Stromal Expression of CD10 in Breast Carcinoma and Its Association with Known Prognostic Factors—A Tissue Microarray-Based Study

Introduction

In India, breast cancer accounts for 18.5% of all newly diagnosed cancers.^[1] As it is an epithelial malignancy, various epithelial prognostic factors have been extensively studied.^[2,3] Cancer cells interact with stromal cells and involve stimulatory and inhibitory factors that regulate cell adhesion and migration. Recent studies suggest that extracellular proteinase regulates growth factors and cytokines contributing to cancer invasion and metastasis.^[4] The matrix molecules play an important role in malignancies and knowledge of such stromal markers will help identification of potential therapeutic targets.^[2,5,6]

The continuous and bilateral molecular crosstalk between normal epithelial cells and stromal cells is affected by several factors like growth factors, chemokines, extracellular matrix (ECM) molecules as well as ECM-modifying enzymes, including matrix metalloproteinases (MMPs) secreted by tumor cells or stromal cells and they represent the most prominent group of proteinases associated with tumor progression.^[2]

MMPs are a family of metallopeptidases that cleave protein components of ECM and thereby play a central role in tissue remodeling.^[7] They also play an important role in secretion of active transforming growth factor β that promotes tumorigenesis. The cleavage products of matrix components (e.g., collagen, laminin) also have chemotactic activity for tumor cells, and thus help in tumor cell migration.^[2]

CD10 is a 90 to 110 kDa zinc-dependent metalloproteinase, also called as “common acute lymphoblastic leukemia antigen”. It is expressed in normal tissues like myoepithelial cells of breast and salivary glands and pulmonary alveolar epithelial cells, but it is not expressed in stroma of normal breast tissue.^[8–10] Stromal CD10 is expressed with higher frequency in malignancy as opposed to borderline and benign phyllodes tumor of breast.^[7,11–13] A knowledge of CD10 role in breast cancer will thus help in identifying specific signals that promote growth, dedifferentiation, invasion, angiogenesis, metastasis, and can serve as target for better therapeutic plan and management.^[2,5,6]

Materials and Methods

A prospective study of prognostic indicators in 62 cases of carcinoma breast was undertaken from August 2015 to June 2017 at a tertiary care hospital in South India with the approval of Institutional ethical committee (JSSMC/PG/1435/2015-16). The formalin fixed paraffin embedded sections of modified radical mastectomy specimens of invasive breast carcinoma (No Special Type) were studied extensively for all the prognostic factors and were correlated with the clinical details (collected with patient's consent).

Nottingham combined histological grading (Elston–Ellis modification of Scarff–Bloom–Richardson grading system)^[14] and tumor, node, metastasis staging according to American Joint Committee on Cancer (AJCC) classification (8th edition)^[15] were followed. Nottingham's Prognostic index (NPI) was calculated using the formula: NPI = 0.2 x S + N + G, where, S is the size of the index lesion in centimeters, N is the node status: 0 nodes = 1, 1-4 nodes = 2, >4 nodes = 3 and G is the grade of tumor: Grade I =1, Grade II =2, Grade III =3. Based on the score, the patients were divided into six NPI groups (►Table 1).^[16]

Manual Tissue Microarray Construction

Hematoxylin & eosin (H&E)-stained sections of formalin fixed paraffin embedded tumor tissue were screened to identify representative areas of malignancy. The corresponding areas on the paraffin blocks were marked and tissue cores of 4 mm were punched out using tissue punch. For each case, two cores were taken from the area of interest. Tissue microarray (TMA) blocks were constructed manually in rows and columns with the tissue cores. Sections from TMA blocks were stained with H&E and immunohistochemical (IHC) stains and studied to verify tumor cell content and to note IHC expression profile.^[17–19] The sequence of steps in construction of a manual TMA paraffin block is illustrated in ►Fig. 1.

Close

Fig. 1

The sequence of steps in construction of a manual tissue microarray paraffin block.

Export to PPT

Results

A total of 62 cases of breast carcinoma were included in the study. Most commonly affected patients were in their fifth decade (32.2%) with a mean age of 51.3 years. Tumors were predominantly of grade III (56.5%) and grade II (40.3%) with only 3.2% of grade I tumors. Tumor necrosis, peritumoral lymphocytic infiltration, lymphovascular invasion (LVI), desmoplasia, and perineural invasion were seen in 58.1, 62.9, 80.6, 85.5, and 17.7% of cases respectively.

Pathological staging (based on AJCC system of classification) of majority of tumors was pT2 (67.7%) followed by 22.6% cases of pT3 and 9.7% cases of pT1. About 66.1% of tumors were ER and PR positive, while 56.5% showed HER2/neu overexpression and 19.4% were triple negative. Based on molecular subtyping, majority (41.9%) were of luminal B type, while there were 24.2% of luminal A, 19.4% of triple negative, and 14.5% of Her2 types. NPI showed 46.8% of cases under Moderate prognostic groups (MPG I & MPG II), 27.4% of very poor prognostic group (VPG), 22.6% of poor prognostic group (PPG), and 3.2% of cases of good prognostic group (GPG).

Scoring of CD10 by IHC and Its Correlation with Known Prognostic Factors

In normal breast tissue, CD10 expression was seen only in myoepithelial cells lining the ducts, while in 82.3% of invasive ductal carcinomas (NST), CD10-positive cells were seen infiltrating surrounding stroma. Of these, 48.4% of cases showed strong positivity (score 2) (►Fig. 2), while 33.9% of cases showed weak positivity (score 1; ►Fig. 3). CD10 scores were correlated with known prognostic factors like tumor size, grade, necrosis, LVI, perineural invasion, lymph node metastasis, and prognostic markers—ER, PR, HER2, molecular subtypes, prognostic groups. ►Table 3 shows the correlation and statistical significance of CD10 scores with these prognostic factors.

Close

Fig. 2

CD10 immunohistochemistry (IHC) staining displaying strong staining intensity in >30% of stromal cells in invasive breast carcinoma—score 2 (CD10 IHC, x200).

Export to PPT

Close

Fig. 3

CD10 immunohistochemistry (IHC) staining showing moderate staining intensity in <30% of stromal cells of invasive breast carcinoma- Score 1 (CD10 IHC, x200).

Export to PPT

Follow-Up

Among the 62 cases of invasive carcinoma breast, we could follow-up 41 cases on the recurrence and breast-cancer-associated deaths for a period of 3 years. The remaining cases were lost to follow-up. Two cases succumbed to disease-specific death by the end of 1.5 and 2 years, respectively, and both cases were strongly CD10 positive and belonged to the Nottingham's VPG. Recurrence was recorded in eight cases, of which five belonged to VPG and three to PPG with strong CD10 expression. Thus, this showed a significant trend toward CD10 overexpression and decreased survival as well as disease recurrence in the patients.

Discussion

The structural and functional integrity of breast is maintained by mutual efforts of the epithelial cells and stromal components. Interaction of tumor cells with stromal cells involves various factors that regulate tumor growth, adhesion, migration and thus affects invasiveness and metastatic potential of cancer cells. This modifying effect of matrix molecules gives matrix a cardinal role in cancer proliferation and metastasis.^[20] More knowledge regarding the role of stroma in breast cancer development and progression will help in identification of new prognostic markers that can serve as potential therapeutic targets.^[2,5]

CD10 is metallopeptidase that plays an important role in tissue remodeling by cleavage of protein components of ECM.^[7] In normal breast, CD10 is expressed only by myoepithelial cells that line the outer layer of the ducts. In invasive breast carcinoma, enzymatic activity of CD10 is upregulated; this leads to an accumulation of CD10–cleaved peptides that inhibit epithelial cell differentiation and result in epithelial mesenchymal transition and malignant proliferation.^[2]

Tumor size is one of the significant prognostic factors in breast carcinoma and there is increased incidence of axillary lymph node metastasis and decreased survival with increasing size of tumor.^[21,22] Majority of patients in the present study had large tumors when they sought medical attention and there was a significant positive correlation with the overexpression of CD10 and tumor size^[5,17,23] however, Iwaya et al^[24] and Arora et al^[25] have not noticed this correlation.

Researchers have observed a significant correlation between CD10 expression and increasing grade^[25] and lymph node involvement.^{[4,5,17,23,24]} Complimenting this finding, a higher intensity of CD10 expression has been noted more often in grade III tumors, N2/N3 tumors compared to grade II tumors, and N0 / N1 tumors. Studies have reported that factors like necrosis, desmoplasia, and LVI are associated with aggressive nature, poor prognosis, and early recurrence/death.^[26-28] The present study has shown a significant association of CD10 expression with these factors, which implies CD10 as a poor prognostic indicator.

Though majority of breast tumors are hormone dependent or express HER2 and have appropriate treatment targeting these pathways, triple-negative tumors form a significant number. Studies have reported a statistically significant correlation between CD10 expression and ER/PR negativity^[2-5,7] and HER2/neu positivity,^[2,3] but data correlating CD10 expression with molecular subtypes is limited. In the present study, CD10 expression was seen more often in triple negative and HER2 molecular subtypes of tumors compared to luminal A and luminal B subtypes and CD10 expression was of higher intensity in triple negative and HER2 types are the aggressive molecular variants, while in luminal A and Luminal B subtypes, expression was of lower intensity. An increased expression was also noted inPPGs.^[2] Studies have reported that a strong positive stromal CD10 expression in breast carcinoma was associated with decreased long-term disease-specific survival and overall survival.^[7,17,23,24]

CD10 can be used as a potential target for novel therapies as it is involved in cleavage of doxorubicin, a critical component in many chemotherapy protocols. As doxorubicin is known for its toxicity, CPI0004Na a prodrug of doxorubicin that can be cleaved by CD10 may improve the antitumor efficacy profile with reduced toxicity in tumors that overexpress CD10.^[8]

Our study shows a statistically significant correlation of CD10 overexpression with poor prognostic factors including larger tumor size, lymph node metastasis, higher histological grade, necrosis, LVI, ER/PR negativity, HER2/neu positivity, triple-negative molecular subtypes, and the PPGs, signifying that CD10 can be used as an independent poor prognostic marker in carcinoma breast.

Stromal CD10 expression in breast cancer is not static and changes with neoadjuvant anthracycline based chemotherapy. A stable or decrease in CD10 expression indicates complete or partial clinical response, while an increase in CD10 expression appears to correlate with poor clinical response. Thus, incorporating CD10 as a biomarker along with known factors into a prognostic index will not only help to predict clinical outcome more accurately but also help in predicting treatment failure in patients receiving neoadjuvant chemotherapy.

Conclusion

Our results are found parallel with other investigators, but it is necessary to further unveil the complex crosstalk between the cancer cells and other cells of the tumor microenvironment to establish high-quality targeted therapy. Larger prospective clinical studies with a longer duration of follow-up and study of survival are essential to validate our findings and help to envisage new therapeutic strategies.