Benign Adenomyoepithelioma of the Breast: Imaging Characteristics

Abstract

Purpose

This study aimed to evaluate the radiological and clinical characteristics of benign adenomyoepitheliomas of the breast.

Materials and Methods

Over the last 20 years, 120 patients were histologically diagnosed with breast adenomyoepithelioma (AME) at our institution. We excluded 43 patients who were incidentally diagnosed during mastectomy for breast cancer, 28 who underwent percutaneous biopsy without further excision, and 8 who had biopsy-confirmed benign AME and were found to have another pathology after complete excision. We retrospectively reviewed the clinical records and radiological findings of the remaining 41 patients with histologically diagnosed benign breast AMEs after complete excision.

Results

All 41 patients underwent US; 38 underwent mammography (MG) and US; and 18 underwent MG, US, and MRI. MG detected 38 cases with a round or oval shape (56%), and mass (89%), were non-circumscribed (62%), hyperdense (53%), and without microcalcifications (95%). Breast US revealed suspicious masses (98%) with a non-circumscribed margin (66%), hypoechogenicity (43%), and intratumoral vascularity (63%). All lesions on breast MRI showed suspicious masses (100%) with ill-defined margins (61%), and 84% showed wash-out kinetics. Benign AMEs showed suspicious features of Breast Imaging Reporting and Data System (BI-RADS) category 4 or 5 in 83%–95% of the MG, US, and MRI. Sixteen of the 41 cases were misdiagnosed on the initial core needle biopsy and two were diagnosed as malignancy.

Conclusion

Benign breast AME often shows suspicious radiological features mimicking a malignant mass on MG, US, and MRI. Differentiating benign AME from other pathologies might be difficult on core needle biopsy, and complete excision is needed for a correct diagnosis.

Graphical Abstract

INTRODUCTION

Breast adenomyoepithelioma (AME) is an uncommon tumor characterized by biphasic proliferation of epithelial and myoepithelial cells in normal breast tissue (1,2). Most AMEs are benign tumors but have the potential for local recurrence after surgical excision or, rarely, of malignant degeneration (3,4). Because of the rarity of these tumors, there have been limited literature describing benign AME. In addition, most studies have focused on the clinical features or histopathologic characteristics of AME rather than the radiological findings (5,6,7).

AME can be a diagnostic challenge because their imaging features are not well known and can mimic breast carcinoma clinically and radiologically (8). Only publications of individual case reports (9,10) or case series with a small number of benign AMEs diagnosed by percutaneous core needle biopsy are available (11,12,13). However, to the best of our knowledge, benign AME may be challenging to diagnose and is sometimes misdiagnosed with a core needle biopsy.

Therefore, it is necessary to analyze the imaging characteristics of benign AMEs identified by complete excision. The purpose of this study was to investigate the imaging characteristics of benign AME of the breast, including mammographic, US, and MRI findings in histologically proven benign AME that were completely excised.

MATERIALS AND METHODS

SUBJECTS

From January 2001 to August 2021, 120 patients had histologically confirmed benign AMEs of the breast. We excluded patients who did not have imaging for the histologically confirmed AME when they were incidentally detected after breast cancer surgery (n = 43), those who had AME confirmed by percutaneous core needle biopsy without subsequent excision (n = 28), and those who had AME confirmed by percutaneous core needle biopsy but had other pathologies after excision (n = 8). Finally, 41 patients were included in the study.

The mean age of the patients was 50 ± 10 years (range, 27–76 years). Clinically, eight patients presented with palpable masses, 15 patients presented screen-detected lesions, four mammography-detected lesions, 11 screening US-detected lesions, 16 patients presented US-detected lesions during the preoperative evaluation of proven breast cancers, and two patients demonstrated abnormalities on PET/CT that were assumed to be malignant during the work-up for cervical cancer. The patients’ demographics included are shown in Fig. 1.

Fig. 1. Flow chart of the study population.

AME = adenomyoepithelioma

This retrospective study was approved by the Institutional Review Board (IRB No. 2021-12-147), and the requirement for informed consent was waived.

IMAGE ANALYSIS

Two radiologists reviewed the mammography, US, and MRI in consensus. On mammography, the lesion was analyzed according to the type of lesion, size, shape, margin, density, presence of calcifications within the mass, and final Breast Imaging Reporting and Data System (BI-RADS) assessment category. On US, the size of the tumor was measured, and the sonographic features were analyzed according to lesion type, shape, margin, orientation, echogenicity, posterior acoustic features, and final BI-RADS assessment category. Other US findings, including calcification, effects of surrounding tissue, and intra-tumoral cystic changes, were also reviewed. In cases where the Doppler US images were available, the presence or absence of vascular flow signals within the lesion was assessed. MRI findings were analyzed according to the lesion type, shape, margin intratumoral cystic change, T2 signal intensity (SI), internal enhancement pattern, enhanced kinetics, and final BI-RADS assessment category.

HISTOLOGIC CONFIRMATION

Percutaneous biopsy was performed using a 14-gauge gun biopsy in 23 patients and fine-needle aspiration in one patient. Every patient who underwent percutaneous biopsy underwent breast-conserving surgery or excisional biopsy. The other 17 patients underwent direct excision after US-guided localization without percutaneous biopsy. Surgical excision was performed in 32 patients, and removal of the mass using a vacuum-assisted device in nine patients. We compared the histologic diagnosis made by percutaneous biopsy and the final diagnosis after the complete excision of the lesion.

RESULTS

Among the 41 enrolled patients, 18 patients underwent mammography, US, and MRI, 20 patients underwent mammography and US, and three had only US images. All 17 lesions that were incidentally detected during the preoperative work-up of known breast cancer and two palpable benign AMEs that were considered as malignant by the initial percutaneous core needle biopsy had breast MRI in addition to mammography and US. Overall, 38 mammograms, 41 breast US, 18 breast MRI, and 2 PET CT images of 41 benign AMEs were analyzed.

Mammography showed positive findings in 18/38 (47%) patients with benign AME. The most common mammographic finding was a mass (89%) with a round or oval shape (56%) but not circumscribed margin (62%) without calcification (95%) (Fig. 2A). Most were assessed as suspicious lesion of BI-RADS category 4 or 5 (83%) (Table 1).

Fig. 2. A 55-year-old woman with a screening mammography-detected mass in the left breast.

A. Mammography shows a newly developed round hyperdense mass with ill-defined margins without calcifications. Her mammography 2 years ago was normal.

B. On breast ultrasonography, the lesion is a 1.9 cm irregular mass with angular margin, heterogeneous isoechogenicity.

C. Doppler ultrasonography shows increased vascularity within the tumor. Based on mammography and ultrasonography, the final assessment was Breast Imaging-Reporting and Data System category 4B. The result of core needle biopsy was intraductal papilloma, however, surgical excision revealed benign adenomyoepithelioma.

Table 1. Mammography Findings of Benign Adenomyoepithelioma of the Breast (n = 38).

Category	Finding		Total (%)
Negative			20 (52.6)
Positive			18 (47.4)
Type	Mass		16 (88.9)
	Shape	Round/oval	9 (56.3)
		Lobular	1 (6.3)
		Irregular	6 (37.5)
	Margin	Circumscribed	6 (37.5)
		Not circumscribed	10 (62.5)
	Non-mass		2 (11.1)
Density	Iso		9 (50)
	Hyper		9 (50)
Calcification	Present		1 (5.6)
	Absent		17 (94.4)
BI-RADS	3		3 (16.7)
	4A		8 (44.4)
	4B		6 (33.3)
	5		1 (5.6)

BI-RADS = Breast Imaging-Reporting and Data System

The most common US finding of benign AME was a mass (98%) with parallel orientation (82%), not circumscribed margin (66%), iso- (45%) or hypo- (43%) echogenicity, and increased vascularity (63%). Nearly equal proportions of masses had round or oval and irregular shape (51% vs. 49%) (Fig. 2B, C). The final BI-RADS assessment of the lesion was equal to or greater than 4A in 85% of patients. The mean lesion size measured on US was 0.9 ± 0.6 cm (0.5–3.1 cm) (Table 2).

Table 2. Ultrasonography Findings of Benign Adenomyoepithelioma of the Breast (n = 41).

Category	Finding		Total (%)
Negative			1 (2.4)
Positive			40 (97.6)
Type	Mass		39 (97.5)
	Shape	Round/oval	20 (51.3)
		Irregular	19 (48.7)
	Orientation	Parallel	32 (82.1)
		Not parallel	7 (17.9)
	Margin	Circumscribed	13 (33.3)
		Not circumscribed	26 (66.7)
	Non-mass		1 (2.5)
Echogenicity	Anechoic		1 (2.5)
	Hypoechoic		17 (42.5)
	Isoechoic		18 (45)
	Complex cystic and solid		2 (5)
	Mixed hyper and hypo		2 (5)
Homogeneity	Heterogeneous		31 (77.5)
	Homogeneous		9 (22.5)
Posterior acoustic feature	No effect		30 (75)
	Enhancement		10 (25)
Surrounding tissue	No effect		40 (100)
	Effect (ductal change, edema)		0 (0)
Calcifications	No calcification		40 (100)
	Calcification		0 (0)
Intratumoral cystic change	Present		5 (12.5)
	absent		35 (87.5)
Vascularity	Avascular		15 (37.5)
	Increased		25 (62.5)
BI-RADS	3		6 (15)
	4A		21 (52.5)
	4B		11 (27.5)
	4C		1 (2.5)
	5		1 (2.5)

BI-RADS = Breast Imaging-Reporting and Data System

On MRI, benign AME showed round or irregular (50% vs. 50%) shape, mass (100%) with a not-circumscribed margin (61%), intermediate or low SI on T2-weighted image (89%), heterogeneous internal enhancement (94%), and mostly wash-out or plateau enhancement on delayed phase (89%), suggesting the possibility of malignancy (Fig. 3). Except one, all lesions (95%), were assessed as suspicious (Table 3).

Fig. 3. A 76-year-old woman with a palpable malignant mass in the right upper outer breast.

A. During the preoperative evaluation for malignancy, another 1.3 cm irregular isoechoic mass with cystic portion was found in the right lower outer breast on ultrasonography.

B-E. Breast MRI shows an irregular mass with intermediate T2 signal (B). Dynamic contrast enhanced images (arrow) (C-E) show heterogeneous enhancement, and wash-out kinetics (arrows). Final assessment after all the imaging studies was Breast Imaging-Reporting and Data System category 4C. Tattooing localization and excision biopsy was performed for this suspicious mass and the pathologic result was benign adenomyoepithelioma.

T2WI = T2-weighted image

Table 3. MR Findings of Benign Adenomyoepithelioma of the Breast (n = 18).

Category	Finding	Total (%)
Background parenchymal enhancement	Minimal	3 (16.7)
	Mild	5 (27.8)
	Moderate	8 (44.4)
	Marked	2 (11.1)
Type	Non-mass	0 (0)
	Mass	18 (100)
Shape	Round/oval	9 (50)
	Irregular	9 (50)
Margin	Circumscribed	7 (38.9)
	Not circumscribed	11 (61.1)
Intratumoral cystic change	Present	1 (5.6)
	Absent	17 (94.4)
Internal enhancement pattern	Heterogeneous	17 (94.4)
	Homogeneous	1 (5.6)
Kinetics	Persistent	2 (11.1)
	Plateau	1 (5.6)
	Washout	15 (83.3)
T2 signal	High	2 (11.1)
	Not high	16 (88.9)
BI-RADS	3	1 (5.6)
	4A	9 (50)
	4B	2 (11.1)
	4C	4 (22.2)
	5	2 (11.1)

BI-RADS = Breast Imaging-Reporting and Data System

Percutaneous biopsy before complete excision was performed in 24 patients (Table 4). Only eight of 24 cases (33.3%) were correctly diagnosed with benign AME by percutaneous core needle biopsy. Among the 16 cases that were misdiagnosed, proliferative lesion with atypia was diagnosed using fine-needle aspiration in one (6%), and the other 15 had a core needle biopsy that diagnosed other pathologies, including two malignant lesions, one invasive ductal carcinoma (58%), and one papillary carcinoma (6%). After complete excision of the lesion or partial mastectomy, the final diagnosis revealed benign AME in all 16 cases. By comparing the histology between the previous core biopsy and the surgical specimen, the results of the core needle biopsy of those lesions were revised as benign AME and atypical papilloma (Fig. 4).

Table 4. Cases Where Benign Adenomyoepithelioma Was not Properly Diagnosed by Initial Core Needle Biopsy (n = 16).

Results of Core Needle Biopsy	Cases (%)
Invasive ductal carcinoma	1 (6)
Papillary carcinoma	1 (6)
Intraductal papilloma	9 (58)
Fibroadenoma or fibroadenomatoid mastopathy	3 (19)
Sclerosing adenosis	1 (6)
Proliferative lesion with atypia	1 (6)

Fig. 4. A 52-year-old woman with a palpable mass in the left breast.

A. On mammography, an irregular hyperdense mass (arrow) is noted at the palpable site.

B, C. Ultrasonography shows an irregular heterogeneous isoechoic mass (B) with increased vascularity within the tumor on Doppler study (C).

D-G. MRI shows a 2.5 cm irregular mass with intermediate T2 signal (D) and heterogenous enhancements (arrow) and wash-out kinetics on dynamic contrast enhanced images (arrows) (E-G). Core needle biopsy result was invasive ductal carcinoma, however, the pathologic result after partial mastectomy was benign adenomyoepithelioma. Re-evaluation of the biopsy specimen after surgery revised the diagnosis to be consistent with benign adenomyolipoma.

T2WI = T2-weighted image

Two lesions detected on PET/CT during the metastatic workup of other malignancies showed peak-standardized uptake value (p-SUV) of 6.7 and 4.2 and were considered malignant.

DISCUSSION

In our study, most benign AMEs (83%–95%) showed suspicious findings on mammography, US, and breast MRI. The likelihood of malignancy was suggested by a BI-RADS 4B or more category in 39% of mammography, 32% of US, and 45% of MRI. The most suspicious imaging characteristics of benign AME were not circumscribed margins and tumor heterogeneity. Another suspicious feature was hypervascularity on Doppler US and wash-out kinetics on MRI. Increased vascularity of the benign AMEs on Doppler US was also reported by Park et al. (12). However, unlike previous studies that reported that 8 of 10 benign AME showed an irregular shape on US (12), the shape of benign AME in our study was round or oval in approximately half of the cases and irregular in the other half.

MRI findings of benign AMEs are not known. Zhang et al. (14) reported MRI findings of three cases, and all showed high SI on T2WI, homogeneous enhancement, and persistent enhancement in 2/3 cases. Another study (11) with MRI findings of three benign AME reported that all cases showed heterogeneous enhancement, wash-out kinetics in 2/3, and plateau pattern in 1/3. In our study of 18 cases of benign AME, MRI showed a mass with a not-circumscribed margin (61%), intermediate to low SI on T2WI (89%), heterogeneous internal enhancement pattern (94%), and wash-out kinetics (84%) or plateau pattern (5%).

The imaging characteristics of benign AME in our study are valuable since this is the largest case series, and above all, only patients with complete excision were included. However, 43 cases were excluded because they were found incidentally in the mastectomy specimen of breast cancer surgery without prior imaging. Even though US was not performed in those 43 cases, a retrograde review of their mammography or MRI could not provide information on the location of AME. Therefore, benign AME may be more common than reported and may have nonspecific benign features that could be ignored.

Core needle biopsy is the most popular and widely used method for diagnosing breast lesion; however, only 8 cases were correctly diagnosed as benign AME before surgery, and 15/23 (65%) were diagnosed with other pathologies in our study. AMEs are considered a variant of intraductal papilloma and can have heterogeneous histologic features with variable components, making the diagnosis difficult with limited tissue samples from the core needle biopsy. The proliferation of myoepithelial cells or combined extensive metaplasia can mimic other benign and malignant lesion (15,16,17).

In our study, the most common misdiagnosis made by core needle biopsy was intraductal papilloma. However, two patients had an initial diagnosis of invasive ductal carcinoma and papillary carcinoma with radio-pathologically concordant malignant findings and underwent a partial mastectomy. They presented with firm palpable masses and highly suspicious findings on all imaging modalities, including breast MRI (BI-RADS category 4C and 5). The size of the lesions were 2.3 cm and 2.5 cm. Limited tissue from percutaneous biopsy with extensive spindle cell proliferation and atypical cells in the background of highly suspicious clinical and radiological findings might lead to overdiagnosis by the pathologist.

There are a few limitations to our study. First, the included patients were only one-third of the total cases that were histologically confirmed. Many cases that were excised without imaging might be too small to detect. However, the mean lesion size in our study was 0.9 cm on US; therefore, the imaging characteristics of benign AME in our study can only represent detected benign AME rather than total benign AME. Second, our research focused on the imaging characteristics in this study, and we did not study the pathologic features. A further in-depth investigation into the pathologic findings of cases in which biopsy and surgery results were discordant would be needed.

In conclusion, benign AME of the breast shows suspicious findings on mammography, US, and MRI with increased vascularity and wash-out kinetics. It is challenging to diagnose using core needle biopsy only, and complete excision is needed for a correct diagnosis.