If you don't remember your password, you can reset it by entering your email address and clicking the Reset Password button. You will then receive an email that contains a secure link for resetting your password
If the address matches a valid account an email will be sent to __email__ with instructions for resetting your password
The Japan Association of Breast and Thyroid Sonology (JABTS) proposed, in 2003, a conceptual classification system for non-mass abnormalities to be applied in addition to the conventional concept of masses, to facilitate detecting ductal carcinoma in situ (DCIS) lesions. The aim of this study was to confirm the utility of this system and to clarify the distribution of these findings in DCIS lesions. Data on 705 surgically treated DCIS lesions from 16 institutions in Japan were retrospectively reviewed. All 705 DCIS lesions could be classified according to the JABTS classification system. The most frequent findings were hypo-echoic areas in the mammary gland (48.6%), followed by solid masses (28.0%) and duct abnormalities (10.2%) or mixed masses (8.1%). Distortion (1.3%), clustered microcysts (1.4%) and echogenic foci without a hypo-echoic area (2.5%) were uncommon. These results suggest that the concept of non-mass abnormalities is useful in detecting DCIS lesions.
With advances in device-related technologies, breast ultrasound has become a more important modality in the diagnosis of breast cancer. Ultrasound is useful for the diagnosis not only of invasive breast cancer, but also of DCIS. However, as DCIS can appear in a wide variety of forms, an understanding of the variations is important for ultrasound detection and diagnosis. The ultrasound features of DCIS have been described in the literature. In general, the ultrasound findings of DCIS are classified into masses, ductal change (ductal abnormalities), calcification alone and architectural distortion (
The Japan Association of Breast and Thyroid Sonology (JABTS) published guidelines for breast ultrasound in 2003 and proposed a conceptual classification system that incorporates both non-mass abnormalities (non-mass image-forming lesions) and the conventional concept of masses, mainly to facilitate DCIS detection (
). The non-mass abnormalities are of five subtypes, including abnormalities of the ducts, hypo-echoic areas in the mammary gland and architectural distortion. These descriptions refer to the variations in ultrasound images of DCIS, in addition to masses. As this JABTS classification system has been widely used in Japan, we conducted a multicenter, retrospective observational study (JABTS BC-02 study) to confirm the utility of the JABTS classification system for DCIS image classification and to clarify the distribution of these findings in DCIS lesions.
This was a retrospective observational multicenter study. Clinical data and static ultrasound images of DCIS patients who underwent surgery from January 2008 to December 2012 were collected. Patient DCIS data were collected even for cases in which DCIS lesions were not detectable by ultrasound. Collected images were evaluated by a Centralized Image Interpretation Committee.
The JABTS classifies breast lesions primarily as masses and non-mass abnormalities. Non-mass abnormalities are defined as lesions that are not recognized as a mass.
Masses are further classified into the subtypes cystic (simple cyst), mixed and solid (Fig. 1). A mixed mass contains both cystic and solid components.
Duct abnormalities usually refer to ductal dilation accompanied by filling of the duct with a solid component (Fig. 2).
Hypo-echoic areas in the mammary gland differ from the surrounding tissue and cannot be recognized as masses (Fig. 3). Because typical DCIS lesions that appear as hypo-echoic areas possess 3-D segmental or focal distributions, where focal means less than a breast segment, we consider the distribution of lesions to be the most important difference between hypo-echoic areas and masses. In contrast, masses usually do not possess such segmental or focal distributions. As with the concept of non-mass enhancements on magnetic resonance imaging (MRI), lesion distribution is the most important factor to consider when applying the concept of hypo-echoic areas in the mammary gland on ultrasound.
Architectural distortion is a term describing a lesion that distorts the breast tissue, but without mass formation (Fig. 4). Architectural distortion is often difficult to visualize on static images, but can easily be recognized at the time of hand-held ultrasound examination because the image moves in real time.
Multiple small cysts are defined as multiple tiny or small cysts existing in the mammary gland. The distribution of such small cysts is important, in that a diffuse distribution suggests the lesion is benign, whereas a clustered or segmental distribution raises a small possibility of malignancy. The term clustered microcysts is used when multiple small cysts are observed clustered together (Fig. 5).
Echogenic foci without a hypo-echoic area are lesions in which only microcalcifications are visible (Fig. 6). Because not all echogenic foci on ultrasound are ultimately confirmed to be microcalcifications on mammography or by histology, we herein employ the term echogenic foci, which is an ultrasound finding, rather than microcalcifications, which is a mammographic or histologic finding. Note that although DCIS with echogenic foci is generally accompanied by an obvious hypo-echoic area, this type of finding is not.
Centralized Image Interpretation Committee
The Centralized Image Interpretation Committee comprised 14 breast ultrasound specialists in Japan, including 2 radiologists, 9 breast surgeons and 3 ultrasonographers. These specialists are also members of the Terminology and Diagnostic Criteria Committee of the JABTS. B-Mode images of each lesion were first analyzed by at least two members of the committee. Lesion type (mass or non-mass abnormality), subtypes, and other findings (for masses: shape, depth width ratio and margin; for non-masses: distribution and echogenic foci) were evaluated. When interpretation was difficult, the images were evaluated with the participation of all members. When a lesion was found to contain more than one subtype, the committee determined the single dominant type. When a mass and a non-mass abnormality were both present, the subtype of the mass, rather than that of the non-mass abnormality, was determined to be the dominant type. When multiple subtypes of non-mass abnormalities co-existed, the more extensive or prominent type was determined to be the dominant type.
Institutional review board
The institutional review board or ethics committee at each hospital approved this retrospective observational study. Informed consent was not required.
Data collection and analysis
Data collection and analysis were performed by the Clinical Research, Innovation, and Education Center of Tohoku University Hospital. This study is registered as No. UMIN000007604.
From January 2008 to December 2012, 809 DCIS lesions were surgically treated at 16 institutions, and 705 (87%) of these lesions (691 women) were observed by ultrasound. The mean age of these 691 women was 56.6 ± 12.7 y (mean ± standard deviation). This set of 705 DCIS lesions was evaluated in this study.
Mammography or ultrasound screening was the most frequent opportunity (57%) for DCIS detection. Patients had presented with symptoms such as masses and nipple discharge in 24% of the lesions. Eighteen percent were discovered incidentally, for example, at a follow-up examination or on further examination for various cancers (Table 1).
Table 1Opportunities for detection of ductal carcinoma in situ
In 17 (2%) of the 705 DCIS lesions, masses and non-mass abnormalities co-existed. In 260 (37%) lesions, only masses were observed and only non-mass abnormalities were observed in 428 (61%) lesions. Therefore, masses were observed in 277 (39%) lesions in total. The corresponding subtype frequencies are listed in Table 2. Counting was performed using two different methods: by considering the frequencies of all observed subtypes, and by considering only the frequency of the dominant type. Two or more subtypes were observed in 61 lesions (including 17 lesions in which masses and non-mass abnormalities co-existed). Two subtypes were observed in 59 lesions, and three subtypes were observed in 2 lesions. The resulting cumulative number of observed subtyped lesions found in the 705 DCIS lesions was 768. In the 17 lesions in which masses and non-mass abnormalities co-existed, the subtype of the mass was used as the dominant type. In 44 lesions in which two or more non-mass subtypes were observed, the Centralized Image Interpretation Committee determined the dominant type (Fig. 7). With both counting methods, hypo-echoic areas in the mammary gland were found to be the most frequent lesion, accounting for approximately 50% of DCIS lesions, followed by solid masses and abnormalities of the ducts or mixed masses as the second and third most frequent lesions. Architectural distortion, multiple small cysts and echogenic foci without a hypo-echoic area were rare.
Table 2Frequencies of subtypes among ductal carcinoma in situ
In 17 DCIS lesions, with co-existing masses and non-mass abnormalities, the subtypes of the masses were used as the dominant type (11 solid masses and 6 mixed masses). In 44 DCIS lesions with two or more non-mass subtypes, the Centralized Image Interpretation Committee determined the dominant type.
Abnormalities of the ducts
Hypo-echoic areas in the mammary gland
Multiple small cysts (clustered microcysts)
Echogenic foci without a hypo-echoic area
DCIS = ductal carcinoma in situ.
∗ Two subtypes were observed in 59 DCIS lesions, and three subtypes were observed in 2 DCIS lesions.
† In 17 DCIS lesions, with co-existing masses and non-mass abnormalities, the subtypes of the masses were used as the dominant type (11 solid masses and 6 mixed masses). In 44 DCIS lesions with two or more non-mass subtypes, the Centralized Image Interpretation Committee determined the dominant type.
All lesions that appeared as “multiple small cysts” also appeared as “clustered microcysts.”
Of the 44 non-mass abnormalities in which multiple subtypes of non-mass lesions were observed, only 1 lesion contained three subtypes: abnormality of the duct, hypo-echoic areas in the mammary gland and multiple small cysts. Among the remaining 43 non-mass abnormalities, the most frequent finding was hypo-echoic areas in the mammary gland (38 lesions, 88%). Furthermore, the second most frequent finding was abnormalities of the ducts (24 lesions, 56%). In 21 of these 43 non-mass abnormalities, hypo-echoic areas in the mammary gland and abnormalities of the ducts co-existed (Table 3).
Table 3Non-mass abnormalities with two subtypes of non-mass lesions (n = 43)
AD = abnormalities of the ducts; HA = hypo-echoic areas in the mammary gland; Dis = Architectural distortion; MSCs = multiple small cysts; EFs = echogenic foci without a hypo-echoic area.
Of the total 705 lesions, 26.4% were comedo type. This type displayed a relatively high rate of hypo-echoic areas in the mammary gland, whereas the rate of masses was low. Echogenic foci were observed in 38.4% of the 705 lesions, and were more frequent in lesions that appeared as abnormalities of the ducts or hypo-echoic areas in the mammary gland, than in lesions that appeared as masses (Table 4).
Table 4Relationship between subtypes of ultrasound images and pathology/echogenic foci
This was a multicenter study in which 16 institutions located across Japan participated. Furthermore, 14 breast ultrasound specialists analyzed the images. To our knowledge, this is the first multicenter study focusing on the ultrasound findings of DCIS. In our study, all 705 DCIS lesions could be classified according to the JABTS classification system. The most frequent findings were hypo-echoic areas in the mammary gland, followed by solid masses and abnormalities of the ducts or mixed masses, in that order, which collectively accounted for 97% of DCIS lesions. In contrast, distortion, multiple small cysts (clustered microcysts) and echogenic foci without a hypo-echoic area were uncommon findings and were associated with only 3% of DCIS lesions. Although the frequencies of these findings differed, our results indicated that DCIS appears in a wide variety of forms. Accordingly, a clear understanding of these ultrasound imaging variations is necessary to detect DCIS. In that regard, the histopathology of DCIS is useful for understanding these variations.
Ductal carcinoma in situ is considered to develop primarily at a terminal duct lobular unit (TDLU) (
). There are 15 to 20 ductal lobular systems within a breast. Studies using 3-D and computerized reconstruction have found that DCIS is mainly unicentric and extends along a complex, branching ductal lobular system in a pyramid-like shape toward the nipple (
reported that one of the histologic appearances of DCIS is due to the “unfolding” of the lobules. A rubber glove can be taken as an illustrative example, as the shape of the lobules is similar to that of a glove. If the glove is inflated like a balloon, the five fingers initially become evident; however, additional inflation gradually changes the glove shape to that of a large round balloon. This process represents the unfolding of the lobules. When this occurs in the TDLU, it is capable of forming a variety of solid or mixed DCIS masses of various sizes. We consider dilation of ducts, as well as the unfolding of the TDLU, caused by intra-ductal spreading, to be the fundamental elements of the morphologic changes in DCIS. A variety of combinations of these elements then yield various forms detectable on ultrasound images. The relationships between the histopathologic findings and ultrasound images of DCIS have not previously been studied. However, it is reasonable to speculate that when DCIS progresses mainly into the central or larger ducts and these ducts thereby enlarge, it becomes possible to visualize the lesion as abnormalities of the ducts. On the other hand, when it progresses mainly into the peripheral ducts and TDLU, detection as a hypo-echoic area in the mammary gland becomes possible. When DCIS progresses in the unfolding process mainly without extensive intra-ductal spreading, it may appear as a mass with a well-defined margin (Fig. 8).
The ultrasound features of DCIS have previously been described in the literature, although the classifications vary slightly among authors. Ultrasound images of DCIS have been classified into masses, ductal change, calcification alone and architectural distortion (
). The finding of a ductal change is similar to abnormalities of the ducts in the JABTS classification. Most of the lesions that we classified into hypo-echoic areas in the mammary gland were classified as masses in those studies; in addition, the most frequent classification was masses.
reported that the most common sonographic features are masses with irregular shapes, microlobulated margins, abrupt interfaces and complex echotextures. These studies were not performed under the concept of “non-mass abnormalities.” However, if these authors had recognized this concept, the results of their studies might have been different.
divided echogenic images of breast disease into “non-tumor image-forming type” (non-mass abnormalities) and “tumor image-forming type” (masses); 70 of 562 breast lesions were non-tumor image-forming type. The authors reported that DCIS accounted for 7 of 135 malignant lesions, and 2 of these 7 DCIS lesions were non-tumor image-forming type. Three recent reports (
) focused on non-mass abnormalities. Shin et al. defined a non-mass lesion as one with minimal or no mass effect, a focal heterogeneity distinguished from the adjacent normal breast parenchyma or calcifications that are not associated with a mass.
divided DCIS into mass-forming malignancies and non-mass-forming malignancies. The non-mass-forming lesions were classified into microcalcifications only, duct dilation and indistinct hypo-echoic areas. The frequencies of these non-mass abnormalities were reported to range from 11% to 42% in DCIS. In our study, approximately 60% of DCIS lesions were recognized as non-mass abnormalities. These results suggest the concept of non-mass abnormalities, along with that of masses, to be useful for describing DCIS.
On MRI, there are three types of enhancing lesions: foci, masses and non-mass enhancement. The last is defined as enhancement of an area that is not a mass, and is classified according to the distribution of the enhancement as focal area, linear, segmental, multiple regions or diffuse. A segmental distribution refers to a triangular or cone-shaped area with the apex at the nipple (
). According to our results, 61% of DCIS lesions appeared as non-mass abnormalities, whereas 39% appeared as masses, on ultrasound. This result is similar to those obtained in MRI studies. As with MRI, incorporating the concept of distribution into the ultrasound image classification is important. Although the distribution of lesions is easily recognized during real-time scanning, it is difficult to illustrate this distribution accurately to others using a static image. Therefore, examiners must accurately describe the distribution in their clinical reports. However, it is probably easier to show the distribution by using an automated whole-breast scan.
Approximately 60% of DCIS lesions appear as non-mass abnormalities. Among the subtypes of non-mass abnormalities diagnosed as DCIS, hypo-echoic areas in the mammary gland and abnormalities of the ducts were found to be very frequent. We conclude that the concept of non-mass abnormalities is as useful as that of masses for detecting DCIS lesions.
We are deeply grateful to Dr. Eisuke Hukuma and Ms. Miki Yamaguchi (Kameda Medical Center), Professor Ryuzou Sekiguchi (Tochigi Cancer Center), Dr. Miki Yamaguchi (Japan Community Health Care Organization Kurume General Hospital), Dr. Youichi Ishibe (Mizushima Kyodo Hospital) and Dr. Etsuo Takada (Dokkyo Medical University) for the collection of data. We extend our special thanks to Dr. Toshikazu Ito (Rinku General Medical Center Izumisano Municipal Hospital), Ms. Nobue Kawachi (St. Luke's International Hospital), Dr. Kazutaka Nakashima (Kawasaki Medical School), Professor Kiyoka Omoto (Jichi Medical University), Mr. Masahiko Tsuruoka (Moriya Keiyu Hospital) and Mr. Norikazu Obane (Sumitomo Hospital) for their work on the Centralized Image Interpretation Committee. We gratefully acknowledge the work of past and present members of the Clinical Research, Innovation, and Education Center of Tohoku University Hospital.
Ductal carcinoma in situ: Terminology, classification, and natural history.