Original Research  n  Breast

Imaging

Note: This copy is for your personal non-commercial use only. To order presentation-ready copies for distribution to your colleagues or clients, contact us at www.rsna.org/rsnarights.

Diffusion-Tensor MR Imaging of the Breast: Hormonal Regulation1

Noam Nissan, MD Edna Furman-Haran, PhD Myra Shapiro-Feinberg, MD Dov Grobgeld, MSc Hadassa Degani, PhD

1

 From the Department of Biological Regulation (N.N., D.G., H.D.) and Unit of Biological Services (E.F.H.), Weizmann Institute of Science, PO Box 26, Rehovot 76100, Israel; and Department of Radiology, Meir Medical Center, Kfar Saba, Israel (M.S.F.). Received September 3, 2013; revision requested October 30; revision received November 7; accepted November 26; final version accepted December 13. Address correspondence to N.N. (e-mail: Noamniss@ gmail.com).

Purpose:

To investigate the parameters obtained with magnetic resonance (MR) diffusion-tensor imaging (DTI) of the breast throughout the menstrual cycle phases, during lactation, and after menopause, with and without hormone replacement therapy (HRT).

Materials and Methods:

All protocols were approved by the internal review board, and signed informed consent was obtained from all participants. Forty-five healthy volunteers underwent imaging by using T2-weighted and DTI MR sequences at 3 T. Premenopausal volunteers (n = 16) underwent imaging weekly, four times during one menstrual cycle. Postmenopausal volunteers (n = 19) and lactating volunteers (n = 10) underwent imaging once. The principal diffusion coefficients (l1, l2, and l3), apparent diffusion coefficient (ADC), fractional anisotropy (FA), and maximal anisotropy (l1– l3) were calculated pixel by pixel for the fibroglandular tissue in the entire breast.

Results:

In all premenopausal volunteers, the DTI parameters exhibited high repeatability, remaining almost equal along the menstrual cycle, with a low mean within-subject coefficient of variance of l1, l2, l3, and ADC (1%–2% for all) and FA (5%), as well as a high intraclass correlation of 0.92–0.98. The diffusion coefficients were significantly lower (a) in the group without HRT use as compared with the group with HRT use (P , .01) and premenopausal volunteers (P , .01) and (b) in the lactating volunteers as compared with the premenopausal volunteers (P , .005). No significant differences in DTI parameters were found between premenopausal volunteers free of oral contraceptives and those who used oral contraceptives (P = .28–0.82) and between premenopausal volunteers and postmenopausal volunteers who used HRT (P = .31–0.93).

Conclusion:

DTI parameters are not sensitive to menstrual cycle changes, while menopause, long-term HRT, and presence of milk in lactating women affected the DTI parameters. Therefore, the timing for performing breast DTI is not restricted throughout the menstrual cycle, whereas the modulations in diffusion parameters due to HRT and lactation should be taken into account in DTI evaluation.  RSNA, 2014

q

Online supplemental material is available for this article.

 RSNA, 2014

q

672

radiology.rsna.org  n  Radiology: Volume 271: Number 3—June 2014

BREAST IMAGING: Diffusion-Tensor MR Imaging of the Breast

M

agnetic resonance (MR) imaging has become a valuable tool for breast imaging, predominantly owing to its high sensitivity for breast cancer detection (1,2). Currently, the clinical protocol for breast MR imaging is based on dynamic contrast material–enhanced (DCE) MR imaging (3), and it is applied as the standard of care for specific indications, such as patients at high risk (4,5). However, hormone-induced changes in normal breast parenchymal enhancement may influence the accuracy of breast cancer diagnosis, imposing restrictions on the selection and timing of the DCE examination (6). Consequently, as a result of menstrual cycle–dependent enhancement of normal breast parenchyma (7,8), it is recommended that DCE examination of premenopausal women should be limited during days 6–16 of the menstrual cycle (4). In addition, because of higher breast tissue perfusion of postmenopausal women who use hormone replacement therapy (HRT) (9), it is recommended that the examination be performed at least 4 weeks after HRT discontinuation (4). During

Advances in Knowledge nn The diffusion-tensor (DT) imaging parameters are not sensitive to changes during the menstrual cycle and are unchanged along the menstrual cycle, exhibiting high repeatability with intraclass correlation of 0.92–0.98. nn DT imaging parameters in the premenopausal women were not significantly different from those in the postmenopausal women who used hormone replacement therapy (HRT), while the diffusion coefficients in the postmenopausal women who did not use HRT declined by 6%–14% (P , .01). nn The diffusion coefficients in the lactating group declined by 13%– 21% (P , .005), as compared with those in the nonlactating premenopausal group.

pregnancy, contrast-enhanced MR imaging is not recommended because gadolinium-based contrast agents are known to cross the placenta (10). Lactating women can be examined safely with DCE MR imaging (11); however, its utility is limited because lactating parenchyma shows increased vascular permeability, which creates a challenge in differentiating changes due to lactation and suspicious findings (12,13). In recent years, diffusion-weighted (DW) MR imaging of the breast was demonstrated to be a useful adjunct method to DCE MR imaging (14), improving the diagnostic accuracy as compared with DCE MR imaging alone (15–17). DW MR imaging is based on the intrinsic contrast due to changes in water diffusion caused by variations in the tortuosity and restriction of tissues (18), and, therefore, does not require injection of external agents. Quantification of DW imaging data sets is achieved by means of analysis that yields a parametric map of the apparent diffusion coefficient (ADC) (19), averaged over three orthogonal directions. ADC values appear to be inversely related to cell density (20) but are not sensitive to architectural and/or structural elements. In contrast, diffusion-tensor (DT) imaging, which is based on applying diffusion gradients in multiple directions, allows diffusion measurements to be used in the investigation of microstructural features, in addition to cellular density (21). Preliminary DT imaging studies of the normal breast (22,23) and breast lesions (24–26) have been reported. These studies indicated lower values of ADC (20–22) and the principal

Implication for Patient Care nn The timing of performing DT imaging examination of the breast is not restricted to a certain phase of the menstrual cycle, whereas the modulations in diffusion parameters due to HRT and lactation should be taken into account in DT imaging evaluation.

Radiology: Volume 271: Number 3—June 2014  n  radiology.rsna.org

Nissan et al

diffusion coefficient (l1) (22) in cancerous lesions, as compared with normal breast tissue and benign lesions. The results regarding the diagnostic capacity of the fractional anisotropy (FA) index were inconclusive (20–22); however, the maximum anisotropy index (l1–l3) was found to be diagnostically relevant and lower in cancers than in normal ductal and/or glandular tissue (22). In addition, vector maps of the first eigenvector and tracking of these maps (27) displayed the main architectural features of the breast ductal tree system. The results achieved with DT imaging for cancer detection (24–26) suggest that DT imaging may have the potential to serve not only as an adjunct method to DCE examination but also as an alternative method when DCE imaging is contraindicated. Thus far, hormone-induced changes in diffusion MR imaging experiments were investigated only with preliminary DW imaging studies, which yielded conflicting results (28–30). Our goal was to investigate the parameters obtained with DT imaging of the breast throughout the menstrual cycle phases, during lactation, and after menopause, in volunteers with and without use of HRT. Published online before print 10.1148/radiol.14132084  Content codes: Radiology 2014; 271:672–680 Abbreviations: ADC = apparent diffusion coefficient DCE = dynamic contrast material–enhanced DT = diffusion-tensor DW = diffusion-weighted FA = fractional anisotropy HRT = hormone replacement therapy OC = oral contraceptives ROI = region of interest SD = standard deviation Author contributions: Guarantors of integrity of entire study, N.N., H.D.; study concepts/study design or data acquisition or data analysis/interpretation, all authors; manuscript drafting or manuscript revision for important intellectual content, all authors; approval of final version of submitted manuscript, all authors; literature research, N.N., E.F.H., M.S.F., H.D.; clinical studies, N.N., E.F.H., M.S.F., H.D.; experimental studies, N.N., M.S.F., H.D.; statistical analysis, N.N., D.G.; and manuscript editing, N.N., E.F.H., M.S.F., H.D. Conflicts of interest are listed at the end of this article.

673

BREAST IMAGING: Diffusion-Tensor MR Imaging of the Breast

Materials and Methods All protocols were approved by the Internal Review Board of Meir Medical Center, Israel, and signed informed consent was obtained from all participants.

Volunteers The study included 45 healthy female volunteers who were recruited prospectively from February 2013 to May 2013. Sixteen premenopausal volunteers (median age, 26 years; age range, 22–34 years) were recruited, including eight volunteers with a regular menstrual cycle (lasting 28–32 days) and eight volunteers who used oral contraceptives (OC). Three types of OC were used: ethinylestradiol and desogestrel based (in four volunteers), ethinylestradiol and levonorgestrel based (in one volunteer), and ethinylestradiol and drospirenone based (in three volunteers). These volunteers underwent imaging four times—once every week during one menstrual cycle. None of them were pregnant or lactating in the year prior to the study. Fifteen premenopausal volunteers completed four imaging examinations, and one volunteer quit the study after one imaging session. The median day for the first examination was day 3 of the menstrual cycle (range, day 1–5), with a median interval of 7 days between the rest of the imaging examinations (range, 5–9 days). Additionally, 19 healthy postmenopausal volunteers (median age, 57 years; range, 54–69 years) underwent imaging once, including eight volunteers who used HRT (estradiol and norethindrone acetate–based tablets or patches) on a regular basis, for a period of 3–19 years, and 11 of the same age group who did not use HRT. Furthermore, 10 lactating volunteers (median age, 35 years; age range, 30–40 years) underwent imaging once. MR Imaging Protocol Images were acquired with a 3-T whole-body MR imaging unit (Magnetom Trio Tim System; Siemens, Erlangen, Germany) that was equipped with a transmitting body coil and a 674

Nissan et al

receiving four-channel breast array coil (Siemens) or a seven-channel breast array coil (Invivo, Orlando, Fla). The imaging protocols included axial fast two-dimensional T2-weighted images with an echo time (msec)/ repetition time (msec) of 122/5500 at a resolution of 0.8 3 0.6 3 1.9–2.5 mm3 and an acquisition time of 4 minutes 26 seconds. Axial DT images with fat suppression and an acceleration factor of two were applied by using the spin-echo twice-refocused echo-planar imaging sequence (31), with diffusion gradients in 30 directions during 5 minutes, b values of 0 and 700 sec/mm2, 120/10 400, and in-plane resolution of 1.9 3 1.9 mm2. In all sequences, the section thickness of 2.5 mm was identical, allowing matched comparison of the T2weighted images and the parametric maps derived from DT images.

The DT imaging parameters were calculated in regions of interest (ROIs) as follows: Initially, ROIs of the entire two breasts were delineated manually on the fat-suppressed images obtained with a null b value, assisted by T2-weighted images of the same section. Then, image analysis was performed by using a signal intensity threshold of four times that above the noise, to include the fibroglandular tissue within the ROIs, excluding pixels with low signal intensity due to partial volume effect. The percentage of the fibroglandular tissue was estimated on central sections of non–fat-suppressed T2-weighted images by using manually delineated ROIs of the entire two breasts, including the surrounding fat, and ROIs of the corresponding fibroglandular tissue.

Image Processing

Statistical analysis included calculation of means, standard deviations (SDs), medians, and interquartile values of all DT imaging parameters in ROIs. Two-way repeated-measures analysis of variance and intraclass correlation coefficients were used to compare the four results of diffusion parameters obtained for premenopausal volunteers who underwent imaging four times during the menstrual cycle (Statistica 12.0; StatSoft, Tulsa, Okla). Within-subject coefficient of variance was calculated separately for each of the premenopausal volunteers by using the mean and SD values of the four examinations. Between-subjects coefficient of variance was calculated on the basis of the mean values of 16 premenopausal volunteers. The two-tailed Student t test (unpaired or paired, as indicated) was applied for evaluating differences between premenopausal volunteers with and without OC use and to compare the groups of premenopausal volunteers, postmenopausal volunteers with or without HRT use, and lactating volunteers (Excel 2010; Microsoft, Redmond, Wash). No adjustment for multiple comparisons has been made. The significance level was set at P , .05.

The DT parameters and the anisotropy indexes were calculated pixel by pixel for each given section by using home-built software programed in C++. The Digital Imaging and Communications in Medicine images were read by the program, followed by the application of a pixel-by-pixel fitting of the diffusion coefficient in each gradient direction, according to the Stejskal-Tanner equation (32). Further nonlinear regression fitting of the directional diffusion coefficients to a symmetric tensor yielded six tensor parameters: Dxx, Dyy, Dzz, Dxy, Dxz, and Dyz. The application of principal component analysis yielded for each pixel the three eigenvectors (v1, v2, and v3) that define the principal diffusion directions in three orthogonal axes, which coincide with the diffusion frame of the tissue, and their corresponding three eigenvalues, arranged from high to low values (l1, l2, and l3), that quantify the principal diffusion coefficients (18). Average diffusivity, defined as ADC, was calculated as the mean of the three eigenvalues. The anisotropy was characterized by the maximal anisotropy index, l1–l3, and by the FA (18).

Statistical Analysis

radiology.rsna.org  n Radiology: Volume 271: Number 3—June 2014

BREAST IMAGING: Diffusion-Tensor MR Imaging of the Breast

exhibiting a mean within-subject coefficient of variance of 1%–2% for l1, l2, l3, and ADC. FA and l1–l3 were slightly less reproducible but still very similar, showing within-subject coefficient of variance of 5%–6%. The high level of proximity between the results of the four DT imaging examinations along the menstrual cycle was further supported by the high intraclass correlation rates for the DT imaging parameters (Table 1). The repeatability of the DT imaging parameters for each of the volunteers, as well as the gradual decline from the first to the third measured eigenvalues (l1, l2, and l3), are demonstrated in Figure 1. Qualitative assessment of the parametric DT imaging maps (l1, l2, and l3 maps and three calculated parameters: ADC, FA, and l1–l3 maps) of a typical premenopausal volunteer, recorded in the second and the fourth week of the menstrual cycle (Fig E1 [online]), demonstrates the equivalence in the spatial distribution of all the parameters. Analysis of the pixel distribution by using histograms in breast ROIs showed not only repeatable means, but also high similarity in the distribution of pixels within imaging examinations during the cycle, as demonstrated for the first eigenvalue (l1) of a premenopausal volunteer (Fig 2).

Figure 1

Figure 1:  Graphs show DT imaging parameters during the menstrual cycle. Mean values are presented for the four examinations in the 15 premenopausal nonlactating volunteers in two groups, eight subjects who were OC free (left) and seven subjects who used OC (right). No significant difference was found between those who used OC and those who did not (P = .28, P =.54, P =.82, and P =.55 for l1, l2, l3, and ADC, respectively).

Results Premenopausal Volunteers and Menstrual Cycle Repeatability Comparison between the premenopausal nonlactating groups with and without OC use did not show any significant differences for all DT imaging

Nissan et al

parameters (Fig 1). Therefore, the statistical analysis of the DT imaging parameters was performed for all 15 premenopausal nonlactating volunteers together as one group (Table 1). The four DT imaging results of each of the 15 volunteers remained almost equal along the menstrual cycle,

Radiology: Volume 271: Number 3—June 2014  n  radiology.rsna.org

Postmenopausal Volunteers We found significantly higher values of l1, l2, l3, and ADC by 6%–12% (P , .01) in the volunteers who used HRT, as compared with the postmenopausal volunteers who did not use HRT. However, FA and l1–l3 did not show significant differences between these groups (Table 2). Comparison of the postmenopausal groups with the nonlactating cohort showed that while there was no significant difference between the DT imaging parameters of the premenopausal volunteers and the postmenopausal volunteers who used HRT, the parameters l1, l2, l3, and ADC were found to be significantly higher by 9%–14% (P , .01) in the premenopausal group as compared with the 675

BREAST IMAGING: Diffusion-Tensor MR Imaging of the Breast

Nissan et al

Table 1 Breast DT Imaging Parameters of Premenopausal Nonlactating Volunteers Parameter Mean value Range Within-subject SD Within-subject CV (%) Intraclass correlation coefficient Between-subjects SD Between-subjects CV (%)

l1 (31023 mm2/sec)

l2 (31023 mm2/sec)

l3 (31023 mm2/sec)

ADC (31023 mm2/sec)

FA

l1–l3 (31023 mm2/sec)

2.44 2.02–2.64 0.04 1.7 0.93 0.16 6.6

1.91 1.56–2.22 0.02 1.2 0.98 0.19 9.9

1.49 1.13–1.92 0.03 1.9 0.98 0.21 14.9

1.95 1.59–2.24 0.02 1.2 0.98 0.18 9.2

0.25 0.16–0.33 0.01 5.3 0.92 0.04 16

0.95 0.65–1.17 0.05 5.8 0.84 0.13 13.7

Note.—Data are mean values in 15 volunteers. For each volunteer, we used the mean value from four examinations performed during one menstrual cycle. Within-subject SD, within-subject coefficient of variance, between-subjects SD, between-subjects coefficient of variance, and intraclass correlation coefficient values were calculated as described in Materials and Methods. CV = coefficient of variance.

Table 2 Breast DT Imaging Parameters of Postmenopausal Volunteers with and without HRT Use Parameter Mean 6 SD in HRT group Range in HRT group Mean 6 SD in group without HRT Range in group without HRT Postmenopausal HRT group vs no HRT group* Postmenopausal HRT group vs premenopausal group* Postmenopausal no HRT group vs premenopausal group*

l1 (31023 mm2/sec)

l2 (31023 mm2/sec)

l3 (31023 mm2/sec)

ADC (31023 mm2/sec)

FA

l1–l3 (31023 mm2/sec)

2.37 6 0.13 2.20–2.51 2.22 6 0.09 2.05–2.36 .01 .31

1.86 6 0.07 1.74–1.94 1.71 6 0.14 1.53–1.98 .005 .36

1.45 6 0.10 1.28–1.64 1.28 6 0.18 1.09 6 1.62 .02 .57

1.90 6 0.07 1.78–1.96 1.73 6 0.13 1.56–1.98 .003 .35

0.25 6 0.04 0.16–0.30 0.28 6 0.05 0.20–0.34 .18 .93

0.92 6 0.17 0.58–1.06 0.94 6 0.15 0.71–1.12 .78 .72

.001

.003

.01

.001

.15

.95

Note.—Unless otherwise indicated, data are mean values 6 SDs in in eight postmenopausal volunteers who used HRT (median age, 58 years) and 11 postmenopausal volunteers who did not use HRT (median age, 57 years). * Data are P values derived from statistical analysis of DT imaging parameters in the same groups.

postmenopausal volunteers who did not use HRT. A summary of the DT imaging parameters obtained for the diverse hormonal status groups is provided (Table 3 and Figure E2 [online]).

The Lactating Breast In comparison to the nonlactating premenopausal volunteers, the lactating breast values of l1, l2, l3, ADC, and l1–l3 were significantly lower by 13%– 21% (P , .005), although FA remained similar (Fig E2 [online]), whereas the percentage of fibroglandular tissue in the lactating group (mean 6 SD, 47% 6 11) was not significantly different from that in the nonlactating premenopausal group (mean, 51% 6 17) (P = 676

.83). The parameters l1, l2, l3, and ADC were highly similar in the 10 lactating volunteers, exhibiting a betweensubjects coefficient of variance of 3%– 5% for these parameters. However, FA and l1–l3 exhibited variations between the volunteers with between-subjects coefficient of variance of 13% and 16%, respectively (Table 3). The diffusion vector map and diffusion coefficient maps of the lactating breast are illustrated in Figure 3 and Figure E3 (online). In both, there is a subjective appearance of direction in the anterior-posterior axis, reflecting the structure of milk ducts heading from the base of the breast toward the nipple. Figure E3 (online) also shows

a lower level of anisotropy and faster diffusivity in the nipple area, as compared with the periphery of the breast. Based also on the T2-weighted image, this is most likely due to broadening of the ducts toward the nipple and consequently reduced water restriction.

Discussion The normal breast tissue undergoes changes in the epithelial and stromal elements in response to different hormonal states. During the menstrual cycle, five phases, varying in the vascularity and histologic features of the epithelial and stromal components, have been described (proliferative, follicular,

radiology.rsna.org  n Radiology: Volume 271: Number 3—June 2014

BREAST IMAGING: Diffusion-Tensor MR Imaging of the Breast

Nissan et al

Figure 2

Figure 2:  Parametric maps and distribution histograms of l1 in an OC-free 27-year-old premenopausal woman throughout the menstrual cycle. Maps depict a central axial section through the two breasts and are overlaid on the corresponding T2-weighted image. Numbers I–IV represent the week of the examination during the menstrual cycle. l1 color-coded scale is given in units of 1023 square millimeters per second.

luteal, secretory, and menstrual) (33). During the cycle, the vascularity of the breast tissue fluctuates in response to hormonal changes. In particular, cyclic vascular dilation occurs when progesterone levels are high (34), resulting in increased diffuse and focal enhancements of normal breast parenchyma at DCE MR imaging (7,8). The most distinct discriminative histologic features were based on differences in luminal and basal myoepithelial cell layers, degree and proportion of myoepithelial cell vacuolization, and presence of mitoses and apoptosis, as well as presence

of edema in interlobular stroma (35). The diffusion parameters determined in our study mainly reflected the motion of water in the large fraction of the extracellular compartment in both the ductal-glandular system and the connective-fibrous tissue. Therefore, it is most likely that the diffusion coefficients and the anisotropy indexes were not sensitive to the previously mentioned microscopic or cellular changes that occurred during the menstrual cycle. Although the edematous process could have reduced the viscosity and therefore affected the rate of diffusion,

Radiology: Volume 271: Number 3—June 2014  n  radiology.rsna.org

it appears from our results that it was not sufficiently robust to change significantly the diffusion coefficients. Because of its clinical relevance, there is a rising interest in the effects of the menstrual cycle on different breast imaging modalities (36–38) and in DCE MR imaging in particular, where the cyclic vasodilatation phenomenon may be interpreted as a false-positive finding, creating a diagnostic challenge (39). More recently, findings of studies on background parenchymal enhancement during the menstrual cycle (40–42) have supported the current recommendation 677

BREAST IMAGING: Diffusion-Tensor MR Imaging of the Breast

Nissan et al

Figure 3

Figure 3:  Direction and vector maps of a lactating breast in a 30-year-old volunteer. Maps depict a central axial section of the right breast and are overlaid on the corresponding T2-weighted image. The direction map represents the direction of the first principal eigenvector in a three-color code: Red indicates left to right, green indicates anterior to posterior, and blue indicates head to feet. The vector map presents in white sticks the direction of the first principal eigenvector. Note that in the direction map and the vector map, the predominant direction (green) of the first eigenvector is toward the nipple, as expected for the prominent ductal tree system of a lactating breast.

Table 3 Breast DT Imaging Parameters of Lactating Volunteers Parameter Mean 6 SD Range Between-subjects CV (%) Lactating group vs premenopausal group* Lactating group vs HRT group* Lactating group vs no HRT group*

l1 (31023 mm2/sec)

l2 (31023 mm2/sec)

l3 (31023 mm2/sec)

ADC (31023 mm2/sec)

FA

l1–l3 (31023 mm2/sec)

2.05 6 0.11 1.91–2.21 5.4 ,.005 ,.005 ,.005

1.64 6 0.06 1.52–1.71 3.7 ,.005 ,.005 .16

1.30 6 0.06 1.18–1.39 4.6 ,.005 ,.005 .72

1.66 6 0.06 1.54–1.73 3.6 ,.005 ,.005 .12

0.23 6 0.03 0.20–0.30 13.0 .25 .39 ,.05

0.75 6 0.12 0.60–0.99 16.0 ,.005 ,.05 ,.005

Note.—Unless otherwise indicated, data are DT imaging parameters and between-subjects coefficients of variance in 10 lactating volunteers (median age, 35 years). CV = coefficient of variance. * Data are P values derived from statistical analysis conducted in the same group.

to limit DCE MR imaging to the second week of the menstrual cycle. In view of the DCE imaging results, menstrual cycle effects on DW imaging of the breast were also examined by Partridge et al (22), who showed nonsignificant 5% variations between four weekly measurements of ADC (n = 6) along one menstrual cycle. More recently, O’Flynn et al (29) reported that no significant difference in ADC was found when comparing the proliferative and secretory phases of the menstrual 678

cycle (n = 13), while on the contrary, a significant difference in ADC values between the luteal and follicular phases of the menstrual cycle was reported by Clendenen et al (30). In our study, we examined the menstrual cycle influence not only on ADC but also on l1, l2, l3, and FA. The results of four DT imaging measurements, each in a different phase of the menstrual cycle, showed a high reproducibility for the diffusion parameters l1, l2, l3, ADC, FA, and l1–l3. The

higher reproducibility rates for ADC compared to FA are in agreement with Tagliafico et al (23). These observations may suggest that DT parameters of the normal breast fibroglandular tissue are minimally affected by the histologic and vascular changes that occur during the menstrual cycle. As can be seen on the parametric maps and the pixel histograms, however, the DT parameter distribution is heterogeneous throughout the breast, and there are differences even between neighboring

radiology.rsna.org  n Radiology: Volume 271: Number 3—June 2014

BREAST IMAGING: Diffusion-Tensor MR Imaging of the Breast

pixels. Thus, it was important to obtain ROIs of the entire breast fibroglandular tissue rather than delineate regional ROIs. During menopause, after the decline in hormone levels, the breast undergoes an atrophy process, resulting in a decreased amount of fibroglandular tissue, which is replaced by adipose tissue. DCE MR imaging showed decreases in both background parenchymal enhancement and fibroglandular tissue after menopause (43). However, the use of HRT was found to affect the breast vascularity (9), and, therefore, HRT discontinuation has been recommended before DCE MR imaging (4). In a DW imaging study, O’Flynn et al reported lower ADC values in the postmenopausal cohort, as compared with the premenopausal volunteers. However, the differences between the postmenopausal women with and without HRT use were not significantly different (29). We also observed a decline in the DT coefficients and ADC values in premenopausal women as compared with postmenopausal women who did not use HRT (P , .0004, P , .005, P , .02, and P , .002 for l1, l2, l3, and ADC, respectively), while the anisotropy indexes FA and l1–l3 remained similar. However, we did find significant differences in the DT coefficients between the postmenopausal women with and without HRT use, as well as similar values for the premenopausal women and postmenopausal women who used HRT. This finding of similar results for premenopausal women and postmenopausal women who used HRT is in accord with the expected hormonal influences of HRT (44). However, further studies with larger cohorts are required to validate these findings. We also found that the prime DT coefficient was significantly lower in the lactating volunteers (P , .005) than in any other group of women, and ADC was lower in that group than in the premenopausal women and the postmenopausal women who used HRT (P , .005). A reduction in ADC could be attributed to the higher viscosity of the milk, which dominates the liquid

environment in the lactating breast, as compared with the normal breast ductal water milieu. This study has clinical and technical limitations: The group sizes were small, and a larger number of volunteers is required to establish the complete range of each group. Additionally, the current experimental protocol of breast DT imaging is highly sensitive to the presence of fat, requiring an “ideal” and homogeneous fat suppression, and it may be affected by echo-planar imaging–related artifacts due to inhomogeneous field and tissue/air susceptibility differences (45), as well as general challenges in diffusion MR imaging, such as eddy current–induced distortions and subject motion (46). In summary, we found that breast DT parameters are not sensitive to menstrual cycle changes, whereas lactation and long-term use of HRT do not affect diffusion parameters. Therefore, the timing for performing breast DT imaging is not restricted throughout the menstrual cycle, whereas the modulations in diffusion parameters due to HRT and lactation should be taken into account at DT imaging evaluation. Acknowledgments: We acknowledge the examinees for their volunteering spirit. The professional work of the MR imaging technicians, Fanny Attar and Nachum Stern, is gratefully acknowledged, as well as the secretarial assistance of Martie Spiegel. Disclosures of Conflicts of Interest: N.N. No relevant conflicts of interest to disclose. E.F.H. No relevant conflicts of interest to disclose. M.S.F. No relevant conflicts of interest to disclose. D.G. No relevant conflicts of interest to disclose. H.D. Financial activities related to the present article: none to disclose. Financial activities not related to the present article: none to disclose. Other relationships: author has a patent for a method and apparatus for imaging of ductal tube tracking in breast cancer diagnosis.

References 1. Berg WA, Gutierrez L, NessAiver MS, et al. Diagnostic accuracy of mammography, clinical examination, US, and MR imaging in preoperative assessment of breast cancer. Radiology 2004;233(3):830–849. 2. Van Goethem M, Schelfout K, Dijckmans L, et al. MR mammography in the pre-operative staging of breast cancer in patients with dense breast tissue: comparison with

Radiology: Volume 271: Number 3—June 2014  n  radiology.rsna.org

Nissan et al

mammography and ultrasound. Eur Radiol 2004;14(5):809–816. 3. Sinha S, Sinha U. Recent advances in breast MRI and MRS. NMR Biomed 2009;22(1): 3–16. 4. Sardanelli F, Boetes C, Borisch B, et al. Magnetic resonance imaging of the breast: recommendations from the EUSOMA working group. Eur J Cancer 2010;46(8):1296– 1316. 5. Lee CH, Dershaw DD, Kopans D, et al. Breast cancer screening with imaging: recommendations from the Society of Breast Imaging and the ACR on the use of mammography, breast MRI, breast ultrasound, and other technologies for the detection of clinically occult breast cancer. J Am Coll Radiol 2010;7(1):18–27. 6. Kuhl C. The current status of breast MR imaging. Part I. Choice of technique, image interpretation, diagnostic accuracy, and transfer to clinical practice. Radiology 2007;244(2):356–378. 7. Kuhl CK, Bieling HB, Gieseke J, et al. Healthy premenopausal breast parenchyma in dynamic contrast-enhanced MR imaging of the breast: normal contrast medium enhancement and cyclical-phase dependency. Radiology 1997;203(1):137–144. 8. Müller-Schimpfle M, Ohmenhaüser K, Stoll P, Dietz K, Claussen CD. Menstrual cycle and age: influence on parenchymal contrast medium enhancement in MR imaging of the breast. Radiology 1997;203(1):145–149. 9. Delille JP, Slanetz PJ, Yeh ED, Kopans DB, Halpern EF, Garrido L. Hormone replacement therapy in postmenopausal women: breast tissue perfusion determined with MR imaging—initial observations. Radiology 2005;235(1):36–41. 10. Vashi R, Hooley R, Butler R, Geisel J, Philpotts L. Breast imaging of the pregnant and lactating patient: imaging modalities and pregnancy-associated breast cancer. AJR Am J Roentgenol 2013;200(2):321–328. 11. Ayyappan AP, Kulkarni S, Crystal P. Pregnancy-associated breast cancer: spectrum of imaging appearances. Br J Radiol 2010;83(990):529–534. 12. Talele AC, Slanetz PJ, Edmister WB, Yeh ED, Kopans DB. The lactating breast: MRI findings and literature review. Breast J 2003;9(3):237–240. 13. Espinosa LA, Daniel BL, Vidarsson L, Zakhour M, Ikeda DM, Herfkens RJ. The lactating breast: contrast-enhanced MR imaging of normal tissue and cancer. Radiology 2005;237(2):429–436.

679

BREAST IMAGING: Diffusion-Tensor MR Imaging of the Breast

Nissan et al

14. Guo Y, Cai YQ, Cai ZL, et al. Differentiation of clinically benign and malignant breast lesions using diffusion-weighted imaging. J Magn Reson Imaging 2002;16(2):172–178.

25. Baltzer PA, Schäfer A, Dietzel M, et al. Diffusion tensor magnetic resonance imaging of the breast: a pilot study. Eur Radiol 2011;21(1):1–10.

36. Miglioretti DL, Walker R, Weaver DL, et al. Accuracy of screening mammography varies by week of menstrual cycle. Radiology 2011;258(2):372–379.

15. Parsian S, Rahbar H, Allison KH, et al. Nonmalignant breast lesions: ADCs of benign and high-risk subtypes assessed as falsepositive at dynamic enhanced MR imaging. Radiology 2012;265(3):696–706.

26. Eyal E, Shapiro-Feinberg M, Furman-Haran E, et al. Parametric diffusion tensor imaging of the breast. Invest Radiol 2012;47(5):284– 291.

37. Wojcinski S, Cassel M, Farrokh A, et al. Variations in the elasticity of breast tissue during the menstrual cycle determined by real-time sonoelastography. J Ultrasound Med 2012;31(1):63–72.

16. El Khouli RH, Jacobs MA, Mezban SD, et al. Diffusion-weighted imaging improves the diagnostic accuracy of conventional 3.0-T breast MR imaging. Radiology 2010;256(1): 64–73. 17. Bogner W, Gruber S, Pinker K, et al. Diffusion-weighted MR for differentiation of breast lesions at 3.0 T: how does selection of diffusion protocols affect diagnosis? Radiology 2009;253(2):341–351. 18. Le Bihan D, Mangin JF, Poupon C, et al. Diffusion tensor imaging: concepts and applications. J Magn Reson Imaging 2001;13(4): 534–546. 19. Koh DM, Collins DJ. Diffusion-weighted MRI in the body: applications and challenges in oncology. AJR Am J Roentgenol 2007;188(6): 1622–1635.

27. Reisert MM, Eyal E, Grobgeld D, Degani H, Hennig J. Diffusion tensor based reconstruction of the ductal tree [abstr]. In: Proceedings of the Nineteenth Meeting of the International Society for Magnetic Resonance in Medicine. Berkeley, Calif: International Society for Magnetic Resonance in Medicine, 2011; 3649. 28. Partridge SC, McKinnon GC, Henry RG, Hylton NM. Menstrual cycle variation of apparent diffusion coefficients measured in the normal breast using MRI. J Magn Reson Imaging 2001;14(4):433–438. 29. O’Flynn EA, Morgan VA, Giles SL, deSouza NM. Diffusion weighted imaging of the normal breast: reproducibility of apparent diffusion coefficient measurements and variation with menstrual cycle and menopausal status. Eur Radiol 2012;22(7):1512–1518.

20. Paran Y, Bendel P, Margalit R, Degani H. Water diffusion in the different microenvironments of breast cancer. NMR Biomed 2004; 17(4):170–180.

30. Clendenen TV, Kim S, Moy L, et al. Magnetic resonance imaging (MRI) of hormoneinduced breast changes in young premenopausal women. Magn Reson Imaging 2013;31(1):1–9.

21. Hagmann P, Jonasson L, Maeder P, Thiran JP, Wedeen VJ, Meuli R. Understanding diffusion MR imaging techniques: from scalar diffusion-weighted imaging to diffusion tensor imaging and beyond. RadioGraphics 2006;26(Suppl 1):S205–S223.

31. Reese TG, Heid O, Weisskoff RM, Wedeen VJ. Reduction of eddy-current-induced distortion in diffusion MRI using a twice-refocused spin echo. Magn Reson Med 2003;49(1): 177–182.

22. Partridge SC, Murthy RS, Ziadloo A, White SW, Allison KH, Lehman CD. Diffusion tensor magnetic resonance imaging of the normal breast. Magn Reson Imaging 2010;28(3): 320–328. 23. Tagliafico A, Rescinito G, Monetti F, et al. Diffusion tensor magnetic resonance imaging of the normal breast: reproducibility of DTI-derived fractional anisotropy and apparent diffusion coefficient at 3.0 T. Radiol Med (Torino) 2012;117(6):992–1003. 24. Partridge SC, Ziadloo A, Murthy R, et al. Diffusion tensor MRI: preliminary anisotropy measures and mapping of breast tumors. J Magn Reson Imaging 2010;31(2): 339–347.

680

32. Stejskal EO, Tanner JE. Spin diffusion measurements: spin echoes in the presence of a time-dependent field gradient. J Chem Phys 1965;42(1):288–292. 33. Vogel PM, Georgiade NG, Fetter BF, Vogel FS, McCarty KS Jr. The correlation of histologic changes in the human breast with the menstrual cycle. Am J Pathol 1981;104(1): 23–34. 34. Weinstein SP, Conant EF, Sehgal CM, Woo IP, Patton JA. Hormonal variations in the vascularity of breast tissue. J Ultrasound Med 2005;24(1):67–72; quiz 74. 35. Ramakrishnan R, Khan SA, Badve S. Morphological changes in breast tissue with menstrual cycle. Mod Pathol 2002;15(12):1348–1356.

38. Lin CY, Ding HJ, Liu CS, Chen YK, Lin CC, Kao CH. Correlation between the intensity of breast FDG uptake and menstrual cycle. Acad Radiol 2007;14(8):940–944. 39. Macura KJ, Ouwerkerk R, Jacobs MA, Bluemke DA. Patterns of enhancement on breast MR images: interpretation and imaging pitfalls. RadioGraphics 2006;26(6): 1719–1734; quiz 1719. 40. Amarosa AR, McKellop J, Klautau Leite AP, et al. Evaluation of the kinetic properties of background parenchymal enhancement throughout the phases of the menstrual cycle. Radiology 2013;268(2):356–365. 41. Kang SS, Ko EY, Han BK, Shin JH, Hahn SY, Ko ES. Background parenchymal enhancement on breast MRI: influence of menstrual cycle and breast composition. J Magn Reson Imaging 2013 Apr 30. [Epub ahead of print] 42. Scaranelo AM, Carrillo MC, Fleming R, Jacks LM, Kulkarni SR, Crystal P. Pilot study of quantitative analysis of background enhancement on breast MR images: association with menstrual cycle and mammographic breast density. Radiology 2013; 267(3):692–700. 43. King V, Gu Y, Kaplan JB, Brooks JD, Pike MC, Morris EA. Impact of menopausal status on background parenchymal enhancement and fibroglandular tissue on breast MRI. Eur Radiol 2012;22(12):2641– 2647. 44. Söderqvist G. Effects of sex steroids on proliferation in normal mammary tissue. Ann Med 1998;30(6):511–524. 45. Furman-Haran E, Eyal E, Shapiro-Feinberg M, et al. Advantages and drawbacks of breast DTI. Eur J Radiol 2012;81(Suppl 1): S45–S47. 46. Jones DK, Cercignani M. Twenty-five pitfalls in the analysis of diffusion MRI data. NMR Biomed 2010;23(7):803–820.

radiology.rsna.org  n Radiology: Volume 271: Number 3—June 2014