Article Text
PDF
Abstract
Encapsulated papillary carcinoma (EPC) of the breast is a distinct histological subtype characterised by malignant epithelial proliferation supported by fibrovascular stalks. Although EPC typically lacks myoepithelial cells, it shows indolent clinical course. The classification of EPC as an in situ, or invasive disease, remains a matter of debate.
Methods In this study, the authors investigated a panel of invasion-associated markers in a series of EPC and compared their expression with control groups of non-papillary ductal carcinoma in situ (DCIS) and conventional invasive carcinomas. The expression pattern of four matrix metalloproteinases (MMP-1, MMP-2, MMP-7 and MMP-9), transforming growth factor receptor beta, vascular endothelial growth factor (VEGF) and E-cadherin were assessed in the tumour cell and/or stromal tissue, and the results were analysed.
Results EPC showed higher expression levels of both MMP-1 and MMP-9 compared with DCIS, and no significant differences were observed between EPC and invasive carcinoma. Expression of MMP-2 and MMP-7 levels were similar in EPC and DCIS, but both showed lower levels compared with invasive tumours. EPC showed higher expression of E-cadherin and transforming growth factor receptor ß1 compared with both DCIS and invasive cancer. No difference in the stromal expression of MMPs or tumour expression of VEGF was detected.
Conclusion EPC exhibits an expression pattern of invasion-associated markers, which is intermediate in nature between DCIS and invasive cancer, providing further support of the unique biological features of EPC, and which may explain its clinically indolent behaviour.
- Human breast carcinoma
- encapsulated papillary carcinoma
- biology
- invasion-associated markers
- breast
- genetics
- oncology
- cytology
- diagnosis
- histopathology
- cancer
- molecular pathology
- molecular biology
- endocrinology
- breast cancer
- steroid receptors
- tamoxifen
- PCR
- immunocytochemistry
Statistics from Altmetric.com
- Human breast carcinoma
- encapsulated papillary carcinoma
- biology
- invasion-associated markers
- breast
- genetics
- oncology
- cytology
- diagnosis
- histopathology
- cancer
- molecular pathology
- molecular biology
- endocrinology
- breast cancer
- steroid receptors
- tamoxifen
- PCR
- immunocytochemistry
Introduction
Papillary carcinoma (PC) of the breast is a distinct histological subtype characterised by malignant epithelial proliferation supported by fibrovascular stalks.1–7 The term, encapsulated papillary carcinoma (EPC), has recently been introduced to define the cystic (intracystic/encysted) variant of PC.5–8 These tumours are typically circumscribed and often encapsulated (separated from surrounding mammary stroma by a fibrous capsule) and lack myoepithelial cells at their periphery.1–5 ,8 ,9 Although EPC has long been regarded as a form of ductal carcinoma in situ (DCIS), the absence of peripheral myoepithelial cells at the tumour–stromal interface,3 ,4 ,10–12 and the reported cases with local stromal/muscle infiltration8 and lymph node8 ,10 ,13 or distant metastases,14–16 may indicate that EPC is an invasive form of breast cancer with an expansile growth pattern and indolent behaviour.3 ,4 ,8 ,17
Degradation of basement membranes and the stromal extracellular matrix (ECM) are crucial steps for tumour invasion and metastasis. Several molecules have been shown to affect the integrity of basement membrane and structure of the surrounding stroma, and therefore, their expression is associated with tumour invasion. Matrix metalloproteinases (MMPs) are zinc-dependent endopeptidases involved in remodelling of the ECM, as well as cell–matrix and cell–cell contacts, facilitating detachment of epithelial cells from the surrounding tissue with subsequent movement of cells, or groups of cells, through ECM.18 ,19 MMPs play key functions in mammary gland morphogenesis and they are upregulated in breast cancer, where they stimulate tumourigenesis, cancer cell invasion and metastasis.20–25 Transforming growth factor β (TGF-ß), which is secreted by tumour cells and/or stromal cells within the peritumoral microenvironment, can contribute to diverse cellular processes such as, cell proliferation, differentiation, motility, adhesion, organisation and programmed cell death. TGF-β, which was initially identified by its potential to induce growth of mesenchymal cells in soft agar26 was found to inhibit the growth of normal and neoplastic epithelial cells.27 In normal cells, TGF- β generally enhances adhesion through increased matrix production and decreased proteolysis.28 In breast cancer, TGF-β is reported to have a biphasic effect, acting as a tumour suppressor in the initial phase, but stimulating invasion and metastasis at later stages.29 ,30 Angiogenesis is necessary for tumour growth, invasion and metastasis. Vascular endothelial growth factor (VEGF) has been implicated as a key mediator of angiogenesis.31 In breast cancer, expression of VEGF is associated with poorer outcome.32 E-cadherin is a transmembrane glycoprotein that mediates epithelial cell-to-cell adhesion. Loss of membranous expression of E-cadherin in ductal carcinoma of the breast is associated with lymph node and distant metastasis.33
In a previous study, we provided clinical and morphological evidence that EPC has pathological characteristics of invasive carcinoma but is associated with an excellent prognosis comparable with that of DCIS.8 In this study, we aimed to evaluate the expression of a panel of seven invasion-associated biomarkers—four MMPs(MMP-1, MMP-2, MMP-7 and MMP-9), TGF-ß, VEGF and E-cadherin in EPC and compare their expression with that observed in non-papillary DCIS and invasive tumours to further characterise the biological feature of EPC and its invasiveness potential.
Methods
Case selection
A retrospective search of the pathology database at Nottingham University Hospital NHS Trust was performed. Cases diagnosed during the period 1990–2004 as encysted/intracystic carcinoma (EPC) were retrieved. Seventeen cases of EPC with representative formalin-fixed paraffin-embedded tissue blocks were available. All patients were women, aged between 42 and 70 years (median 65 years). All cases were oestrogen-receptor positive and none of them had regional lymph node involvement. Tumour size ranged from 6 mm to 60 mm (median 20 mm). Only one case had nuclear grade 3 features, while the remaining cases were grade 1 or grade 2. All cases were reviewed by one pathologist (EAR), and a diagnosis of EPC was confirmed in all cases. In addition, 14 excision specimens of grade-matched non-papillary DCIS, and an unselected consecutive series of 140 conventional non-papillary invasive breast carcinomas of different histological types and grades that were prepared as tissue microarrays (TMA) were included as control groups. Cases were subject to histological review to confirm the diagnosis.
Immunohistochemistry
Tissue sections (4 μm) from formalin-fixed paraffin full-face (EPC and DCIS) and TMA (invasive tumours) were immunohistochemically stained using a panel of seven biomarkers (MMP-1, MMP-2, MMP-7, MMP-9, TGF-β1, VEGF and E-cadherin) using Leica Novolink polymer detection reagents. Primary antibodies, optimal concentration, antigen retrieval method and incubation time are summarised in table 1. Briefly, the slides were de-waxed and re-hydrated. After antigen retrieval, the slides were rinsed with Tris-buffered saline. Endogenous peroxidase activity was blocked using 100 μl of peroxidase block (Novolink reagent) for 5 min.
Details of antibodies used in the current study
After the slides were incubated with optimised primary antibodies, they were washed twice with Tris-buffered saline for 5 min each. Following post-primary block, 100 μl of polymer was added on the slides for 30 min each to enhance the penetration of subsequent polymer reagent and to detect any tissue-bound primary antibody. Then, 100 μl of 3, 3′-diaminobenzidine solution was applied. Sections were counterstained with Novocastra haematoxylin and coverslipped, mounted in DPX and air dried for 20 min. Positive and negative controls were included in each experiment. Positive control was used according to the supplier's datasheets, while negative control was stained without the primary antibody.
Immunohistochemically prepared slides were scored using a semi-quantitative visual approach. The slides were scanned for immunostaining evaluation using a light microscope and scored by two observers (MT and EJ) independently, and discordant cases were reviewed by a third observer (EAR). Expression of all markers was determined in the membranes (E-cadherin), cytoplasm (other markers) in the tumour cells. Both the percentage (0–100%) and the intensity of staining (0=negative, 1=weak, 2=moderate and 3=strong) were considered. Both, per cent and intensity, were used to develop the histochemical score (H-score) as follows: H-score=Intensity (0–3) × per cent of expression (0–100) to give a final score of 0–300. Stromal expression was scored as positive and negative.
This study was approved by Nottingham Research Ethics Committee 2 under the title of ‘Development of a molecular genetic classification of breast cancer.’
Statistical analysis
The difference in the expression of the various biomarkers between EPC, DCIS and invasive carcinoma was determined using Mann–Whitney U test (continuous variables) and χ2 and Fisher's exact tests (categorical). A cut-off of ≥10% expression is considered for positivity. Statistical significance was set at p<0.05. All statistical analyses were performed using SPSS V.16 software (SPSS Inc.).
Results
In this study, 17 EPC, 10 non-papillary DCIS and 138 conventional non-papillary invasive carcinomas were informative for all markers. All markers except E-cadherin showed cytoplasmic expression in the malignant cells with peritumoural stromal expression detected in some cases. The malignant cells showed E-cadherin membrane with or without cytoplasmic expression; only membranous expression was considered in the analysis.
Table 2 shows the expression of the biomarkers in EPC compared with DCIS and invasive carcinomas. EPC showed expression of MMP-1 and MMP-9 intermediate between DCIS and invasive carcinomas; higher than DCIS but lower than invasive carcinomas (figures 1 and 2). Although EPC showed lower expression of MMP-9 compared with invasive carcinomas, the difference was not statistically significant. Similarly, EPC showed higher expression of MMP-7 and lower expression of MMP-2 compared with invasive carcinomas; however, the difference between EPC and DCIS was not statistically significant. EPC showed higher expression of TGF-β1 compared with both DCIS and invasive carcinomas. No difference between EPC and either DCIS or invasive carcinomas was detected regarding the expression of VEGF and E-cadherin.
The percentages of tumour expression (%) of the seven biomarkers in malignant cells of encapsulated papillary carcinoma (EPC), non-papillary ductal carcinoma in situ (DCIS) and conventional invasive carcinoma of the breast
Expression of matrix metalloproteinase 9: (A) A case of encapsulated papillary carcinoma showing moderate cytoplasmic expression. (B) A case of low-grade non-papillary duct carcinoma in situ showing weak cytoplasmic expression. (C) A case of invasive carcinoma showing moderate to strong cytoplasmic expression of matrix metalloproteinase 9.
Expression of transforming growth factor receptor β (TGFβ): (A) A case of encapsulated papillary carcinoma showing strong cytoplasmic expression of TGFβ. (B) A case of low-grade non-papillary duct carcinoma in situ showing negative TGFβ expression. (C) A case of invasive carcinoma showing weak to moderate cytoplasmic expression of TGFβ.
Table 3 shows the stromal expression levels of the different markers. No statistically significant difference was detected between EPC and either DCIS or invasive carcinomas regarding stromal expression of the different markers.
Peritumour stromal expression in encapsulated papillary carcinoma (EPC), ductal carcinoma in situ (DCIS) and invasive carcinoma
Discussion
Although EPC was initially perceived as a variant of in situ carcinoma, recent studies have demonstrated that EPC consistently lacks a myoepithelial cell layer surrounding the tumour nodules.3 ,4 ,10–12 In addition, there is a potential, albeit infrequent, tendency to exhibit local stromal/muscle infiltration,8 and lymph node8 ,10 ,13 or distant metastatic spread.14–16 These observations have changed our perception of this unique type of breast cancer. Moreover, although some cases of EPC are surrounded by a thick fibrous capsule, we have recently demonstrated in an observational study that this capsule is mainly a reactive process rather than an expansion of the native basement membrane material surrounding the index breast ducts.8 Therefore, some authors, including our group, have concluded that EPC is an invasive form of breast cancer with an expansile growth pattern and indolent behaviour.3 ,4 ,8 ,17
This study investigated expression of a panel of seven biomarkers of known relevance to invasion and metastasis in mammary EPC and their expression levels compared with those observed in non-papillary DCIS and non-papillary conventional types of invasive carcinomas. Our aim was to further understand the biological status of EPC, and whether their indolent ‘in situ-like’ behaviour can be explained by difference in the expression of invasion and metastasis-associated molecules.
Our results showed that EPC has levels of the four members (MMP-1, MMP-2, MMP-7 and MMP-9) of the intermediate between those observed in DCIS and invasive carcinoma. However, invasive carcinoma showed higher expression of MMP-7 and MMP-9 compared with DCIS. Similar results have been reported by Gonzalez et al 34 using 95 DCIS and 63 patients with invasive carcinomas.
It was reported that TGF-β1 protein is responsible for collagen deposition and the scirrhous nature of some gastric carcinoma.35 Our results demonstrated that EPC express higher levels of TGF-β1 compared with both DCIS and invasive carcinomas. These results may indicate that TGF-β1 play a role in the development of the thick fibrous capsule seen around EPC nodules and supports our previous hypothesis that the surrounding capsule around EPC is a reactive rather than a thickened native basement membrane surrounding the ducts distended by the neoplastic papillary proliferation.8
VEGF, which is one of the most potent angiogenic factors, acts as a mitogen for endothelial cells and increases vascular permeability.36 Our findings indicated that VEGF expression is expressed in all cases of EPC and DCIS and in the majority of invasive carcinomas, and no significant differences were observed. These results are consistent with previous studies, which showed that around 90% of in situ carcinomas express VEGF,36 and no difference between DCIS and invasive carcinomas.37 Invasion is a complex multistep process, and stromal alterations similar to those seen in invasive breast carcinomas are already evident in association with some cases of DCIS and even in some benign lesions such as radial scars.38 Similarly, in this study, we have demonstrated that all cases of EPC and DCIS, and the majority of invasive carcinomas, show membranous expression of E-cadherin in the malignant cell population, and the difference was not statistically significant.
This study has limitations. The number of cases of EPC and grade-matched DCIS was limited. Therefore, caution should be taken when interpreting the results and drawing conclusions from the findings of this study. To compensate for the use of TMA in invasive carcinoma cases, a large number of cases was included.
In conclusion, this study provides further evidence that EPC is an indolent invasive carcinoma with biological features intermediate between in situ and invasive carcinomas. Not only the absence of myoepithelial cells, but also the expression of some invasion-associated proteins may support the invasive nature of EPC. This study showed that EPC express higher levels of TGF-β1 than conventional in situ and invasive carcinomas, and therefore, may provide further explanation for the reactive nature of the fibrous capsule surrounding EPC nodules. The difference between EPC and invasive carcinomas regarding expression of some invasion-associated proteins may explain, at least in part, the indolent nature of EPC.
Take-home messages
-
Encapsulated papillary carcinoma (EPC) is a rare type of breast cancer. EPC often lacks myoepithelial cells and it demonstrates a potential, albeit infrequent, tendency to exhibit invasive features.
-
EPC exhibits an expression pattern of invasion-associated markers, which is intermediate in nature between DCIS and invasive cancer, providing further support of the unique biological features of EPC, and which may explain its clinically indolent behaviour.
References
Footnotes
-
EAR and MT contributed equally to the project.
-
Competing interests None.
-
Patient consent Obtained.
-
Ethics approval The ethics approval was provided by the Nottingham Researchl Ethics Committee 2.
-
Provenance and peer review Not commissioned; externally peer reviewed.