat (Fig. 1). In the case of a cyst identified in the slice of interest, the ROI was drawn to avoid the cyst or an alternative nearby slice outside the cyst was chosen. The breast parenchymal ADC was calculated as the mean of the voxels in the ROI.
parenchymal enhancement on DCE-MRI was categorized as minimal, mild, moderate, or marked (in increasing order). Each breast was given a background parenchymal enhancement assessment, and the highest assessment was used. Of the 103 women, 102 were given the same background parenchymal enhancement level for both breasts. The instances of differing bilateral density and background parenchymal enhancement assessments were in two different women.
Patient Characteristics
Statistical Analyses
Patient characteristics were recorded from the medical record. In the radiology reports, mammographic density was categorized at the time of interpretation as either fatty, scattered fibroglandular tissue, heterogeneously dense, or extremely dense (in increasing order). Each breast was given a density assessment, and the highest assessment was used. Of the 103 women, 102 were given the same density level for both breasts. Background
Intrasubject right- and left-breast ADC values were compared by Pearson correlation coefficient and were then averaged together for comparison of mean breast ADC with subject age (
