Imaging findings with clinicopathological correlation in breast carcinoma in young women

Imaging Evaluation and Interpretation

All 71 patients underwent an ultrasound of both breasts, axillae, and supraclavicular regions; 66 patients had bilateral mammograms; and three had a dynamic contrast-enhanced MRI. US was performed on an Esaote machine using a 4–15 MHz frequency linear probe. In most of the cases, mammograms were performed after ultrasound revealed BIRADS 4/5 lesions. In a few cases, a mammogram was the first investigation done if there was clinical suspicion of malignancy. Routine craniocaudal and mediolateral oblique views with 2D and tomosynthesis were obtained in all cases undergoing mammograms. Other views, like spot compression or magnification, were obtained as needed. The machine used was Hologic Selenia Dimensions, having 3D Digital Breast Tomosynthesis. MRI is not performed routinely at our institute for diagnostic purposes. Only 3 of the 71 patients underwent MRI for the purpose of problem-solving. A three Tesla GE scanner (Discovery 750 w) was used with a dedicated breast coil. Patients were scanned in the prone position. The following are the sequences obtained and the scan parameters used: an axial, turbo spin-echo T2W sequence (TE/TR of 85/6223 ms, flip angle of 111 degrees, FOV 320 mm, matrix 416 × 320 mm, 2 NEX, slice thickness of 4.5 mm, acquisition time of 3 minutes 20 seconds) and a pre- and dynamic post-contrast sequence using the T1W 3D VIBRANT sequence (TE/TR of 2.1/4.3 ms, flip angle of 10 degrees, 1.2 mm slice thickness with no gap, FOV 330 mm, and matrix 320 × 320). Post-contrast images were obtained after a bolus injection of 0.1 mmol/kg of Omniscan (Gadodiamide). All ultrasound, mammography, and MRI findings were retrospectively interpreted by two radiologists. In cases of disagreement, a consensus was reached by discussion. All lesions were evaluated using the morphological criteria described for all three modalities according to the American College of Radiology (ACR) Breast Imaging Reporting and Data System (BIRADS) atlas, fifth edition. We also correlated imaging descriptors with the index tumor grade and molecular subtypes.

Pathological Analysis

Core needle biopsy using 14G automatic or semi-automatic guns was done under ultrasound guidance from BIRADS 4 and 5 lesions. In case of multiple suspicious lesions, biopsy was done from the most suspicious lesion and the most distant lesion from the index lesion. Lesions in both breasts were biopsied using different biopsy guns; to rule out synchronous malignancy. At least five cores were obtained and transferred in 10% neutral buffered formalin solution to the pathology department for review. The pathological specimens were reviewed by two pathologists with experiences of more than 10 and 8 years, the former having expertise in breast pathology. The specimens were evaluated for the presence or absence of malignancy, histological type, and histological grade. Estrogen receptor (ER), progesterone receptor (PR), and human epidermal growth factor receptor 2 status were determined using immunohistochemistry. The Allred scoring system was used to determine the ER and PR status.^[5] The results were classified as positive when the total score was more than or equal to 3. For HER2 status, tumors with a 3+ score were classified as HER2 positive, and tumors with a 0 or 1+ score were classified as negative. In equivocal cases (+2), Fluoroscent in situ hydbridization (FISH) was done to confirm the HER2 status. In our study, antigen Ki67 was also evaluated, which indicates tumor activity. Using the result of immunohistochemistry, the tumors were classified as ER-positive, HER2 enriched, and triple-negative. ER positive were further divided into luminal A and B based on Ki67 activity using a cutoff value of 15%.^[6]

Clinicopathological characteristics

All 71 patients initially underwent clinical examination by a breast surgeon. The mean age was 35.16 ± 3.85 years, and the age range was from 20 to 39 years. Figure 1 shows the distribution of cases across various age groups. Table 1 summarizes the clinicopathological findings of these cases. Four patients had positive family histories in first-degree relatives. Two had a personal history of breast cancer, one in the same breast and the other in the contralateral breast. Almost 92% of the patients presented with a palpable lump ranging from 1.5 to 5.0 cm on clinical examination. Associated findings, including skin thickening and edema, peau d’orange appearance of the breast, skin ulceration, fixity to skin and chest wall, nipple retraction, and ulceration, were seen in 20.3% of our cases. Palpable axillary nodes were seen in 42.2%, and supraclavicular nodes in 12.6% of our cases. 50 (70%) women presented with locally advanced disease at the time of presentation, and 8 (11%) had metastatic disease.

Close

Figure 1:

Bar diagram showing distribution of breast cancer cases in our study across various age groups.

Export to PPT

The most common histological type was intraductal carcinoma, no-specific type (IDC-NST), present in 82.4% of the cases, followed by intraductal carcinoma with lobular differentiation, seen in 5.4% of the patients. The incidence of other types has been summarized in Table 1. Most (65.7%) of the patients had grade 1 tumors as per the Nottingham grading system.^[7] 31.4% had grade 2 tumors, and only 2.8% had grade 3 tumors.

We used the molecular classification of breast cancer as per the Gallen consensus 2011^[8] and divided the breast cancers into the following four types: triple negative breast cancer was seen in 42.6% of the cases, which was the most common type, followed by luminal type A, seen in 35.2%. The other two types were luminal type B (13.2%), and the least common type was HER2 enriched (9%).

Imaging features

Unilateral breast involvement was found in 68 cases, and synchronous malignancies were seen in 3 of the cases, making the total number of breasts with cancer 74. Both breasts had an equal incidence of involvement. Positive ipsilateral axillary nodes were seen in 68.9% of the patients on ultrasound. The frequency of involvement of other lymph nodes is as follows: ipsilateral supraclavicular nodes 18.9%, ipsilateral infraclavicular or level III lymph nodes 18.9%, and contralateral axillary nodal involvement was seen in 8.1% of the cases. The malignancy was multifocal-multicentric in 32 (43.2%) cases. On imaging, skin and nipple involvement was seen in 33 (44.5%) cases, and skin thickening due to infiltration or lymphatic obstruction was noted in 28 (37.8%) cases. All 71 cases were given BIRADS category 4 or 5. BIRADS 5 was the most common final assessment category given in 55 (74.3%) of the cases, BIRADS 4A in 3 (4%), BIRADS 4B in 5 (6.7%), and BIRADS 4C in 11 (14.8%) of the cases. These features are summarized in Table 2.

US

All 71 patients underwent Ultrasound (USG) evaluation, and the findings are shown in Table 3. The mean size of the tumor on ultrasound B-mode was 41.6 ± 21.5 mm. No abnormality was seen on ultrasound in one patient in whom mammography showed an area of architectural distortion. The most common finding was a mass seen in 69 (93.2%) of the 74 breasts. Most of these masses were irregular in shape (60, 81.0%), with a hypoechoic-heterogeneous echo pattern (63, 91.3%), spiculated margins (30, 43.3%), and no posterior features (25, 36.2%) or shadowing (24, 34.8%) [Figures 2-4]. The other less common presentations on USG included dilated ducts (2, 2.7%) and non-mass lesions (2, 2.7%).

Close

Figure 2:

A 38-year-old female presented with a palpable lump in the right breast in the upper outer quadrant. (a-b) MLO and CC view of mammogram of right breast reveal An irregular high-density mass with spiculated margins and associated skin retraction and invasion. Diffuse skin and trabecular thickening noted secondary to lymphatic obstruction. (c) Corroborative grayscale ultrasound shows an irregular heterogeneous mass with spiculated margins and shadowing. (d) USG of the ipsilateral axilla revealed irregular level I lymph node with extracapsular breach. (e) H&E, 20x: grade 3 invasive breast carcinoma, NST. (f) Ki67 labeling index: 70%–80%, grade 3 IBC, NST. MLO: Mediolateral oblique, CC: Cranio caudal, USG: Ultrasound, H&E: Hematoxylin and eosin, NST: Non specific type.

Export to PPT

Close

Figure 3:

A 33-year-old woman who presented with a palpable lump in the left breast. (a) Grayscale USG image of the left breast shows an irregular shaped, heterogeneous, anti-parallel mass with microlobulated margins (blue arrow) and posterior enhancement. (b) Tomosynthesis slice showed a partially visualized high-density irregular mass in the posterior third of the breast. (c) H&E, 20x: ER/PR: negative (Allred score 0/8), HER2Neu negative in grade 3 carcinoma was found on core needle biopsy of the lump in the above patient. USG: Ultrasound, H & E: Hematoxylin and Eosin, ER: Estrogen receptor, PR: Progestreone receptor.

Export to PPT

Close

Figure 4:

Vague pain as presentation in 39 year old female: Dynamic contrast enhanced MRI revealed an irregular spiculated mass in upper outer quadrant of right breast which was hypointense on (a) T2 and showed heterogeneous enhancement on (b-c) post contrast images with (d) type III kinetic curve. (e) The corroborative USG showed mass as irregular, non-parallel, heterogeneous, spiculated and with no posterior features. It was seen as an area of architectural distortion better appreciated on (f) tomosynthesis. (g) Biopsy confirmed it as invasive ductal carcinoma, NST. H&E, 20x: Photomicrograph showing nests and sheets of tumor cells exhibiting marked nuclear pleomorphism, enlarged nuclei with vesicular chromatin, and increased mitosis—grade 3 invasive breast carcinoma, NST. (h) IHC for ER and PR, 10x: shows ER+/PR+ (Allred score 5 + 3 = 8). MRI: Magnetic resonance imaging, USG: Ultrasound, H&E: Hematoxylin and eosin, IHC: immunohistochemistry, NST: Non specific type, ER: Estrogen receptor, PR: Progestreone receptor.

Export to PPT

TNBC can present as a rounded circumscribed mass with posterior enhancement, which are features of benign tumors. Figure 5a shows a well-circumscribed solid cystic mass with posterior acoustic enhancement. BIRADS 4C category was assigned by USG on the basis of the sudden increase in size of the lesion and the presence of suspicious ipsilateral lymph nodes. The lesion came out to be invasive ductal carcinoma, grade 3, TNBC. Figure 6 shows a similar oval-shaped mass lesion with circumscribed margins, internal cystic spaces, and posterior enhancement. Due to the chronicity of the lesion and absence of suspicious lymph nodes, a BIRADS category of 4B was assigned on USG. The lesion turned out to be a benign phyllodes tumor on histopathological examination (HPE).

Close

Figure 5:

A 25-year-old female presented with a rapidly growing lump in the right breast associated with pain. (a) A grayscale ultrasound of the right breast shows a large, rounded, circumscribed, solid-cystic mass with associated posterior enhancement. (b) The CC view of the mammography shows a large rounded high-density lesion with circumscribed margins occupying all quadrants of the breast. CC: Cranio-caudal.

Export to PPT

Close

Figure 6:

A 37-year-old female presented with a lump in her left breast for 5 years. A grayscale USG of the left breast shows a parallel-oriented oval-shaped lesion with circumscribed margins, heterogeneous echotexture, and posterior enhancement (yellow arrow). Small internal cystic spaces are seen (blue arrow).

Export to PPT

Mammography

A total of 66 patients had mammography done, and the findings are shown in Table 4. Most (44, 63.7%) of the breasts were heterogeneously dense as per ACR categories. The most common presentation was mass, seen in 63 (91.3%) of the cases. Most of these masses were irregularly shaped (80.9%), had high density (79.3%), and had spiculated margins (53.9%) in mammography [Figures 2-5]. Other findings included asymmetry and architectural distortion, each seen in three cases (4.3%). Microcalcifications alone were not seen in any of the patients. It was seen in association with mass, asymmetry, or architectural distortion.

MRI

Only 3 out of the 71 patients underwent dynamic contrast-enhanced MRI. 2 of the 3 patients had heterogeneous fibroglandular tissue as per ACR categories. The background parenchymal enhancement was mild in two cases and moderate in one case. Two cases presented as masses as shown in Figure 4, and one presented as suspicious non-mass enhancement.

USG

Similar to previous studies^[16,20,21] ultrasound was highly sensitive (sensitivity of 98.6%) in picking up the abnormalities and showed a negative scan in only one patient, who had architectural distortion on the mammogram. The commonest abnormality on USG was a mass with an irregular shape and non-circumscribed margins, the most common being spiculated margins. This was in congruence with previous studies done by Foxcroft LM et al. and Huang J et al.^[20,21] The majority of the masses showed either no posterior features or shadowing.

Mammography

Mammography is usually not performed below the age of 40 years because the breasts are dense and obscure findings. In our study, mammography was performed after clinical or ultrasound suspicion of malignancy. Most of the breasts were heterogeneously dense as per ACR categories of breast densities, as was seen in previous studies.^[17,18,20] The commonest presentation on mammography was mass, which was high density, had an irregular shape and non-circumscribed margins, and was most commonly spiculated. This was in line with previous studies.^{[8,13,20–23]}

Microcalcifications were seen in association with other findings in 32% of the cases, compared to 26.8%^[17]–49.1%^[18] seen in other studies. Isolated microcalcification was not seen in any of the cases in this study. One reason for the low number of microcalcifications seen in our study could be the lower number of HER2-enriched molecular subtypes.

MRI

MRI is not routinely performed at our institute for breast cancer before surgeries. It was performed only in three cases in our study and showed a mass in two cases and suspicious non-mass enhancement in one case. The sample size is too low to draw any conclusions.

Molecular Subtypes

As per various studies in the literature, the percentage of TNBCs has been reported to be higher in young patients. This has been corroborated in our study as well, which showed triple negative cancer (42.6%) to be the most common molecular subtype, followed by luminal type A (35.2%).^[12,24] Other studies have reported luminal type A^[14,25,26] and luminal type B^[27] as the most common subtypes in women younger than 40 years. The differences could be attributed to ethnic differences.

Limitations

Our study had certain limitations. Firstly, the sample size was small. Secondly, BRCA mutation testing was not done in any of the patients due to cost constraints. Thirdly, only three patients underwent dynamic MRI, MRI due to logistic issues which also highlights the concerns specific to resource constraint settings.