AME of the breast is rare, especially malignant AME. Malignant AME generally occurs in 50 or 60-year-old women. Age is an important factor in prognosis, and young age at onset is a protective factor for recurrence, metastasis and death (DFS, hazards radio, 1.063; OS, hazards radio, 1.082). These lesions were mostly unilateral without significant differences between the left and right sides, and site is independent of both DFS and OS (P > 0.05 for both). Patients with a malignant AME of the breast could present with a palpable mass[3, 4]. The patient is usually preceded by a longstanding stable mass, followed by a period of rapid growth[5]. As expected, the maximum tumor diameter is closely associated with DFS and OS (P < 0.05 for both). As such, surgical intervention is called as soon as possible. Like the case in our present report, several patients are asymptomatic and merely detected by mammogram or ultrasound screening which show the relatively non-specific oval mass. It may be impossible to identify malignant AME by radiographic features alone. However, imaging techniques are non-conclusive but indispensable. Almost all of these cases presented a BI-RADS classification of 4 or higher, which means biopsies could not be omitted and reduces the occurrence of missed diagnoses simultaneously [6].
In our present case, the result of preoperative needle-biopsy showed an intraductal papilloma, while the final postoperative pathological examination demonstrated that the resected mass was a malignant AME, which extremely misleading our treatment. The study revealed that these tumors are morphological polymorphism and classified into lobulated, papillary, tubular and mixed patterns[5]. Papillary patterns are easily confused with intraductal papilloma, and some scholars have even suggested that AME is a variant of intraductal papilloma[7]. The diagnosis of intraductal papilloma is supported if myoepithelial cells are seen only in the lateral aspect of the papillary region without significant proliferation of nest-like, small nodular cells. In contrast, if the myoepithelial cells are significantly and diffusely proliferating, the lesion is more likely to be a malignant AME[8, 9]. As for intraductal papillary carcinoma, it is easy to differentiate due to the absence of myoepithelium. These lesions are quite heterogeneous with variable components on pathology, so limited tissue sampling could potentially lead to misdiagnosis[10]. Never can we ignore the false negative of biopsy, and we should emphasize radiology-pathology concordant. In cases of ultrasound abnormalities (BI-RADS 4 or higher), extended resection could reduce incidence of missed diagnoses.
The diagnosis of malignant AME generally depends on immunohistochemistry after complete resection. The conclusions of Moritani et al., combining high molecular weight cytokeratin and any one of α-SMA, calponin, and p63 would be a good panel for the diagnosis of AME[11]. In our case, the tumor was positive for α-actin, calponin, and P63, and focally positive for CK5/6, consistent with related reports. The diagnosis of malignant AME should base on the diagnosis of AME, whose malignant component often shows an infiltrative growth pattern, marked cytological atypia, a high mitotic rate, significant necrosis, and a high Ki-67[5, 8, 9].
Due to this pathology is extremely rare, no guidelines are available for its therapeutic approach. After univariate analyze, we found a significantly reduced risk of relapse among breast-conserving patients, as compared with patients with biopsy only. Complete surgical excision with adequate margins is the treatment of choice that minimize the risk of recurrence and metastasis, as reported previously [5, 12-16]. However, we summarized the literatures and found that the incidence of recurrence and metastasis is still close to 50% even after complete resection, and once recurrence and metastasis occur, the mortality rate is as high as 66%. For malignant AME, it seems that complete excision alone may not be sufficient. The metastases occur mainly via the blood system, with lungs, brain and bones as the main target organs. The proportion of involved axillary lymph nodes was 4/34 and axillary node status seems to be associated with worse OS based on statistical analysis (hazard ratio, 4.443; P = 0.043). Reviewing these four patients, lymphatic metastases may be associated with excessive tumor and delayed treatment. On the one hand, as mentioned previously, the maximum tumor diameter is closely related to prognosis. On another hand, the consequence of delayed treatment is rapid growth of tumor size. Therefore, axillary staging may not be necessary in small tumor, which is consistent with the relevant literature[14, 17].
After statistical analysis, all adjuvant treatments were proven not to prolong DFS or OS (hazard ratio, 0.138; P = 0.024). We have carefully reviewed these cases, 7 cases were given chemotherapy or radiotherapy after recurrence and metastasis, and all of these patients eventually died after metastasis. We may learn from this that remedial radiotherapy or chemotherapy does not work after the occurrence of recurrence and metastasis. At the same time, among the 9 patients treated with chemotherapy and/or radiotherapy after initial surgery, there was only one case with a bad ending. The patient delayed the treatment after biopsy, and the tumor had progressed significantly when she received treatment[17]. This may indicate that these patients may benefit from appropriate postoperative adjuvant therapy[9]. Some papers suggest that the endocrinotherapy has no clear benefit, due to the ER and PR tend to be negative[18]. For the start, not all patients received adjuvant therapy. Secondly, Patients receiving adjuvant therapy might have a higher degree of malignancy themselves. This means we need more evidence to validate the value of adjuvant therapy.
In the present study, the initial lesion was found and resected completely, which was only 0.7 cm in size. Following immediate mastectomy and chemotherapy, local recurrence still occurred. Thus, it triggered further exploration on the causes of the recurrence. First of all, the malignant component of the case is myoepithelial carcinoma. Some scholars have noticed that the increased proportion of the myoepithelial component compared to the epithelial component may lead to a worse prognosis, due to heterogeneity between proportion of epithelial and myoepithelial cells can affect treatment resistance[16].This may explain the local recurrence in our case after chemotherapy. In addition, we performed lumpectomy without attention to margins status identified from frozen section analysis, due to the benign biopsy pathology. Treatment strategies for malignant AME should be made together in a multidisciplinary way to avoid over- or under-treatment of the patients, especially radiology-pathology concordant.