Fibroglandular tissue and quantitative background parenchymal enhancement on magnetic resonance breast images are inversely correlated with menopause in Thai women
Article Category: Brief communication (Original)
Published Online: Jan 31, 2017
Page range: 519 - 526
DOI: https://doi.org/10.5372/1905-7415.0904.422
Keywords
© 2015 Anchukorn Rijiravanich, Jatuporn Chayakulkheeree
This work is licensed under the Creative Commons Attribution-NonCommercial-NoDerivatives 3.0 License.
Radiographically, breast consists of 2 main compositions namely fibroglandular tissue (FGT) and fat. FGT composes of stroma or fibrous connective tissue and the glandular epithelial-myoepithelial cells that line the ducts of the breast [1]. In mammography, FGT shows as regions of brightness, which is referred to as ‘mammographic density’. It is well known that high breast density decreases the sensitivity of breast cancer detection in mammography [2, 3]. This can be a problem for mammographic interpretation, especially in Asian women because they generally have higher breast densities than women of European ancestry [4, 5]. In such cases, dynamic contrast enhanced magnetic resonance imaging (DCE-MRI) of breast can be performed [6].
Dynamic breast MRI has become the most sensitive of all imaging techniques for detecting breast cancer [7, 8]. However, enhancement on DCE breast MRI is not limited to malignancies or discrete benign lesions of the breast. Normal breast parenchyma also enhances with contrast agent administration and is termed ‘background parenchymal enhancement’ (BPE) [9], which is highly variable over the course of the dynamic series [7]. Previous studies have shown that the degree of BPE does not necessarily correlate with the amount of fibroglandular parenchyma present in mammography, but rather depends on the patient’s hormonal status, age, and the amount of glandular tissue [7, 8, 10-14]. Because breast density in mammography reduces the technique’s sensitivity, moderate/marked BPE affects the detection and staging of breast cancer by increasing the false negative rate, reducing sensitivity for multifocal/ multicentric disease, and decreasing the accuracy of evaluating the extent of the tumor [7, 15, 16].
To our knowledge, there has not yet been a report of BPE in Asian women. This study sought to determine the qualitative and quantitative correlations between the amount of FGT and BPE on MR breast images from Thai women, and correlate these variables with their menopausal status.
Images from Thai women who underwent DCE breast MRI at our institution between March 2012 to July 2014 were included in this retrospective study. We excluded images from women who were imaged during the third or fourth week of their menstrual cycle, in a perimenopausal period, and who were taking chemotherapy and/or hormonal therapy, or had been within 6 months before MRI. The breast image to be reviewed was chosen by exclusion criteria, which included the presence of a breast implant, bilateral enhancing lesions, and ipsilateral enhancing lesions, i.e. mass-like and nonmass-like enhancement. In patients with breast cancer, contralateral nonirradiated breast, and/or nonoperative breast images were evaluated. Thus, only the normal side of breast was studied. Images from 95 patients were included in the study, including from 81 women in postmenopausal status and 14 women in premenopausal status.
This study was approved by the institutional review board of the Faculty of Medicine, Chulalongkorn University, Bangkok, Thailand (approval No. 318/56).
DCE breast MRI were acquired on a 1.5 T scanner (Siemen Magnetom Espree, Siemens Medical Solutions, Erlangen, Germany) with a dedicated breast coil. The patient had been placed in a prone position. A transverse 3-dimensional high-resolution T1-weighted fast gradient-echo sequence [Echo time/ repetition time 4.75/1.45 ms; flip angle 10°; field of view 360 × 360 mm2; slice thickness 3 mm; voxel size 1.0 mm × 0.8 mm × 1.3 mm] was performed before administration of contrast agent (0.1 mmol/kg gadolinium diethylenetriamine pentaacetic acid (Gd-DTPA) (gadobutrol 1.0 mmol/ml) at a flow rate of 2.0 ml/s), followed by repeated performance of the same sequence at 1, 2, 3, 4, 5, and 6 min after administration of contrast agent. Additional axial T2-weighted fat-suppressed fast gradient-echo sequences were performed before administration of contrast material.
MRI-FGT was visually identified on precontrast T1-weighted fat-suppressed images of all axial slices and the volume and distribution compared with the entire breast volume. The amount of FGT, based on qualitative evaluation, was then classified according to the American College of Radiology (ACR) Breast Imaging Reporting and Data System (BI-RADS) 2013 lexicon categories as: almost entirely fat, scattered FGT, heterogeneous FGT, or extreme FGT [17].
BPE was visually identified on postcontrast T1-weighted fat-suppressed fast gradient-echo dynamic images based on a combination of early postcontrast and late phase subtraction images [14]. BPE was categorized qualitatively according to the ACR BIRADS 2013 BPE lexicon as minimal, mild, moderate, or marked [17, 18]. The intensity of enhancement was subsequently quantified by manually tracing a 1.0 cm2 square-shaped region of interest on the most enhancing part of the normal breast parenchyma (
Figure 1
All images were evaluated by consensus of 2 radiologists with up to 6 years’ experience in breast MRI. Both were blinded to the patients’ information.
IBM Statistics for Windows, version 20 (IBM Corp, Armonk, N Y, USA) was used for statistical analyses. A Pearson chi-square test was used to determine association between categorical MRI-FGT and BPE. Then correlation between different MRI-FGT, BPE, and quantitative BPE (PEK) were determined using a Spearman rank correlation test and Kruskal–Wallis test. One-way analysis of variance was used to describe the relationship between age and MRI-FGT. The impact of menopausal status on MRI-FGT and BPE was assessed using a Mann–Whitney
The images from 95 patients were included in this study, 14 (15%) patients were premenopausal and 81 (77%) were postmenopausal. The mean patient age was 56.4 ± 19.4 (range 32-82) years. Mostly, patients showed scattered FGT (26 (27%)) and heterogeneous FGT (44 (46%)), and minimal BPE (41 (43%)) to mild BPE (39 (41%)), respectively. All of the patients showed a gradually increased BPE signal intensity pattern (type 1). There was a significant correlation between age and the amount of MRI-FGT (
In the premenopausal patients, there was no woman who had almost entirely fatty breast or scattered FGT (
Description of the 95 normal breast magnetic resonance imaging analyses by visually assessment of fibroglandular tissue and background parenchymal enhancement classified according to the American College of Radiology Breast Imaging Reporting and Data System 2013
| Menopausal status | Background parenchymal enhancement | Total | |||
|---|---|---|---|---|---|
| Minimal | Mild | Moderate | Marked | ||
| Postmenopause | |||||
| Fibroglandular tissue | |||||
| Almost entirely fat | 8 | 2 | 0 | 0 | 10 |
| Scattered | 11 | 11 | 4 | 0 | 26 |
| Heterogeneous | 15 | 17 | 4 | 0 | 36 |
| Extreme | 6 | 12 | 0 | 9 | |
| Total | 40 | 31 | 10 | 0 | 81 |
| Premenopause | |||||
| Fibroglandular tissue | |||||
| Heterogeneous | 0 | 51 | 2 | 8 | |
| Extreme | 1 | 30 | 2 | 6 | |
| Total | 1 | 81 | 4 | 14 | |
There was no postmenopausal woman who had marked BPE. Minimal or mild BPE patterns were most commonly found (
Figure 2
Figure 3
Neither qualitative BPE nor quantitative PEK revealed any significant correlation with the amount of FGT, when tested among all patients, either premenopausal or postmenopausal (
Correlation between magnetic resonance imaging fibroglandular tissue and background parenchymal enhancement (BPE), including the visually assessed BPE classified according to the American College of Radiology Breast Imaging Reporting and Data System 2013 and quantitative enhancement kinetics including the enhancement at 90 seconds (Eearly), at 6 minutes (Epeak), and signal enhancement ratio (SER).
| BPE | Eearly | Eearly | SER | |
|---|---|---|---|---|
| Correlation coefficient | 0.179 | 0.002 | 0.103 | –0.116 |
| Significance (2-tailed) | 0.083 | 0.985 | 0.322 | 0.265 |
| 95 | 95 | 95 | 95 |
The amount of FGT and the degree of BPE were significantly different between pre and postmenopausal women (
Comparison of magnetic resonance imaging fibroglandular tissue (FGT), visually assessed background parenchymal enhancement (BPE), enhancement at 90 s (Eearly), enhancement at 6 min (Epeak), and signal enhancement ratio (SER) between pre and postmenopausal women.
| FGT | BPE | Eearly | Epeak | SER | |
|---|---|---|---|---|---|
| Mann–Whitney | 243 | 270 | 430 | 369 | 538.5 |
| Wilcoxon | 3564 | 3591 | 3751 | 3690. | 643.5 |
| –3.6 | -3.4 | –1.5 | –2.1 | –0.3 | |
| Asymp significance (2-tailed) | 0.000 | 0.001 | 0.15 | 0.04 | 0.77 |
Like previous studies demonstrating the degree of background enhancement in breast MRI does not correlate with mammographic breast density [8, 10, 11], this study demonstrates no correlation between degree of BPE and amount of FGT on breast MRI examinations of both pre and postmenopausal Thai women. Eearly, a quantitative measurement of BPE, has previously shown not to be significantly different between groups with mammographic breast density [13]. However, a weak, but significant correlation has been reported by King et al. [14] in both premenopausal and postmenopausal women (
The effect of menstruation on the amount of fibroglandular parenchyma and BPE we found in Thai women were consistent with previous studies of women of European ancestry, showing that premenopausal women had generally more fibroglandular tissue and degree of BPE compared with postmenopausal women [11, 13]. A paired study design in which BPE and MRI-FGT were examined both before and after menopause in the same woman showed that both BPE and MRI-FGT were visibly decreased after menopause. These results suggest that menopause effects a substantial decrease in BPE and fibroglandular tissue [14]. We found that the quantitative BPE parameter, Epeak, was significantly higher in premenopausal women than postmenopausal women. The Eearly and SER were not significantly different between pre and postmenopausal Thai women, suggesting that they were not affected by the menopause. These finding are consistent with those in a study of Turkish women [10] that found Eearly was not affected by menopause. By contrast, a study of Canadian women found a significantly higher Eearly in premenopausal women compared with postmenopausal women [13]. In our opinion, this inconsistency possibly reflects differences related to ethnicity, not only in the structural components of breast tissue, but also a response to hormones demonstrated by normal BPE on a DCE MRI series. There is evidence that BPE is generally higher in women <45-50 years old than in older women [11, 15] and that mammographic breast density, FGT assessed on MRI, and BPE decrease significantly with increasing patient age [14, 20]. The present study also clearly demonstrated a significant effect of age on FGT by breast MRI. The effect of age on BPE was not clear.
Mammographic dense breast has long been known as a major risk factor for breast cancer, having an estimated 4-6-fold increased risk of developing breast cancer compared with fatty breasts [21, 22]. FGT is seen in breast MRI and as expected, MRI density is strongly correlated with visually assessed and computed mammographic density [23-26].
To date, the effects of BPE on breast lesion detection remain controversial. Previously, it was believed that normal fibroglandular tissue, which enhances strongly, is associated with higher rate of abnormal interpretation [15]. Effects of moderate or marked BPE on breast MRI interpretation are such that the radiologist may take time and have difficulty in interpreting any lesion. However, BPE and MRI-FGT have been reported to have no influence on the detection of either benign or malignant lesions, or mass or nonmass enhancement lesions [27]. Before applying this finding to our daily practice, it needs to be noted that the examination was based on diffusion-weighted images. The apparent diffusion coefficient (ADC) for normal breast tissue are not influenced by BPE, while ADC is significantly lower in predominantly fatty breasts compared with dense breasts.
King et al. have assessed breast cancer risk using BPE and MRI-FGT in a case control study [28]. Breast MRI screening of 1275 women at high risk of breast cancer was retrospectively reviewed. While BPE varies markedly between different women and decreases significantly after menopause [13, 14], increasing BPE shows a strong prediction for breast cancer risk, regardless of the menopausal status and essentially independent of MRI-FGT.
The BPE kinetic variables used in the present study were Eearly, Epeak, and SER [19]. We found a significant correlation of Eearly and Epeak with the visually assessed degree of BPE, classified according to BIRADS [17], in a gradually increasing degree of correlation. This suggests that Eearly and Epeak, might be good candidates for quantitative assessment of BPE.
The main limitation of the present study was its retrospective nature that led to a limited number of included patients in the premenopausal group. We were dependent on patient medical records for items such as the date of the most recent menstrual period. Because we wanted to avoid the influence of hormonal status, we limited the study to only include images from premenopausal women who were imaged during the first or second week of their menstrual cycle, not in a perimenopausal period, not taking chemotherapy and/or hormonal therapy, or within the 6 months before imaging. Another limitation was that the MR images were reviewed by consensus between 2 radiologists, precluding evaluation of interobserver variability. Nevertheless, all the images were acquired using the same MRI scanner and protocol settings and the readers were blinded to the patient information.
In conclusion, this study provides quantitative analysis of the BPE of normal FGT from Thai women in correlation with the amount identified on breast MR images. Neither visual nor quantitative measurement of BPE kinetics showed any relationship with the amount of the FGT. However, the effects of menopause on normal FGT and its enhancement were demonstrated. Menopause is significantly correlated with a lower amount of FGT, visual BPE, and Epeak.
We found a significantly higher degree of visual BPE and larger Epeak in premenopausal women. This information may be useful to alert radiologists to consider these variables carefully in breast MRI in premenopausal women.