Terminology and classification of non-mass abnormalities on breast US

The incidence of breast cancer, which is the most common cancer in women, is increasing worldwide. Breast tumors are often discovered as breast masses on imaging; however, when accompanied by bloody nipple discharge or detected as calcification on mammography performed for screening purposes, they do not appear as a mass, and in many cases, they appear as non-mass abnormalities of the breast. Non-mass lesions are defined in the ACR BI-RADS breast magnetic resonance imaging (MRI) lexicon [17], but not in ultrasound imaging.

In Japan, JABTS and JSUM previously used the word “lesion,” which may be more popular worldwide, but “lesion” indicates disease. However, many non-mass lesions include aberrations from the normal that are not diseases. Therefore, JABTS and JSUM decided to use the term non-mass “abnormalities” based on the changes on images. After two revisions, the current terms were mainly categorized as hypoechoic areas in the mammary gland, ductal abnormalities, architectural distortion, multiple small cysts, and echogenic foci without a hypoechoic area [15]. Histological types presenting as non-mass abnormalities in the breast comprise a wide range of malignancies, including ductal carcinoma in situ, invasive lobular carcinomas, and invasive ductal carcinomas, and various benign lesions, such as fibrocystic change, sclerosing adenosis, and intraductal papilloma [18]. Thus, the JABTS taxonomy is too complex to accommodate the diagnosis of its diverse histology.

Many studies have classified non-mass lesions as ductal abnormalities, architectural distortions, or calcifications on US. It has also been reported that the diagnostic ability for non-mass abnormalities is not as satisfactory as that for masses, with fair sensitivity and poor specificity [1,2,3,4,5]. This is attributed to the overlap of a few US findings between low-grade ductal carcinomas in situ and benign diseases. The addition of color Doppler US or elastography to classical B-mode imaging reportedly improves the diagnostic specificity [19,20,21,22]. In the JABTS guidelines, calcification, vascularity, and elasticity are considered additional findings that supplement the B-mode findings [15].

B-mode US, color Doppler US, and elastography of various non-mass abnormalities of the breast

Figure 5 shows a segmental hypoechoic area in the mammary gland with echogenic foci on B-mode US (Fig. 5a). Category 5 was assigned to the lesion. It demonstrated vascularity of (3 +) on color Doppler US (Fig. 5b), and an elasticity score of 3 was assigned on elastography (Fig. 5c). After surgery, the tumor was histologically diagnosed as ductal carcinoma in situ. Figure 6 shows a focal hypoechoic area in the mammary gland on B-mode US (Fig. 6a). Category 4 was assigned to the lesion. It demonstrated vascularity of (3 +) on color Doppler US (Fig. 6b), and an elasticity score of 3 was assigned on elastography (Fig. 6c). After surgery, the tumor was histologically diagnosed as invasive lobular carcinoma. Figure 7 shows a regional hypoechoic area in the mammary gland on B-mode US (Fig. 7a). Category 3b was assigned to the lesion. No vascularity was observed on color Doppler US (Fig. 7b), and an elasticity score of 2 was assigned on elastography (Fig. 7-c). On core needle biopsy, the lesion was histologically diagnosed as fibrocystic change.

Fig. 5

B-mode ultrasonography (US), color Doppler US, and elastography in a case of ductal carcinoma in situ. a A segmental hypoechoic area in the mammary gland with echogenic foci is seen on B-mode US. b It demonstrated vascularity of (3 +) on color Doppler US. c An elasticity score of 3 was assigned on elastography

Fig. 6

B-mode ultrasonography (US), color Doppler US, and elastography in a case of invasive lobular carcinoma. a A focal hypoechoic area in the mammary gland was seen on B-mode US. b It demonstrated vascularity of (2 +) on color Doppler US. c An elasticity score of 3 was assigned on elastography

Fig. 7

B-mode ultrasonography (US), color Doppler US, and elastography in a case of fibrocystic disease. a A regional hypoechoic area in the mammary gland was seen on B-mode US. b It demonstrated no vascularity on color Doppler US. c An elasticity score of 2 was assigned on elastography

Diagnostic ability of color Doppler US and elastography adjunct to B-mode US

The addition of color Doppler US and SE to B-mode imaging in this study improved sensitivity without compromising specificity because in the case of invasive breast cancers, tumor cells infiltrating the stroma are accompanied by vascular hyperplasia, and in the case of noninvasive ductal carcinomas, the higher the degree of malignancy, the higher the vascular hyperplasia in the stroma, and the greater the vascularity (Figs. 5, 6) [23, 24]. Thus, color Doppler US is incomparable for the diagnosis of breast cancer with improved sensitivity. In contrast, benign lesions in the mammary gland are relatively soft and exhibit low scores on elastography, making it superior for diagnosing benign lesions, thus improving the specificity (Fig. 7) [3, 25]. Color Doppler and elastography work in a complementary manner, resulting in improved sensitivity while maintaining specificity.

For breast masses, the addition of color Doppler US and elastography to B-mode US has considerably improved specificity while maintaining sensitivity [14]. In this study, the specificity did not improve as much for non-mass abnormalities as for masses. This is because fibrocystic changes, which are benign lesions, often show a decrease in the elasticity, whereas noninvasive ductal carcinoma, which is a malignant lesion, often maintains its elasticity. Therefore, the addition of SE to B-mode and color Doppler US did not improve specificity and decreased sensitivity.

Vascularity of blood flow into an internal solid component of the ducts was not a statistically significant finding for distinguishing between 35 benign and 19 malignant lesions. Benign internal solid component of the duct included at least eight cases of intraductal papilloma, which has a relatively thick vascular stalk and exhibits abundant vascularity for its size. Therefore, we presume that no significant difference was present in the vascularity between the benign and malignant internal solid component of the ducts in this study.

B-mode features and benign/malignant differential diagnosis

To differentiate between benign and malignant lesions, the JABTS guidelines assign categories based on B-mode distribution and spread of non-mass abnormalities, considering the presence of echogenic foci. However, statistically significant differences in the odds ratios were observed only in the presence of echogenic foci in the hypoechoic areas of the mammary gland and the configuration of the internal solid component in the mammary ducts. No difference was observed between benign and malignant non-mass abnormalities with respect to distribution and spread on B-mode US. The sensitivity of categories 3b, 4, and 5, which are indicated for tissue biopsy, as malignant tumors was 83.7%, and the specificity and accuracy were 32.8% and 59.5%, respectively, thereby indicating poor diagnostic ability. This is because benign lesions, such as epithelial hyperplasia, adenosis, and fibrosis, were classified as category 3b or higher, and the current category classification may be inappropriate for excluding benign lesions and indications for biopsies.

Color Doppler US and elastography features and benign/malignant differential diagnosis

As described above, the diagnostic ability of B-mode ultrasound is insufficient for non-mass abnormalities, and we attempted to improve its diagnostic ability by incorporating color Doppler US and SE. The odds ratio of vascularity (2 +) and (3 +) versus (1 +) and (–) was 7.84, and that of penetrating blood flow was 7.46, which were considered significant findings for malignancy. Namely, vascularity evaluation using color Doppler US in this study was excellent for determining benign or malignant diseases, and it improved the sensitivity from 84.6%83.7% to 93.1%. It is relatively difficult to determine whether non-mass abnormalities are abnormalities or normal variations using B-mode images. In this sense, the increase in sensitivity resulting from adding color Doppler US to B-mode US in this study is considered meaningful. In contrast, the odds ratio of elasticity scores 3, 4, and 5 versus elasticity scores 1 and 2 was 4.49. The elastography findings in this study often overlapped between benign and malignant lesions, and the discrimination between benign and malignant lesions was inferior to that achieved using color Doppler US. Although no benign lesions with an elasticity score of 5 were detected, benign lesions with an elasticity score of 3 or 4 accounted for 50% of all lesions. Hence, adding elastography did not significantly improve the specificity.

Distribution of vascularity by age between benign and malignant non-mass lesions

Patients with benign non-mass abnormalities aged under 55 years tended to have more lesions with high vascularity than those aged 55 years and older. We previously reported that vascularity was significantly higher in younger individuals under 50 years of age than in those over 50 years of age upon performing US examinations for benign breast masses using B-mode plus color Doppler US [10]. Even in benign breast diseases presenting with non-mass abnormalities, recognizing that vascularity of lesions can be higher in younger individuals suggests the potential for improving specificity.

Future scope considering US of non-mass abnormalities of the breast.

Regarding B-mode US, no statistically significant difference was observed in the distribution and spread of lesions, which is thought to be the basic characteristic feature for differentiating between benign and malignant lesions. The accuracy rate was 59.5%, indicating that B-mode US alone could not differentiate benign and malignant tumors. Color Doppler US and SE were then added to B-mode US; although the sensitivity improved, the specificity did not change. Moreover, although the number of missed breast cancers could be reduced, they could not contribute to the avoidance of unnecessary biopsies.

Non-mass abnormalities of the mammary gland originally included lesions of various histological types, regardless of whether they were benign or malignant. Therefore, to diagnose non-mass abnormalities of the breast, it is necessary to comprehensively evaluate not only US but also clinical, mammography, and MRI findings to determine whether the tumor is benign or whether biopsy should be performed.

Limitations

This study had several limitations. First, as mentioned earlier, non-mass abnormalities in the breast include a wide variety of histological types; therefore, an analysis for each subtype should be performed. Second, owing to their diversity, there are limitations to determining whether non-mass abnormalities are benign or malignant using uniform findings and criteria. Third, the cutoffs for vascularity on color Doppler US and the elasticity score on elastography have not been validated for determining whether a tumor is benign or malignant. Changing the cutoff automatically changes the sensitivity and specificity; therefore, setting an optimal cutoff is necessary. Finally, we considered the lesions to be benign during the 1-year follow-up. BI-RADS recommends 2-year follow-up, and we consider the 1-year follow-up to be insufficient. However, in actual clinical practice, there are few cases in which benign lesions can be followed up for 2 years. Therefore, if only those who have undergone strict follow-up observation for 2 years are considered eligible, many patients will be excluded.