MBC is marked mainly by histological and molecular diversity, as well as aggressiveness and chemoresistance, which related to epithelial-mesenchymal transition and cancer stem cell characteristics [2, 4, 6, 15]. Although few image features of MBC were reported, MBC still deserves our attention and further research for its high mortality. Therefore, we try to evaluate the clinical value of conventional ultrasound and clinicopathology in diagnosing MBC, as well as their capacity to distinguish MBC from IDC.
There was a younger median age for patients with MBC (51 years old) than those with IDC (52 years old) in our study. However, only 11 (55%) patients in the MBC population were older than 50 years old, contrary to previous studies which showed that most MBC patients were older than 50 years old[6, 8, 9, 16]. It might be related to the small sample size of MBC in our single-center study. Li et al. demonstrated that patients over 58 years old show unfavorable survival with poor OS and cancer-specific survival[17]. Older MBC patients generally had a higher-risk histological subtype, which resulting in lower survival[18]. As revealed in our study and other reports, patients with MBC had higher tumor grade, higher Ki67 proliferation index, higher HR negativity, higher probability of TNBC, but similar expression of HER2, compared with IDC [6, 16, 19, 20]. Elimimian et al. compared different rare TNBCs, they found that triple-negative MBC had the lowest ER and PR expression (19.0% and 14.1%, respectively), but the highest HER2 expression (4.4%); a high level of expression of the HER2 oncogene might explain why MBC was associated with an increased risk of metastasis to the lungs and bones [21]. A study including 75 patients diagnosed with MBC found that most of them expressed basal markers, epidermal growth factor receptor, epithelial-mesenchymal transition marker vimentin, and the stem cell marker CD44; disease-free survival was significantly correlated with basal-like marker expression[15]. Although the mechanism is not fully understood, the aggressiveness and adverse outcomes of MBC may be explained by these clinicopathologic features.
Several studies revealed that MBC had a worse prognosis, such as a poor response to NST, a higher recurrence rate, and a lower OS rate[2–6]. The pCR rate was 18.2% among MBC cases in our study, which is not in line with prior literature estimate of 11%[3]. We followed 20 patients with MBC over the last decade, postoperative local recurrence occurred in two cases, distant metastases occurred in six cases, and death due to MBC occurred in five cases. The 5-year OS rate was 65%, which is consistent with 63.1% reported by Elizabeth B et al[21]. Moreno et al.indicated that the 5-year OS rates for MBC were lower than for TNBC and other breast cancer across all clinical stages: stage I disease were 85%, 87% and 91%; stage II were 73%, 77%, and 87%; and stage III were 43%, 53%, and 75% [4].Elizabeth B et al. concluded that the poor prognosis of MBC might be associated with advanced stage, lung metastasis, older age, and not receiving chemotherapy or radiation therapy[21].
MBC has no specific imaging appearances, which can present with imaging features similar to malignant tumors and probably even benign masses. Findings from multimodal imaging of five MBC cases, including ultrasound, mammography, magnetic resonance imaging(MRI), and positron emission tomography–computed tomography, indicated that as compared to IDC, MBC had more benign imaging features, such as round or oval shapes, circumscribed margins and fewer axillary lymph node metastases[22]. An analysis of 19 MBC patients indicated that mammography and ultrasound showed benign image features, while MRI with T2-weighted imaging manifested specific features of signal intensity based on apparent diffusion coefficient and time-intensity curves analysis[23]. Another article with 65 MBC patients came to similar conclusions, MBC demonstrated benign or moderately malignant features on imaging methods, MRI might comparatively provide more malignant features than other imaging modalities [24]. MBC can be diagnosed by T2-weighted imaging with high signal intensity due to cystic or necrotic components, although it must be differentiated from mucinous carcinoma and, less frequently, necrotic IDC[22, 25]. Mammographically, the most frequent presentations of MBC were round/ oval shape, high density masses, and calcifications were not a frequent finding. But MBC with osseous differentiation could present as densely calcified lesion on mammography[16, 26]. Günhan-Bilgen et al. analysed eight MBC cases, all masses were highly dense, none of the masses showed prominent spiculations, microcalcifications or architectural distortion on mammography[8]. One study of 67 women with MBC found that the most common mammographic findings were oval or round (52.5%), not-circumscribed (71.6%), and lacked calcification (68.7%)[27]. However, when MBC had an IDC component, masses might represent an irregular or spiculated margin[28].
On sonographic examination, the most characteristic manifestations of MBC were masses with complex cystic and solid echo pattern and posterior acoustic enhancement[28–31]. In our study, there were 10 (50.0%) MBC lesions had complex cystic and solid component and 19 (95.0%) MBC manifested as posterior acoustic enhancement, which was consistent with previous studies. The cystic areas might be caused by necrosis, hemorrhage, or cystic degeneration[9]. In our cohort, 45.0% of the MBCs were larger than 5cm in diameter, while only 1.5% of IDCs were larger than 5cm. It might indicate that the growth speed of MBC is significantly faster than that of IDC. Unlike IDC, which was mostly irregular in shape(98.5%), there were 45.0% of MBCs had regular shape, generally oval or round. The commonest margin performances of MBC were circumscribed margin(40.0%) or microlobulated margin(45.0%), none presented with spiculated or angular margin which usually seen in IDC(33.5% and 66.5% respectively). Our current results were basically consistent with previous studies, the shapes of MBC might be oval, round, or occasionally irregular, and the margins might be circumscribed or non-circumscribed, the latter was most commonly microlobulated[23, 31]. In order to distinguish MBC from IDC, we used MLP to analyse the importance of all these sonographic and clinicopathologic features. We found that masses with posterior acoustic enhancement, circumscribed margin and big size might be more likely to be MBC, while masses with spiculated or angular margin and hypoecho might possible be IDC.
The small sample size of our study is the major limitation of this retrospective case series in a single-center study. This study could only involve 20 patients due to MBC's rare occurrence. Future evaluation of our results will require external validation with a larger sample size.