Mammographic and ultrasonographic features of triple-negative breast cancer: a comparison with other breast cancer subtypes

Abstract

Background

Triple-negative breast cancer (TNBC) is known to be associated with aggressive biologic features and a poor clinical outcome. Therefore, early detection of TNBC without missing cancer is mandatory to improve prognosis.

Purpose

To retrospectively evaluate the mammographic and sonographic features of TNBC compared to ER (+) cancers and HER2 (+) cancers.

Material and Methods

From June 2011 through June 2012, mammographic and sonographic features of 281 surgically confirmed ER (+) cancers (n = 153), HER2 (+) cancers (n = 83), and TNBC (n = 45) were retrospectively reviewed by two radiologists in consensus. The clinicopathological features were also compared between the three subtypes. Additionally the 45 TNBC cases were analyzed using morphologic criteria of ACR BI-RADS lexicon.

Results

Most TNBC (93.3%, 42 of 45) were mammographically seen and presented with mass or focal asymmetry without microcalcifications (P < 0.001). Ultrasonographically TNBC were presented as hypoechoic nodule without microcalcifications (P < 0.001). Palpable symptom (P < 0.001), a lower incidence of ductal carcinoma in situ (P < 0.001), invasive tumor size that is >2 cm (P = 0.028) and high histologic grade (P < 0.001) were significantly associated with TNBC. With regard to morphologic features of 45 TNBC cases, TNBC were most likely to be hyperdense mass (89.3%) with oval (68.9%) or lobular shape (28.6%) and indistinct (42.9%) or circumscribed margin (32.1%) on mammogram. Ultrasonographically TNBC were most likely to be a hypoechoic mass (82.2%) with irregular (68.9%) or oval (28.9%) shape and microlobulated (46.7%), circumscribed (17.8%) or indistinct margin (17.8%) and parallel orientation (68.9%).

Conclusion

Being familiar with combined mammographic and ultrasonographic features of TNBC may be useful to avoid false-negative cases of TNBC.

Keywords

Breast neoplasms mammography ultrasonography

Breast cancer is a heterogeneous disease composed of different biological subgroups that have various morphologies, clinical outcomes, and responses to systemic therapy. Currently, gene expression analysis based on the extent of the expression of the estrogen receptor (ER), progesterone receptor (PR) and human epidermal growth factor receptor 2 (HER2) has identified several subtypes with common molecular features: Luminal A and B (both ER-positive), normal breast-like, basal-like, and HER2-positive (all three ER-negative) (1, 2). Clinically, breast cancers are categorized into one of three major subtypes, using the three standard immunohistochemical markers, to facilitate targeted therapy; these subtypes are ER (+) cancer, HER2 (+) cancer, and triple-negative breast cancer (TNBC) (3, 4). Indeed, the basal-like subtype is composed almost entirely of triple-negative breast cancer, which is negative for ER, PR, and HER2 (5).

Triple-negative breast cancer (TNBC) is known to be associated with aggressive histologic features, a poor clinical outcome, and a shorter survival compared with other breast cancer subtypes (6 –8). Contrary to its aggressive nature, TNBC has been described as having benign morphology by conventional imaging modalities (9 –13). Therefore, being familiar with the mammographic and ultrasonographic imaging features of TNBC may help examiner to detect TNBC early without missing false-negative cases. However, to the best of our knowledge, there are a few studies reporting the imaging features of TNBC, especially compared to non-TNBC. In this study, we retrospectively evaluate the clinicopathological, mammographic and ultrasonographic features of triple-negative breast cancer in comparison with ER (+) cancer and HER2 (+) cancer.

Material and Methods

Study population

Our institutional review board approved this study, and the requirement for informed consent was waived. From June 2011 through June 2012, a total of 363 patients underwent surgery to treat breast cancer. Among these patients, 36 were excluded because three patients did not undergo immunohistochemical analysis and 34 patients did not undergo SISH (silver in situ hybridization) for determination of the HER2 status. In addition, 45 patients were excluded due to the non-availability or poor quality of their preoperative mammographic or ultrasonographic images. Finally, 278 patients (mean age, 49.8 years; range, 25–84 years) with 281 breast cancers were included in our study. The indications of initial mammography or ultrasound were as follows: 55.0% (153 of 278) were due to palpability, 39.9% (111 of 278) were for asymptomatic screening, 3.2% (9 of 278) were due to pain and 1.8% (5 of 278) were due to nipple discharge.

Clinicopathologic data

The patients' medical records were reviewed to identify the clinical symptoms (e.g. the presence of a palpable mass). The pathological reports were also reviewed to determine the tumor size, histological grade, and presence of axillary lymph node metastasis. The patients' ER, PR, and HER2 statuses were determined by immunohistochemical analysis. For the immunohistochemical analysis, formalin-fixed, paraffin-embedded tissue sections were immunohistochemically stained. The Allred score was used to determine the ER and PR statuses. The results were classified as positive when the total score, expressed as the sum of the proportion and immunointensity scores, was 3 or more.

With regard to the HER2 evaluation, tumors with a 3+score were classified as HER2-positive and tumors with a 0 or 1+score were classified as negative. In tumors with a 2+score, gene amplification using SISH (silver in situ hybridization) was used to determine the HER2 status. The HER2 expression was considered positive if the ratio of the HER-2 gene copies to the chromosome 17 signal was>2.

Imaging analysis

Mammography was performed using a Selenia system (Lorad, Bedford, CT, USA). Standard two-view mammography (mediolateral oblique and craniocaudal) was performed, with additional views as deemed necessary. Breast ultrasound (US) was performed using 5–12-MHz transducers with an IU-22 (Philips Medical Systems, Best, The Netherlands) by two breast imaging radiologists.

Two breast radiologists with 4 and 8 years of clinical experience (MYK and NC) retrospectively reviewed the mammographic and ultrasonographic images in consensus without knowledge of the clinicopathological findings of the cases. On mammograms, lesions were described as mass or focal asymmetry with or without microcalcifications. If the lesion was associated with microcalcifications, the microcalcification was described as fine linear or pleomorphic, amorphous, or punctate. Breast density was also rated as fatty, scattered fibroglandular, heterogeneously dense, or dense. On ultrasonography, lesions were described as mass with or without microcalcifications. Also, both mammographic and ultrasonographic images of TNBC were reviewed by two radiologists in consensus. For image interpretation of TNBC, American College of Radiology Breast Imaging Reporting and Data System Lexicon was used for morphologic criteria (14).

Statistical analysis

The clinicopathological, mammographic, and ultrasonographic features were compared between the three tumor subtypes using the chi-square or Fisher's exact tests for categorical variables and Student's t-test for continuous variables. All statistical analyses were carried out using SPSS version 17.0 for Windows (SPSS, Chicago, IL, USA), and P values<0.05 were considered significant.

Results

Out of a total of 281 breast cancers, 45 (16.0%) were TNBC, 153 (54.4%) were ER-positive cancer, and 83 (29.5%) were HER2-positive cancer (Table 1). The presence of palpable symptom was significantly associated with TNBC compared with non-TNBC (82.2% in TNBC, 50.3% in ER (+) cancer, and 48.2% in HER2 (+) cancer). The mean invasive tumor size was 2.1 cm (range, 0.1–6.0 cm) for ER (+) cancers, 2.0 cm (range, 0.1–6.4 cm) for HER2 (+) cancers, and 2.3 cm (range, 0.1–4.3 cm) for TNBC, respectively. The invasive tumor size that is >2 cm was more often in TNBC (63.6%, 28 of 45) compared to ER (+) cancers (41.2%, 56 of 153) and HER2 (+) cancers (41.9%, 26 of 83). There was no significant difference in the operation type between the three tumor subtypes (P = 0.275). Axillary lymph node metastases were identified in 30.2% (41 of 153), 41.0% (25 of 83), and 36.4% (16 of 45) of ER (+) cancer, HER2 (+) cancer, and TNBC, respectively. The histologic grade was significantly different across the tumor subtypes. A significantly higher percentage of histologic grade 3 was noted in TNBC (73.3%, 33 of 45) than in ER (+) cancer (19.0%, 29 of 153) or HER2 (+) cancer (31.3%, 26 of 83) (P < 0.001).

Table 1

Clinicopathological and imaging features according to the tumor subtype

Features	ER (+) (n = 153)	HER2 (+) (n = 83)	TNBC (n = 45)	P value
Mean age (years)*	50.4 (30–76)	51.2 (26–84)	47.9 (25–80)	0.202
Palpability				< 0.001
No	76 (49.7)	43 (51.8)	8 (17.8)
Yes	77 (50.3)	40 (48.2)	37 (82.2)
Operation type				0.275
Partial mastectomy	121 (79.1)	60 (72.3)	31 (68.9)
Total mastectomy	32 (20.9)	23 (27.7)	14 (31.1)
Pathology				< 0.001
Invasive ductal carcinoma	121 (79.1)	61 (73.5)	40 (88.9)
Invasive lobular carcinoma	10 (6.5)	1 (1.2)	0 (0.0)
Ductal carcinoma in situ	15 (9.8)	21 (25.3)	1 (2.2)
Lobular carcinoma in situ	2 (1.3)	0 (0.0)	0 (0.0)
Mucinous carcinoma	4 (2.6)	0 (0.0)	0 (0.0)
Medullary cancer	0 (0.0)	0 (0.0)	3 (6.7)
Adenoid cystic cancer	0 (0.0)	0 (0.0)	1 (2.2)
Invasive micropapillary cancer	1 (0.7)	0 (0.0)	0 (0.0)
Tumor size (cm)				0.028
≤2	80 (58.8)	36 (58.1)	17 (37.8)
≥2	56 (41.2)	26 (41.9)	28 (62.2)
Axillary lymph node positivity				0.311
No	95 (69.9)	36 (59.0)	29 (64.4)
Yes	41 (30.2)	25 (41.0	16 (35.6)
Histologic grade				< 0.001
Grades I and II	107 (69.9)	36 (43.4)	11 (24.4)
Grade III	29 (19.0)	26 (31.3)	33 (73.3)
Unknown	17 (11.1)	21 (25.3)	1 (2.2)
Mammographic abnormality				0.003
No	31 (20.3)	5 (6.0)	3 (6.7)
Yes	122 (79.7)	78 (94.0)	42 (93.3)
Abnormal mammographic finding				< 0.001
Mass or focal asymmetry only	69 (56.6)	26 (33.3)	32 (76.2)
Mass or focal asymmetry with microcalcifications	36 (29.5)	32 (41.0)	7 (16.7)
Microcalcifications only	17 (13.9)	20 (25.6)	3 (7.1)
Type of microcalcifications				0.452
Fine linear or pleomorphic	39 (73.6)	43 (82.7)	7 (70.0)
Punctate or amorphous	14 (26.4)	9 (17.3)	3 (30.0)
Breast density				0.524
Fatty or scattered fibroglandular	20 (13.1)	15 (18.1)	8 (17.8)
Heterogeneously dense or dense	133 (86.9)	68 (81.9)	37 (82.2)
Sonographic abnormality				0.165
No	2 (1.3)	4 (4.8)	0 (0.0)
Yes	151 (98.7)	79 (95.2)	45 (100.0)
Abnormal sonographic finding				< 0.001
Mass only	111 (73.5)	37 (16.8)	38 (84.4)
Mass with calcifications	40 (26.5)	42 (53.2)	7 (15.6)

Data are the numbers of subjects, with percentages in parentheses

*Numbers in parentheses are age ranges

Pathologically invasive ductal carcinoma was the most common histologic type in ER (+) cancer (79.1%, 121 of 153), HER2 (+) cancer (73.5%, 61 of 83), and TNBC (88.9%, 40 of 45). Significantly lower incidence of ductal carcinoma in situ (DCIS) was observed in TNBC (2.2%, 1 of 45) compared to ER (+) cancer (9.8%, 15 of 153) and HER2 (+) cancer (25.3%, 21 of 83). All medullary (n = 3) and adenoid cystic (n = 1) carcinomas were TNBC.

By mammography, TNBC, ER (+) cancer, and HER2 (+) cancer were detected in 93.3% (42 of 45), 79.7% (122 of 153), and 94.0% (78 of 83) of the patients (P = 0.003), respectively. Masses or focal asymmetries without microcalcifications (76.2% in TNBC vs. 33.3% in HER2 (+) cancer vs. 56.6% in ER (+) cancer; P < 0.001) were significantly associated with TNBC, compared to ER (+) cancer and HER 2(+) cancer (Fig. 1). The types of microcalcifications and breast density were not different between the tumor subtypes (P = 0.452, P = 0.524). By ultrasonography, ER (+) cancer, HER2 (+) cancer, and TNBC were detected at levels of 98.7%, 95.2%, and 100% (P = 0.165), respectively. The presence of mass or focal asymmetry lacking microcalcifications (84.4% in TNBC vs. 16.8% in ER (+) cancer vs. 73.5% in HER2 (+) cancer; P < 0.001) was significantly associated with TNBC.

Fig. 1

A 26-year-old woman with palpable left breast mass. Left craniocaudal (a) and mediolateral oblique view (b) of mammograms showed an oval circumscribed hyperdense mass in the left upper outer quadrant and enlarged axillary lymph nodes with absence of fatty hilum, suggesting metastasis. Ultrasonographic images (c, d) showed a 3 cm oval microlobulated hypoechoic mass with posterior enhancement, which was classified as BI-RADS category 4. Two enlarged lymph nodes with cortical thickening were seen in left axilla. 14G gun biopsy was performed for the mass, which was confirmed to be invasive carcinoma. She underwent breast conserving surgery, which revealed a 2.8 cm medullary carcinoma with histologic grade III, and none of the six dissected axillary lymph nodes had metastasis

Mammographically, of the 45 TNBC cases, 28 (62.2%) presented as a mass and 17 presented with focal asymmetry. Table 2 shows the mammographic findings of the 28 TNBC cases that presented as a mass. The mass margin was more commonly indistinct (42.9%, 12 of 28) and less commonly circumscribed (32.1%, 9 of 28). The mass shape was more commonly oval (35.7%, 10 of 28) and less commonly lobular (28.6%, 8 of 28). Most TNBC had high density (89.3%, 25 of 28), and the others had equal density (10.7%, 3 of 28). Ultrasonographically, all 45 TNBC cases presented as discrete masses. The TNBC mass shape was more commonly irregular (68.9%, 31 of 45) and less commonly oval (28.9%, 13 of 45) (Table 3). The mass margin was more commonly microlobulated (46.7%, 21 of 45) and less commonly circumscribed (17.8%, 8 of 45), indistinct (17.8%, 8 of 45) or angular (15.6%, 7 of 45). Most TNBC cases were hypoechoic (82.2%, 37 of 45), while the remaining cases were complex echoic (17.8%, 8 of 45). Comparing the TNBC cases, 62.2% (28 of 45) had posterior acoustic features, of which posterior enhancement was the most common (53.3%, 24 of 45). The lesion boundary in TNBC was more commonly an echogenic halo (66.7%, 30 of 45), while the remaining cases had an abrupt interface (33.3%, 15 of 45). The BI-RADS category was based on the mammographic and ultrasonographic findings. The BI-RADS category for the TNBC cases was most commonly category 6 (53.3%, 24 of 45) and less commonly category 4 (26.7%, 12 of 45) or category 5 (20.0%, 9 of 45).

Table 2

Mammographic findings of 28 triple-negative breast cancers presenting as masses

Features		Patients (n, %)
Shape	Round	5 (17.9)
	Oval	10 (35.7)
	Lobular	8 (28.6)
	Irregular	5 (17.9)
Margin	Circumscribed	9 (32.1)
	Microlobulated	1 (3.6)
	Obscured	5 (17.9)
	Indistinct	12 (42.9)
	Spiculated	1 (3.6)
Density	High density	25 (89.3)
	Isodense	3 (10.7)
	Low density	0 (0.0)
	Radiolucent	0 (0.0)

Data are the numbers of subjects, with percentages in parentheses

Table 3

Ultrasonographic findings of 45 triple-negative breast cancers

Features		Patients (n, %)
Shape	Oval	13 (28.9)
	Round	1 (2.2)
	Irregular	31 (68.9)
Orientation	Parallel	31 (68.9)
Orientation	Non-parallel	14 (31.1)
Margin	Circumscribed	8 (17.8)
	Indistinct	8 (17.8)
	angular	7 (15.6)
	Microlobulated	21 (46.7)
	Spiculated	1 (2.2)
Lesion boundary	Abrupt interface	15 (33.3)
Lesion boundary	Echogenic halo	30 (66.7)
Echo pattern	Anechoic	0 (0.0)
	Hyperechoic	0 (0.0)
	Complex echoic	8 (17.8)
	Hypoechoic	37 (82.2)
	Isoechoic	0
Posterior acoustic feature	None	17 (37.8)
	Enhancement	24 (53.3)
	Shadowing	3 (6.7)
	Combined	1 (2.2)
BI-RADS category	Category 4	12 (26.7)
	Category 5	9 (20.0)
	Category 6	24 (53.3)

Discussion

Triple receptor-negative cancer (TNBC) is a subtype of breast cancer characterized by a lack of expression of ER, PR, and HER2 (1). Many authors have reported that TNBC shows aggressive biologic features and is associated with high incidence of interval cancer due to rapid growth pattern (4 –8). Compared with non-TNBC, TNBC cases are of high histologic grade, high nuclear grade, and unfavorable histology, such as anaplastic or metaplastic carcinoma (15, 16).

Histopathologically, similar with prior reports, invasive tumor size that is > 2 cm (62.2%, 28 of 45) and high histologic grade (33 of 45, 75%) were more often in TNBC compared to non-TNBC in our study. As a result, TNBC was associated with palpable symptom (82.2%) and subsequently detected clinically rather than with imaging. And, there was no difference in the rate of axillary lymph node positivity between the tumor subtypes (30.2% in ER (+) cancer, 41.0% in HER2 (+) cancer, and 36.4% in TNBC; P = 0.311). Generally the larger the tumor, the rate of lymph node positivity increase, but this relationship was not observed in TNBC. Foulke et al. explained this phenomenon by mode of spread of TNBC and it is thought that TNBC spread hematogenously due to early vascular invasion (17).

Our results showed that most TNBC cases (93.3%, 42 of 45) were mammographically detected, and only 6.7% (3 of 45) were mammographically occult. As TNBC often presented clinically with a palpable mass (82.2%), mammographic examinations were frequently diagnostic rather than screening. By mammography, TNBC most commonly appeared as a mass or focal asymmetry lacking microcalcifications. The mass margin was more commonly indistinct (42.9%) and less commonly circumscribed (32.1%). These results are consistent with prior studies, which reported that TNBC was most likely a mass without calcifications by mammography (average 57.1%, 352 of 616) (9 –12). Traditionally mammographic microcalcification is a hallmark feature of DCIS (18). In our results, a significantly lower incidence of DCIS was observed in TNBC compared with ER (+) cancer and HER2 (+) cancer (25.3%, 21 of 83), which may explain why TNBC is less likely to be associated with calcifications by mammography. Similarly, Yang et al. reported that TNBC is less likely to be associated with DCIS, which is concordant with a lack of mammographic calcifications (12).

With regard to ultrasonographic findings of TNBC, it was reported that TNBC shows circumscribed or lobulated margin more often rather than spiculated or angular margin (9, 13). Wojcinski et al. described this smooth appearance as a pushing border that is associated with non-infiltrative process by rapid tumor growth. In our study, TNBC was likely to be hypoechoic mass (82.2%) or complex mass (17.8%) with microlobulated (46.7%) or circumscribed (17.8%) or indistinct (17.8%) or angular (15.6%) margin (13). Mass lesion with circumscribed margin might be misinterpreted as benign. In such circumstances, we believe that microlobulated margin might be useful to avoiding false-negative cases of TNBC. Unlike prior studies, echogenic halo was observed more often in TNCB (66.7%) in our study. We suppose that the echogenic halo results from peritumoral edema caused by severe peritumoral inflammation which is representatively seen in high grade invasive carcinoma (19). TNBC were more likely to be associated with posterior acoustic enhancement (53.3%). It is well-known that high grade, highly cellular circumscribed carcinomas tends to have enhanced through-transmission (19). With regard to orientation of the tumor, TNBC were more likely to be horizontally orientated (68.9%), which is usually associated with benign nodule. This finding can be explained by a relatively large tumor size of TNBC at presentation. Traditionally a taller than wide shape is characteristic for small carcinoma (20). A growing tumor with pushing margins (expanding and not invading) would respect the limits of a breast segment or lobe, the functional unit of the breast. Indeed, in our study, invasive tumor size that is >2 cm was more often in TNBC (62.2%) compared to non-TNBC (41.2% for ER (+) cancer, 41.9% for HER2 (+) cancer) (P = 0.028).

Our study has several limitations. First, our study is retrospective, and its sample size was small. Second, we did not evaluate inter-observer variability regarding the classification of mammographic and ultrasonographic features which were carried out in consensus in our study. Third, we did not analyze the imaging features of TNBC according to the cancer stage, which might influence the imaging features at diagnosis. Fourth, we did not include the findings of other imaging modalities such as MRI, which may be more sensitive for the diagnosis of TNBC.

In conclusion, combined mammographic and ultrasonographic features of mass or focal asymmetry without microcalcifications that is hypoechoic mass with microlobulated or circumscribed margin can be suggestive of presence of TNBC. Ultrasonographically microlobulated margin can be helpful in avoiding false-negative cases of TNBC, especially when the nodule is horizontally-oriented. Being familiar with combined mammographic and ultrasonographic findings of TNBC would be useful for early detection of TNBC without missing cancer.

Footnotes

Conflict of interest: None.

References

Perou

Sørlie

Eisen

Molecular portraits of human breast tumours. Nature 2000;406:747–52

Sørlie

Perou

Tibshirani

Gene expression patterns of breast carcinomas distinguish tumor subclasses with clinical implications. Proc Natl Acad Sci U S A 2001;98:10869–74

Haffty

Yang

Reiss

Locoregional relapse and distant metastasis in conservatively managed triple negative early-stage breast cancer. J Clin Oncol 2006;24:5652–7

Dent

Trudeau

Pritchard

Triple-negative breast cancer: Clinical features and patterns of recurrence. Clin Cancer Res 2007;13:4429–34

Rakha

Reis-Filho

Ellis

. Basal-like breast cancer: A critical review. J Clin Oncol 2008;26:2568–81

Pal

Childs

Pegram

. Triple negative breast cancer: Unmet medical needs. Breast Cancer Res Treat 2011;125:627–36

Bauer

Brown

Cress

Descriptive analysis of estrogen receptor (ER)-negative, progesterone receptor (PR)-negative, and HER2-negative invasive breast cancer, the so-called triple-negative phenotype: A population-based study from the california cancer registry. Cancer 2007;109:1721–8

Collett

Stefansson

Eide

A basal epithelial phenotype is more frequent in interval breast cancers compared with screen detected tumors. Cancer Epidemiol Biomarkers Prev 2005;14:1108–12

Lee

Kim

Triple-negative breast cancer: Correlation between imaging and pathological findings. Eur Radiol 2010;20:1111–7

10.

Dogan

Gonzalez-Angulo

Multimodality imaging of triple receptor-negative tumors with mammography, ultrasound, and MRI. Am J Roentgenol 2010;194:1160–6

11.

Krizmanich-Conniff

Paramagul

Patterson

Triple receptor-negative breast cancer: Imaging and clinical characteristics. Am J Roentgenol 2012;199:458–64

12.

Yang

Dryden

Broglio

Mammographic features of triple receptor-negative primary breast cancers in young premenopausal women. Breast Cancer Res Treat 2008;111:405–10

13.

Wojcinski

Soliman

Schmidt

Sonographic features of triple-negative and non-triple-negative breast cancer. J Ultrasound Med 2012;31:1531–4

14.

American College of Radiology. Illustrated breast imaging reporting and data system (BI-RADS). Reston, VA: American College of Radiology; 2003

15.

Rakha

El-Sayed

Green

Prognostic markers in triple-negative breast cancer. Cancer 2007;109:25–32

16.

Reis-Filho

Milanezi

Steele

Metaplastic breast carcinomas are basal-like tumours. Histopathology 2006;49:10–21

17.

Foulkes

Brunet

Stefansson

The prognostic implication of the basal-like (cyclin E high/p27 low/p53 /glomeruloid-microvascular-proliferation) phenotype of BRCA1-related breast cancer. Cancer Res 2004;64:830–5

18.

Stomper

Connolly

. Ductal carcinoma in situ of the breast: Correlation between mammographic calcification and tumor subtype. Am J Roentgenol 1992;159:483–5

19.

Stavros

. Ultrasound of solid breast nodules: distinguishing benign from malignant. In: Breast ultrasound. Philadelphia, PA: Lippincott Williams & Wilkins; 2004:445–525

20.

Stavros

Thickman

Rapp

Solid breast nodules: Use of sonography to distinguish between benign and malignant lesions. Radiology 1995;196:123–34