| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
Original Research |
1 Present address: Division of Diagnostic Imaging, Breast Center, Kameda Medical
Center, 929 Higashi-cho, Kamogawa, Chiba, Japan 296-8602.
2 Department of Radiology, The Jikei University School of Medicine, 3-25-8
Nishi-Shimbashi, Minato-ku, Tokyo 105-8461, Japan.
Received June 10, 2005;
accepted after revision August 22, 2005.
Address correspondence to M. Tozaki
(e-tozaki{at}keh.biglobe.ne.jp).
Abstract
 Top Abstract IntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
MATERIALS AND METHODS. Retrospective review was performed of 30 consecutive lesions showing non-masslike enhancement. MRI was performed on a 1.5-T scanner using the volumetric interpolated breath-hold examination sequence. The distribution patterns were classified into three categories: single quadrant/solitary lesion (linear), single quadrant/grouped lesion (focal, regional, segmental), and multiquadrant lesion (multiple regions, diffuse). The presence of a ductal pattern was assessed in the enhancing lesions after the tumor distribution had been decided. In addition to the BI-RADS MRI descriptors, the presence of clustered ring enhancement was also assessed in heterogeneous enhancing lesions.
RESULTS. The most frequent morphologic finding among the benign lesions was a linear pattern (50%) followed by homogeneous internal enhancement (42%), whereas a segmental pattern (56%) (p = 0.003), heterogeneous internal enhancement (44%), and clustered ring enhancement (44%) (p = 0.01) were the most frequent findings in malignant lesions. The features with the highest positive predictive value for carcinoma were a segmental distribution (100%), a clustered ring enhancement (100%), and a clumped internal architecture (88%). Using this interpretation model, the positive predictive value for carcinoma was 94%.
CONCLUSION. A combination of BI-RADS MRI descriptors and clustered ring enhancement criteria is useful for the differential diagnosis of lesions showing non-masslike enhancement.
Keywords: BI-RADS MRI descriptors • breast • breast cancer • high spatial resolution • MRI
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
The recently published BI-RADS [13] included the first edition of MRI lexicon in relation to breast imaging. Many cases of DCIS are reported to exhibit non-masslike enhancement [14]. In regard to the patterns of MRI findings among lesions showing non-masslike enhancement, segmental or clumped linear and ductal enhancement were reported to be more frequent in DCIS than in benign lesions [14, 15]. However, a standardized classification scheme for the interpretation of lesions showing non-masslike enhancement does not exist.
The goal of the present study was to assess a new interpretation model based on BI-RADS MRI descriptors and high-spatial-resolution MR images.
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
For the study reported here, a retrospective review of 30 consecutive histopathologically diagnosed breast lesions showing non-masslike enhancement was performed. The patients ranged in age from 30 to 72 years, with a mean age of 49 years.
The histopathologic diagnoses, established by core biopsy (n = 25), excisional biopsy (n = 3), or lumpectomy (n = 2), were 18 cases of carcinoma and 12 benign lesions. Thirteen and five of the breast cancer patients underwent mastectomy and lumpectomy, respectively. The histologic types of carcinoma included invasive ductal carcinoma with DCIS (n = 2) and DCIS (n = 16), and the 12 benign lesions included atypical ductal hyperplasia (n = 2), intraductal papilloma (n = 1), and fibrocystic disease (n =9).
MRI
MRI was performed using a 1.5-T scanner (Symphony, Siemens Medical
Solutions) with maximum gradient field strength of 30 mT/m. All patients were
examined in a prone position using a double CP (circular polarized) breast
array coil. A transverse, fat-suppressed, T2-weighted fast spin-echo sequence
was performed with the following parameters: TR/TE, 3,500/78; field of view,
20 cm; matrix size, 256 x 256; slice thickness, 5 mm with a 1-mm gap.
T2*-weighted first-pass perfusion images were obtained in the
transverse plane before, during, and after the bolus injection of 0.1 mmol
gadopentetate dimeglumine/kg at a rate of 3 mL/s followed by a 20-mL saline
flush using an automatic injector.
A 3D fat-suppressed VIBE (volumetric interpolated breath-hold examination) sequence was obtained before and 60 seconds, 100 seconds, and 4 minutes after the start of the IV administration. The MRI parameters for the VIBE sequence were as follows: 3.7/1.7; flip angle, 25°; field of view, 27 cm; matrix, 256 x 218; receiver bandwidth, 490 Hz/pixel; mean partition thickness, 1.2 mm; and time of acquisition, 35 seconds. The section thickness varied depending on the size of the breast and ranged from 1 to 1.5 mm without a gap. The affected single breast was examined on the first- and third-phase dynamic images, acquired at 60 seconds and 4 minutes, respectively, and both breasts were examined on images obtained in the second phase at 100 seconds. If incidental suspicious enhancement was detected in the contralateral breast during the second phase, additional images of both breasts were obtained immediately during the subsequent third phase. None of the patients in this study had lesions visualized as incidental enhancement in the contralateral breast.
Image Interpretation
One experienced breast radiologist evaluated all the cases; the radiologist
was unaware of any clinical information or the histopathologic diagnosis.
Figure 1 illustrates the interpretation method used in this study. First, the distribution patterns were classified into three categories: single quadrant/solitary lesion (linear), single quadrant/grouped lesion (focal, regional, segmental), and multiquadrant lesion (multiple regions, diffuse). Second, the presence of a ductal pattern was assessed in the enhancing lesions (positive or negative). Third, internal enhancement (homogeneous, heterogeneous, stippled, clumped, reticular) was evaluated. Fourth, in addition to the BI-RADS MRI descriptors, the presence of clustered ring enhancement was assessed in heterogeneous enhancing lesions (positive or negative).
|
Statistical Analysis
For analysis of group differences from dichotomous variables, the Fisher's
exact test was used. A p value of less than 0.05 was considered to
indicate a statistically significant difference.
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
|
Table 2 shows the pattern of ductal enhancement. The most frequently observed ductal pattern among the malignant lesions was a branching ductal pattern (67%), whereas the most frequently observed ductal pattern among the benign lesions was a linear ductal pattern (100%) (p = 0.03). None of the internal enhancements showed a statistically significant association with benign or malignant lesions.
|
Figure 2 illustrates the new interpretation model proposed in this report, which includes BI-RADS MRI descriptors and clustered ring enhancement. Six benign lesions and two DCIS exhibited a linear ductal pattern. The histologic diagnosis of one benign lesion with a clumped internal architecture was atypical ductal hyperplasia (Fig. 3). All lesions exhibiting a branching ductal pattern were diagnosed as DCIS. The histologic characteristics of these lesions included scattered ductal carcinoma with surrounding atrophic breast tissue (Figs. 4A, 4B, and 4C). The enhancing lines corresponded to single or several ducts. The internal architecture of the nonbranching lesions was classified as stippled, clumped, or heterogeneous. All lesions showing a stippled pattern were diagnosed as fibrocystic disease, whereas all lesions showing a clumped pattern were diagnosed as DCIS. The clumped structures histologically corresponded to clusters of multiple ducts filled with intraductal tumor cells (Figs. 5A, 5B, and 5C). These lesions were characterized by enclosed ducts with fibrous stroma. The histologic types of carcinoma with heterogeneous internal enhancement included two invasive ductal carcinomas with DCIS and six cases of DCIS. The two benign lesions with heterogeneous internal enhancement were diagnosed as fibrocystic disease. All carcinomas with heterogeneous internal enhancement also showed clustered ring enhancement (Figs. 6A, 6B, 7A, and 7B). The histologic characteristics corresponding to clustered ring enhancement were crowded ductal carcinoma in five comedo-type DCIS and three noncomedo-type DCIS. The positive predictive value for carcinoma was 94% (17/18).
|
|
|
|
|
|
|
|
|
|
|
|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
An international group of breast MRI experts supported by the American College of Radiology and the Office of Women's Health had been developing a lexicon for breast MRI [16, 17]. In these reports, the distribution patterns of non-masslike enhancement were classified as follows: foci of enhancement (dotlike), linear nonspecific, linear ductal, segmental, regional, diffuse patchy, or diffuse nonspecific. Moreover, in the BI-RADS MRI terminology, the distribution patterns of non-masslike enhancement were classified as follows: focal area, linear, ductal, segmental, regional, multiple regions, and diffuse. In some cases with ductal enhancement, however, it was difficult to distinguish this characteristic precisely from other distribution descriptors (Figs. 4A, 4B, and 4C). We previously reported an original tumor distribution of breast carcinomas visualized using coronal multiplanar reconstruction of MDCT [18, 19]. The coronal images were useful for classifying the distribution pattern of breast tumors. Using a combination of this tumor classification and the BI-RADS MRI terminology, the distribution patterns of non-masslike enhancements could be classified into three categories: single quadrant/solitary lesion (linear), single quadrant/grouped lesion (focal, regional, segmental), and multiquadrant lesion (multiple regions, diffuse). The difference between this classification and the BI-RADS MRI terminology is that "ductal enhancement" was not included in the tumor distribution. Because "duct" is an anatomic or pathologic term, we used the term "ductal pattern" as morphologic terminology. When evaluating the morphologic characteristics of ductal branching, sagittal or transverse planes were thought to be more exact than the coronal plane. Therefore, we evaluated the presence of a ductal pattern using transverse and sagittal multiplanar reformations after the tumor distribution had been decided based on the findings of coronal images. Moreover, in the BI-RADS MRI terminology, "linear" is defined as any linear enhancement that may not conform to a duct. However, we regarded "linear" as any linear enhancement regardless of the possibility that it may conform to a duct. After the distribution of the enhancing lesion had been decided, we also evaluated whether the lesion conformed to a duct.
Regarding the plane of acquisition, sagittal and transverse images are recommended because these views can be correlated with similarly positioned mediolateral and craniocaudal mammographic views [20]. Until now, the severe technical constraints of MR scanners have made it necessary to choose between temporal and spatial resolution. Dynamic breast MRI, popular in European countries, attempts to distinguish between benign and malignant lesions according to the enhancement kinetics at a high temporal resolution. Static breast MRI, popular in the United States, attempts to achieve the same goal by evaluating the morphologic patterns at a high spatial resolution. Dynamic breast MRI is usually performed axially in bilateral breasts. Static breast MRI is usually performed sagittally in a single breast. However, Orel and Schnall [20] predicted that the acquisition of both high-spatial-resolution images and high-temporal-resolution images might ultimately dominate MRI protocols for the breast. Recently, the development of new MRI protocols of the breast has enabled simultaneous acquisition of high-spatial-resolution and high-temporal-resolution images. One of the representative advanced protocols is a 3D fat-suppressed, gradient-recalled echo technique with volumetric interpolation, first described by Rofsky et al. [21]. The 3D-VIBE sequence has made it possible to evaluate the distribution and morphologic characteristics of breast tumors using near-isotropic multiplanar reformations [22, 23].
Morakkabati-Spitz et al. [15] reported that segmental or linear enhancement was the most frequent manifestation of DCIS on dynamic MRI; however, the overall positive predictive value of this sign was only moderate (34%). Thirteen cases of fibrocystic disease with segmental distribution were included in their study. However, our experience does not agree with that study's findings. In our study, all cases with segmental enhancement were diagnosed as malignant lesions. The apparently higher positive predictive value of segmental enhancement in our study as compared with that reported by Morakkabati-Spitz et al. may be related to the difference in the plane of the MR images. We assessed the morphologic characteristics using the coronal, transverse, and sagittal images, and based on the assessment in these three planes, an apparently triangular region of enhancement, apex pointing toward the nipple, was defined as "segmental." In particular, we attached great importance to the coronal plane because the distribution of the ductal system can be precisely evaluated on coronal images. Ultimately, we suggest that the ductal pattern and the internal architecture should be evaluated in lesions with a segmental pattern.
In this study, the most frequently observed ductal pattern in the malignant lesions was a branching-ductal pattern (67%), whereas the most frequently observed ductal pattern in the benign lesions was a linear-ductal pattern (100%). We found that a branching-ductal pattern was an indicator of malignancy despite the internal architecture. The histologic characteristics of these lesions included scattered ductal carcinoma with surrounding atrophic breast tissue. The lines of enhancement corresponded to single or several ducts (Figs. 4A, 4B, and 4C). However, the frequency of malignant lesions in the linear-ductal pattern was 25% (2/8). This frequency is similar to those of previous reports [15, 24].
With respect to the patterns of internal enhancement, the features with the highest positive predictive value for carcinoma were clustered ring enhancement (100%) and clumped internal architecture (88%). Liberman et al. [14, 24] reported that the feature with the highest positive predictive value for malignancy was clumped internal enhancement in lesions showing non-masslike enhancement. Histologically, the clumped structures corresponded to clusters of multiple ducts enclosed with fibrous stroma (Figs. 5A, 5B, and 5C). In contrast, the histologic characteristics corresponding to clustered ring enhancement were crowded intraductal carcinoma. Thus, clustered ring enhancement reflects the enhancement of ductal carcinoma and the periductal space. Comedo-type DCIS was more frequent than noncomedo-type DCIS. This differentiation between benign and malignant lesions has not been previously reported. Our results indicate that high-spatial-resolution images could depict DCIS occupying a localized region as clustered ring enhancement. On the basis of these results, we find that the presence of clustered ring enhancement may indicate malignancy in a manner similar to the presence of clumped internal architecture.
To our knowledge, this is the first report of an interpretation model combining BI-RADS MRI descriptors and clustered ring enhancement criteria. However, our study has several limitations. First, only a relatively small group of 30 lesions was evaluated. Second, preparation of models for interpretation, focusing specific attention on morphologic characteristics, was investigated in this study. However, kinetic patterns were not evaluated. Although the description of BI-RADS MRI focuses on the morphology of the lesions, kinetic assessment provides crucial information regarding the nature of the lesions [13]. It would appear to be necessary to add kinetic information to the morphologic characteristics and to further evaluate and update diagnostic criteria in the future.
In conclusion, the features with the highest positive predictive value for the diagnosis of malignancy were segmental distribution, clustered ring enhancement, and a clumped internal architecture in lesions showing non-masslike enhancement. Using this interpretation model, the positive predictive value for carcinoma was 94% (17/18). A combination of BI-RADS MRI descriptors and clustered ring enhancement criteria is useful for the differential diagnosis of lesions showing non-masslike enhancement.
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
CiteULike 
Complore 
Connotea 
Del.icio.us 
Digg 
Reddit 
Technorati What's this?
This article has been cited by other articles:
|
|
 |
|
  E. J. Stern Seek and You Shall Find Am. J. Roentgenol., August 1, 2006; 187(2): 265 - 265. [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
