Clinicopathologic Implications of EpCAM and Sox2 Expression in Breast Cancer

Original Study

Clinicopathologic Implications of EpCAM and Sox2 Expression in Breast Cancer Nehad M.R. Abd El-Maqsoud, Dalia M. Abd El-Rehim Abstract Epithelial cell adhesion molecule (EpCAM) and sex-determining region Y-box protein 2 (Sox2) are aberrantly expressed in breast cancer and other cancers. This is a retrospective study aimed at evaluating EpCAM and Sox2 expression in 19 cases of ductal carcinoma in situ (DCIS), 126 cases of invasive breast cancer (IBC), and 86 samples of matched lymph node (LN) metastasis using immunohistochemistry. Our findings highlight the role of EpCAM and Sox2 in breast cancer development and progression. Background: The purpose of this study was to investigate the clinicopathologic significance of EpCAM and Sox2 expression in breast cancer and to study their correlation during breast cancer progression. Patients and Methods: EpCAm and Sox2 expression were assessed using immunohistochemistry in ductal carcinoma insitu (DCIS), invasive breast cancer (IBC) and matched lymph node metastasis (LNM), if present. Results: EpCAM overexpression was found in 63.2% of DCIS, 72.2% of IBC and 74.4% of LNM. In IBC cases, EpCAM overexpression was associated with high grade (P < .001), large tumor size (P ¼ .051), poor Nottingham Prognostic Index (NPI) (P ¼ .006), histological tumor types (P ¼ .044) and the triple negative phenotype (P ¼ .008). LNM frequently reflected the expression phenotype of the matched primary tumors with no significant differences between LNM and their primary tumors (P ¼ .564). Sox2 expression was detected in 47.4%, 33.3% and 54.7% of DCIS, IBC and LNM respectively. In DCIS group, Sox2 expression was significantly associated with comedo type (P ¼ .037), negative ER (P ¼ .012) and PR (P ¼ .037) and the triple negative phenotype (P ¼ .006). In IBC cases, Sox2 expression showed significant associations with high grade (P ¼ .045), nodal spread (P ¼ .037), poor NPI (P ¼ .018) and the triple negative phenotype (P < .001). LNM showed significantly higher Sox2 expression rates than primary tumors (P < .001). Significant positive associations between EpCAM overexpression and Sox2 positivity in DCIS (P ¼ .027), IBC (P ¼ .001) and LNM (P < .001) were found. Conclusion: This study emphasized the potential role of EpCAM and Sox2 in breast carcinogenesis and revealed their involvement during breast cancer progression and LN metastases. Clinical Breast Cancer, Vol. 14, No. 1, e1-9 ª 2014 Elsevier Inc. All rights reserved. Keywords: Breast cancer, EpCAM, Immunohistochemistry, Sox2

Introduction Breast cancer is the most common cancer diagnosed among women, and it is one of the most common causes of cancer mortality worldwide.1 In Egypt, it ranks first, constituting 34.26% of female cancers according to the Egyptian National Cancer Institute registry,2 and is a more biologically aggressive disease compared with that in western countries.3 Breast cancer is a heterogeneous disease encompassing a wide variety of pathologic entities with different outcomes and responses Pathology Department, Faculty of Medicine, Minia University, El-Minia, Egypt Submitted: Apr 7, 2013; Revised: Sep 2, 2013; Accepted: Sep 4, 2013; Epub: Sep 27, 2013 Address for correspondence: Dalia M. Abd El-Rehim, PhD, Department of Pathology, Faculty of Medicine, Minia University, El-Minia, Egypt 61111 E-mail contact: [email protected]

1526-8209/$ - see frontmatter ª 2014 Elsevier Inc. All rights reserved. http://dx.doi.org/10.1016/j.clbc.2013.09.006

to therapy.4 It is initiated and maintained by a subpopulation of cells, tumor-initiating cells (TICs) or cancer stem-like cells (CSCs) that share many biological properties similar to those of embryonic stem cells. CSCs have the ability to self-renew and differentiate into multiple cell lineages, thereby generating tumor heterogeneity. The presence of CSCs in breast cancer is likely one of the main causes of tumor progression, metastasis, recurrence, and resistance to therapy.5-7 Therefore, detection of these cells in breast cancer tissues might be a reasonable approach to predicting metastasis and recurrence. Furthermore, elimination of CSCs could be a necessary step for an effective cure for breast cancer. Several genes and transcriptional factors are involved in maintaining the stemness phenotype of embryonic stem cells and have been shown to be associated with tumors and TICs/CSCs. Epithelial cell adhesion molecule (EpCAM)8,9 and sex-determining region Y-box protein 2 (Sox2)10,11 have recently been found to have

Clinical Breast Cancer February 2014

- e1

EpCAM and Sox2 Expression in Breast Cancer tumorigenic potential and have been reported as putative TIC/CSC markers in several malignancies, including breast cancer. EpCAM is a 40-kDa type I transmembrane glycoprotein composed of a large extracellular domain, one transmembrane region, and a small intracellular domain of 26 amino acids. It is abundantly expressed by the majority of human epithelial carcinomas, including colorectal, breast, lung, prostate, ovarian, and endometrial cancers,12,13 and it has attracted major attention as a potential target for antibody-based cancer therapies. Recently, a fully human monoclonal antibody against EpCAM with antitumor activity in EpCAM-positive metastatic breast cancer has been developed.14 EpCAM is unique among the family of cell adhesion molecules. Recent insights revealed that it is involved in promoting cancer cell proliferation, migration, and invasiveness.15 EpCAM overexpression negatively modulates E-cadherinemediated adhesions by disrupting the link between a-catenin and F-actin, thereby loosening cell-cell adhesion.16 In addition, its overexpression is an important regulator of cancer cell proliferation and pluripotency through its capacity to signal in combination with members of the Wnt pathway and its ability to activate reprogramming factors such as Sox2, octamer-binding transcription factor 4 (Oct-4), and Nanog.9,17 Sox2, a member of the SOX family of proteins, is one of the crucial transcription factors that regulate self-renewal and pluripotency in embryonic stem cells. Stemness pathways may reactivate during tumorigenesis, with aberrant Sox2 expression in several tumors, including breast cancer.11,18-20 In breast cancer, it is involved in early-stage carcinogenesis as well as in later events such as invasion and metastasis, supporting the notion that activation of Sox2 is part of the malignant progression.11,21,22 There is little information in the literature regarding the role of EpCAM and Sox2 expression during breast cancer development and progression. Knowledge about the role of EpCAM and Sox2 in the process of breast carcinogenesis, progression, and metastasis needs further elucidation. The purpose of the current study was to study the clinicopathologic significance of EpCAM and Sox2 expression in human breast cancer and to further investigate their correlation during breast cancer progression.

Material and Methods Tissue Specimens

e2

-

Formalin-fixed and paraffin-embedded surgical specimens of 19 cases of ductal carcinoma in situ (DCIS), 126 cases of invasive breast cancer (IBC), and 86 samples of matched lymph node metastasis (LNM) were collected randomly and diagnosed from 2007 to 2012 from the archives of the Pathology Department, Minia University Hospital. Sixty cases (47.6%) of invasive carcinomas showed a concomitant in situ component along with the invasive component. The diagnosis and histopathologic results were reviewed according to established criteria using the World Health Organization criteria.4 Clinical and pathologic information was obtained from the patients’ medical records. The patients’ mean age  standard deviation was 46.52  5.04 (range, 39-56 years) and 53.37  8.26 (range, 35-69 years) for DCIS and IBC, respectively. Data regarding clinicopathologic features are summarized in Table 1.

Clinical Breast Cancer February 2014

Table 1 Clinicopathologic Characteristics of the 19 Patients With Ductal Carcinoma Insitu and 126 Patients With Invasive Breast Cancer Clinicopathologic Characteristics

DCIS (n [ 19)

IBC (n [ 126)

No.

%

No.

%

<50

12

63.2

38

30.2

50

7

36.8

88

69.8

Premenopausal

14

73.7

46

36.5

Postmenopausal

5

26.3

80

63.5

Age (years)

Menopausal Status

pT Stage pT1

e

e

7

5.6

pT2

e

e

73

57.9

pT3

e

e

46

36.5

Grade Low

3

15.8

G1

8

6.3

Intermediate

6

31.6

G2

47

37.3

10

52.6

G3

71

56.3

Positive

e

e

40

31.7

1-3 positive

e

e

43

34.1

>3 positive

e

e

43

34.1

High LN stage

NPI Good

e

e

8

6.3

Moderate

e

e

53

42.1

Poor

e

e

65

51.6

9

47.4

Ductal

107

84.9

10

52.6

Lobular

12

9.5

Other types

7

5.6

Histologic Type Noncomedo Comedo ER Negative

9

47.4

56

44.4

Positive

10

52.6

70

55.6

PR Positive

10

52.6

68

54

Negative

9

47.4

58

46

Negative

11

57.9

75

59.5

Positive

8

42.1

51

40.5

Yes

5

26.3

36

28.6

No

14

73.7

90

71.4

HER2/neu

Triple negative

Abbreviations: DCIS ¼ ductal carcinoma in situ; ER ¼ estrogen receptor; HER2 ¼ human epidermal growth factor receptor 2; IBC ¼ invasive breast cancer; LN ¼ lymph node; NPI ¼ Nottingham Prognostic Index; PR ¼ progesterone receptor.

Nottingham Prognostic Index The Nottingham Prognostic Index (NPI) is a well-established and widely used method of predicting survival after operable primary breast cancer. The NPI was calculated using the following equation: NPI ¼ 0.2  tumor size (cm) þ grade (1-3) þ LN stage (1-3). According to the index, cases were designated as good,

Nehad M.R. Abd El-Maqsoud, Dalia M. Abd El-Rehim moderate, and poor. A score of  3.40 indicates a good prognosis, 3.41 to 5.40 a moderate prognosis, and > 5.40 a poor prognosis.23

detection of EpCAM and Sox2 expression, respectively. Negative control is obtained by adding phosphate-buffered saline instead of primary antibodies.

Immunohistochemical Staining A streptavidin-biotin immunoperoxidase complex procedure was used for staining. In brief, 4-mm-thick sections were deparaffinized with xylene and rehydrated with graded alcohol solutions. After inactivation of endogenous peroxidase with 0.3% hydrogen peroxide in methanol for 30 minutes, antigen retrieval was achieved by microwave treatment in sodium citrate buffer (0.01 M, pH 6.0) for 10 minutes. For EpCAM, antigen retrieval in Proteinase K (Novocastra) for 10 minutes at 37 C was performed. Tissue sections were then incubated with primary antibodies for EpCAM (clone UV1D9, diluted 1:50; Novocastra, Leica Microsystems, Buffalo Grove, IL), Sox2 (clone 57CT23.3.4, diluted 1:200; Abcam, Cambridge, MA), estrogen receptor (ER) (clone 1D5, diluted 1:80; Dako, Carpinteria, CA), progesterone receptor (PR) (clone PgR 636, diluted 1:100; Dako, Carpinteria, CA) and HER2 (diluted 1:250; Dako, Carpinteria, CA) for 2 hours. After incubation with a biotinylated secondary antibody for 30 minutes at room temperature, visualization of the reaction was detected using 3,3’-diaminobenzidine (DAB) as a chromogen. Finally, sections were counterstained with hematoxylin and dehydrated, cleared, and mounted. Each staining batch included positive and negative control sections.

Positive and Negative Controls Samples of human breast cancer and squamous cell carcinoma of the lung were used as positive controls for immunohistochemical

Scoring System and Cutoff Points EpCAM expression was evaluated according to Spizzo et al by calculating a total immunostaining score as the product of a proportion score and an intensity score.24 The proportion score describes the estimated proportion of positively stained tumor cells (0, none; 1, < 10%; 2, 10%-50%; 3, 51%-80%; 4, > 80%). The intensity score represents the estimated staining intensity (0, no staining; 1, weak; 2, moderate; 3, strong). The total score ranges from 0 to 12, with EpCAM overexpression defined as a total score > 4. Regarding Sox2 expression, cases were considered positive when any neoplastic cell displayed definite nuclear staining.21 Cases were considered positive for either ER or PR when > 10% of tumor cells showed nuclear expression. Only cases with continuous strong membranous HER2 staining of > 30% of cells (3þ) were considered positive.25

Statistical Analysis To test associations between categorical variables, Chi-square (c2) and Fisher exact tests were conducted. Multivariate analysis was performed using logistic regression to determine whether any of the variables tested could be independently associated with EpCAM or Sox2 expression. Differences in EpCAM and Sox2 expression between primary tumors and their corresponding

Figure 1 Representative Immunohistochemical Staining of Epithelial Cell Adhesion Molecule (EpCAM) in Preinvasive and Invasive Breast Carcinoma and Matched Lymph Node Metastasis (LNM) Overexpression of EpCAM (Expression Score > 4) in (A) Ductal Carcinoma in Situ (DCIS) and (B) Invasive Breast Cancer (IBC). (C) EpCAM Overexpression in Concomitant DCIS and IBC. (D and E) Invasive Ductal Carcinoma and its LNMs Show EpCAM Overexpression. (Diaminobenzidine [DAB] as Chromogen; Hematoxylin Counterstain. Original Magnification 3100 and 3400)

Clinical Breast Cancer February 2014

- e3

e4

DCIS (no [ 19) No. Clinicopathologic Characteristics

P Value

EpCAM Negative/ Over Low (%) expression (%)

IBC (n [ 126) P Value

Sox2

No.

P Value

EpCAM Negative/ Over Low (%) expression (%)

Negative (%) Positive (%)

P Value

Sox2 Negative (%) Positive (%)

Age (years) <50

12

4 (33.3)

8 (66.7)

50

7

3 (42.9)

4 (57.1)

4 (28.6)

10 (71.4)

.526

5 (41.7)

7(58.3)

5 (71.4)

2(28.6)

.220

38

10 (26.3)

28 (73.7)

88

26 (29.5)

62 (70.5)

.713

22 (57.9)

16 (42.1)

62 (70.5)

26 (29.5)

.170

Menopausal Status Premenopausal

14

Postmenopausal

5

3 (60)

.237

2 (40)

6 (42.9)

8 (57.1)

4 (80)

1 (20)

.184

46

11 (23.9)

35 (76.1)

80

25 (31.3)

55 (68.8)

.380

28 (60.9)

18 (39.1)

56 (70)

24 (30)

.295

pT Stage pT1

e

e

e

e

e

e

7

3 (42.9)

4 (57.1)

6 (85.7)

1(14.3)

pT2

e

e

e

e

e

e

73

25 (34.2)

48 (65.8)

53(72.6)

20 (27.4)

pT3

e

e

e

e

e

e

46

7 (15.2)

39 (84.8)

25(54.3)

21 (45.7)

.245

.051

.066

Grade 3

2 (66.7)

1 (33.3)

3 (100)

0 (0)

G1

8

6 (75)

2 (25)

7 (87.5)

1 (12.5)

6

3 (50)

3 (50)

4 (66.7)

2 (33.3)

G2

47

19 (40.4)

28 (59.6)

36 (76.6)

11 (23.4)

10

2 (20)

8 (80)

3 (30)

7 (70)

G3

71

10 (14.1)

61 (85.9)

41 (57.7)

30 (42.3)

High

.073

<.001

Low Intermediate

.045

LN Stage Positive

e

e

e

e

e

e

40

15 (37.5)

25 (62.5)

1-3 Positive

e

e

e

e

e

e

43

13 (30.2)

30 (69.8)

29 (67.4)

14 (32.6)

>3 Positive

e

e

e

e

e

e

43

7 (16.3)

36 (83.7)

23 (53.5)

20 (46.5)

.089

32 (80)

8 (20)

.037

NPI Good

e

e

e

e

e

e

8

5 (62.5)

3 (37.5)

7 (87.5)

1 (12.5)

Moderate

e

e

e

e

e

e

53

19 (35.8)

34 (64.2)

41 (77.4)

12 (22.6)

Poor

e

e

e

e

e

e

65

11(16.9)

54 (83.1)

36 (55.4)

29 (44.6)

5 (55.6)

4 (44.4)

.109

7 (77.8)

2 (22.2)

.037

2 (20)

8 (80)

3 (30)

7 (70)

.006

.018

Histologic Type Noncomedo Comedo

9 10

Ductal

107

26 (24.3)

81 (75.7)

70 (65.4)

37 (34.6)

Lobular

12

7 (58.3)

5 (41.7)

.044

10 (83.3)

2 (16.7)

Other types

7

2 (28.6)

5 (71.4)

4 (57.1)

3 (42.9)

56

9 (16.1)

47 (83.9)

70

26 (37.1)

44 (62.9)

.394

ER Positive

9

Negative

10

2 (22.2)

7 (77.8)

5 (50)

5 (50)

2 (20)

8 (80)

5 (55.6)

4 (44.6)

.210

2 (22.2)

7 (77.8)

8 (80)

2 (20)

.012

.009

28 (50)

28 (50)

56 (80)

14 (20)

37 (54.4)

31 (45.6)

<.001

PR Positive

10

Negative

9

.109

3 (30)

7 (70)

7 (77.8)

2 (22.2)

.037

68

12 (17.6)

56 (82.4)

58

23 (39.7)

35 (60.3)

.006

47 (81)

11(19)

.002

EpCAM and Sox2 Expression in Breast Cancer

Clinical Breast Cancer February 2014

Table 2 Correlation of EpCAM and Sox2 Expression and Clinicopathologic Characteristics of 19 Patients With Ductal Carcinoma Insitu and 126 Patients With Invasive Breast Cancer

Test of significance: c2 and Fisher exact test. P values  .05 are considered statistically significant. Abbreviations: DCIS ¼ ductal carcinoma in situ; EpCAM ¼ epithelial cell adhesion molecule; ER ¼ estrogen receptor; HER2 ¼ human epidermal growth factor receptor 2; IBC ¼ invasive breast cancer; LN ¼ lymph node; NPI ¼ Nottingham Prognostic Index; PR ¼ progesterone receptor.

<.001 19 (21.1)

23 (63.9) 13 (36.1) .008 32 (88.9)

59 (65.6) 31 (34.4)

4 (11.1) 36

90

.006 4 (28.6)

5 (100) 0 (0) 5 (100)

7 (50)

0 (0)

7 (50)

5 Yes

No

Triple negative

14

.047

0.311 4 (50) 4 (50) 8

8 (72.7) 3 (27.3) 11

Negative

Positive

HER2/neu

10 (71.4)

.463 3 (37.5)

6 (54.5) 5 (45.5)

5 (62.5)

Negative (%) Positive (%) Negative/ Over Low (%) expression (%) Clinicopathologic Characteristics

71 (78.9)

.124 13 (25.5)

29 (38.7) 46 (61.3) .199 51 (68)

40 (78.4) 11 (21.6)

24 (32) 75

51

Negative/ Over Low (%) expression (%)

P Value Sox2 P Value No.

EpCAM

DCIS (no [ 19) Table 2 Continued

38 (74.5)

Negative (%) Positive (%)

P Value No.

EpCAM

IBC (n [ 126)

Sox2

P Value

Nehad M.R. Abd El-Maqsoud, Dalia M. Abd El-Rehim lymph node metastases (LNMs) were assessed using the McNemar test. Statistical analyses were carried out using SPSS, version 11.0 (SPSS Inc, Chicago IL). Probabilities were considered statistically significant at P  .05.

Results EpCAM Expression in DCIS, IBC, and Corresponding LNMs Positive expression of EpCAM was mainly localized to the cell membrane. EpCAM expression was dichotomized into 2 groups, EpCAM overexpressing (total score > 4) and EpCAM nonoverexpressing (total score, 0-4). Using these appropriate cutoff points, EpCAM overexpression was seen in 12 of 19 (63.2%) cases of DCIS, 91 of 126 (72.2%) cases of IBC, and 64 of 86 (74.4%) cases of LN metastasis. We noticed that all EpCAM-positive IBC maintained the same expression scores in in situ components as in invasive components (Fig. 1A-E). Table 2 summarizes the associations between EpCAM expression and clinicopathologic features in DCIS cases. Significant association with the triple-negative phenotype was found when all triplenegative DCIS displayed EpCAM overexpression (P ¼ .047). No significant associations were seen in relation to other clinicopathologic features. Table 2 summarizes the associations between EpCAM expression and clinicopathologic variables in IBC cases. Significant positive associations were identified between EpCAM expression and histologic grade (P < .001), large tumor size (P ¼ .051) and poor NPI (P ¼ .006). A significant difference in EpCAM expression in relation to histologic tumor type was also found (P ¼ .044) when invasive lobular carcinoma showed significantly lower EpCAM overexpression rates compared with other types. There was no significant association between EpCAM overexpression and LN stage, patient age, or menopausal status. An inverse significant association with ER (P ¼ .009) and PR (P ¼ .006) expression was found, whereas no significant association in relation to HER2 expression was seen. A highly significant positive correlation between EpCAM and the triple-negative phenotype was seen (P ¼ .008). Associations with clinicopathologic features were further assessed by multivariate analysis. Tumor grade was an independent predictor of EpCAM overexpression (P ¼ .030) (Table 3). To assess whether the expression of EpCAM undergoes changes during breast cancer progression, 86 pairs of primary tumors and their corresponding metastases to LNs were compared (Table 4). EpCAM expression in LNM was highly correlated with that of the primary tumor; 74 of 86 (86.05%) of the pairs had the same score levels in both groups. However, 7 of 86 (8.14%) changed EpCAM status from overexpression to non-overexpression and 5 of 86 (5.81%) changed EpCAM status from non-overexpression to overexpression. No significant difference was found between primary tumors and their corresponding LNMs regarding EpCAM expression (P ¼ .564).

Sox2 Expression in DCIS, IBC, and Corresponding LNM Although normal breast epithelial cells displayed no Sox2 staining, breast carcinoma cells were strongly positive for Sox2. Nuclear Sox2 expression was detected in 9 of 19 (47.4%), 42 of 126 (33.3%), and 47 of 86 (54.7%) cases in DCIS, IBC, and LNM respectively.

Clinical Breast Cancer February 2014

- e5

EpCAM and Sox2 Expression in Breast Cancer Table 3 Associations of EpCAM and Sox2 Expression and Clinicopathologic Variables in Multivariate Analysis Invasive Breast Cancer EpCAM

Sox2

Clinicopathologic Characteristics

P Value

Odds Ratio

95% CI

P Value

Odds Ratio

95% CI

Age Menopausal status pT Stage Grade LN Stage NPI Histologic Type ER PR HER2/neu Triple negative

.461 .596 .914 .030 .148 .341 .707 .560 .442 .041 .071

2.056 0.606 1.059 4.152 2.271 0.409 0.930 1.684 0.571 3.484 0.164

0.303-13.961 0.095-3.863 0.372-3.019 1.145-15.056 0.747-6.902 0.065-2.582 0.636-1.360 0.292-9.697 0.137-2.382 1.054-11.514 0.023-1.167

.959 .800 .352 .969 .533 .831 . 687 .972 .764 .485 .028

0.953 1.253 1.573 1.025 1.388 1.217 0.918 0.970 0.810 1.582 0.133

0.151-5.995 0.220-7.127 0.607-4.077 0.292-3.599 0.495-3.897 0.199-7.447 0.604-1.394 0.186-5.070 0.203-3.222 0.436-5.735 0.022-0.808

Abbreviations: CI ¼ confidence interval; EpCAM ¼ epithelial cell adhesion molecule; HER2 ¼ human epidermal growth factor receptor 2; LN ¼ lymph node; NPI ¼ Nottingham Prognostic Index; PR ¼ progesterone receptor; Sox2 ¼ sex-determining region Y-box protein 2.

In IBC cases, the same Sox2 expression state was noticed in both the invasive component and the in-situ component (Fig. 2A-E). In the DCIS group, Sox2 expression showed a significant positive association with the presence of comedo necrosis (P ¼ .037), negative ER and PR status (P ¼ .012 and P ¼ .037, respectively) and the triple-negative phenotype (P ¼ .006). There was no significant relationship between Sox2 expression and other clinicopathologic variables. An association between positive Sox2 expression and high-grade DCIS was found; however, such a relation did not reach the significant level (P ¼ .073) (Table 2). Table 2 summarizes the associations between Sox2 expression and clinicopathologic characteristics in IBC. Positive Sox2 expression was significantly associated with high grade (P ¼ .045), advanced nodal spread (P ¼ .037), poor NPI (P ¼ .018), negative ER status (P < .001), negative PR status (P ¼ .002), and the triplenegative phenotype (P < .001). No significant associations were seen in relation to other clinicopathologic features. A positive association between Sox2 expression and larger tumor size was noticed; however, it did not reach the significant level. On multivariate analysis, the triple-negative phenotype was the only independent factor affecting Sox2 expression (P ¼ .028) (Table 3). To further study whether Sox2 plays a role in the development of LNM, we analyzed Sox2 expression in 86 pairs of metastatic LNs and their corresponding primary tumors. Sox2 expression was detected in all LNMs from Sox2-positive primary tumors. Interestingly, Sox2 was additionally detected in 13 of 86 (15.12%)

LNs derived from primary tumors devoid of Sox2 expression (Table 4). Thus, the frequency of Sox2 positivity in LN samples was significantly higher than that in primary tumors (P <.001). Table 5 shows associations between EpCAM and Sox2 in DCIS, IBC, and LNM. Strong positive correlations between EpCAM overexpression and Sox2 positivity were found in DCIS (P ¼ .027), IBC (P ¼ .001), and LNM (P < .001). Most of Sox2-positive lesions showed EpCAM overexpression.

Discussion In breast cancer, EpCAM expression has received increasing attention as an oncogene and a potential prognostic factor. It promotes cancer cell proliferation, migration, and invasiveness,26,27 which may explain why EpCAM overexpression is associated with poor prognosis in patients with breast cancer.24,28,29 The present study showed EpCAM overexpression in 63.2% of DCIS cases, indicating that it may be an early tumor-promoting event in breast carcinoma. EpCAM was abundantly expressed in high grade and comedo-type DCIS, but the relationship was insignificant probably because of the small size of the study group. Our data also showed EpCAM overexpression in 72.2% of IBC cases. A wide range of EpCAM overexpression rates in breast cancer (17.6%-100%) has been reported in the literature.25,28-30 These variations could be attributed to different immunohistochemical protocols, scoring systems, and cutoff points used and different populations examined in different studies.

Table 4 Comparison of the Expression of EpCAM and Sox2 in 86 Pairs of Primary Breast Cancers and Lymph Node Metastasis Positive Expression Rates EpCAM Sox2

Change in Expression Pattern

Primary (%)

Metastases (%)

P > M (%)

66 (76.7) 34 (39.5)

64 (74.4) 47 (54.7)

7 (8.14) 0

a

P [ M (%)b

P < M (%)c

P Value

74 (86.05) 73 (84.88)

5 (5.81) 13 (15.12)

.564 <.001

Test of significance: McNemar test. P values  .05 are considered statistically significant. Abbreviations: EpCAM ¼ epithelial cell adhesion molecule; P ¼ primary; M ¼ metastasis; Sox2 ¼ sex-determining region Y-box protein 2. a Higher expression scores in the primary tumors. b Equal expression. c Higher expression in LNM.

e6

-

Clinical Breast Cancer February 2014

Nehad M.R. Abd El-Maqsoud, Dalia M. Abd El-Rehim Figure 2 Immunohistochemical Expression of sex-Determining Region Y-box Protein 2 (Sox2) in Preinvasive and Invasive Breast Carcinoma and Matched Lymph Node Metastasis (LNM). Sox2 Immunoreactivity in (A) Ductal Carcinoma in situ (DCIS) and (B) Invasive Breast Cancer (IBC). (C) Sox2-Positive Expression in Concomitant DCIS and IBC. (D and E) Sox2 Expression in a Primary Tumor and the Corresponding LNM. (Diaminobenzidine [DAB] as Chromogen; Hematoxylin Counterstain. Original Magnification 3100 and 3400)

Consistent with previous reports,24,28,29 the current study identified a significant positive association between EpCAM overexpression and less-differentiated tumors, larger tumors, poor NPI, and negative hormone receptor status. On multivariate analysis, tumor grade was an independent predictor of EpCAM overexpression. Of particular interest, the current work reported lower EpCAM expression rates in lobular carcinoma compared with ductal carcinoma. The same finding was reported in previous studies.12,29 Martowicz et al recently demonstrated that the role of EpCAM in breast cancer strongly depends on the phenotype of cancer cells, ie, their status of epithelial-to-mesenchymal transition.31 Cancer cells with an epithelial phenotype need EpCAM overexpression to support proliferation, invasion, and tumor angiogenesis. In contrast, mesenchyme-like tumor cells typically have low EpCAM expression

and they grow independently of EpCAM signaling. These authors also added that the phenotype of lobular breast cancer resembles that of cancer cells undergoing epithelial-to-mesenchymal transition. Accordingly, targeting EpCAM in patients with lobular breast cancer might even result in an adverse effect, and great attention should be given to studying the efficacy of EpCAM-targeting agents in the subgroup of patients with lobular breast cancer. Recently, so-called triple-negative (ER-, PR-, and HER-2 negative) breast cancer has received great attention because of its resistance to existing targeted therapies such as endocrine therapy or trastuzumab.32 The current study identified a significant positive correlation between EpCAM overexpression and this phenotype in both IBC and DCIS cases. The predominant expression of EpCAM in triple-negative breast cancer is of great clinical relevance because

Table 5 Associations Between EpCAM and Sox2 Expression in Different Pathologic Lesions Sox2 Lesion

No.

DCIS

7 12 35 91 22 64

Primary tumor Metastases

Negative (%) EpCAM EpCAM EpCAM

Negative/low overexpression Negative/low overexpression Negative/low overexpression

6 4 31 53 17 22

(85.7) (33.3) (88.6) (58.2) (77.3) (34.4)

Positive (%) 1 8 4 38 5 42

(14.3) (66.7) (11.4) (41.8) (22.7) (65.6)

P Value .027 .001 <.001

Test of significance: c2 test and Fisher exact test. P values  .05 are considered statistically significant. Abbreviations: DCIS ¼ Ductal carcinoma in situ; EpCAM ¼ epithelial cell adhesion molecule; Sox2 ¼ sex-determining region Y-box protein 2

Clinical Breast Cancer February 2014

- e7

EpCAM and Sox2 Expression in Breast Cancer

e8

-

it may be a basis for treating this subgroup of patients with adecatumumab, an antibody directed against EpCAM.14 This study evaluated EpCAM expression in LNM relative to matched primary breast carcinomas. A great proportion of LNM lesions showed EpCAM overexpression and they frequently reflected the expression phenotype of their primary tumors. In agreement with our findings, EpCAM expression in synchronous and metachronous metastases was reported to be highly correlated with that of their primary tumors in several tumors, including breast cancer.12 However, another study reported EpCAM upregulation in metastases compared with the matched primary breast cancer.30 Taken together, these findings suggest the role of EpCAM in promoting LN and distant metastases and validate EpCAM as a therapeutic target in metastatic breast cancer. Sox2 is a reprogramming factor that induces pluripotency and is found to play a crucial role in the promotion of tumor initiation and progression.11,19 Consistent with previous studies, the present work identified Sox2 expression in DCIS cases with a higher expression rate than seen in invasive carcinomas, suggesting a role of Sox2 in the initial stages of breast carcinogenesis. In this group, Sox2 expression was significantly associated with comedo type, negative hormone receptor status, and the triple-negative phenotype. Positive association was also seen with high-grade DCIS; however, such a relation did not reach the significant level, probably because of the restricted number of cases. The data presented here need to be confirmed in a larger cohort of DCIS cases with various grades and histologic types. In IBC cases, Sox2 expression was identified in about one third of cases, which is consistent with a previous study.22 However, Sox2 expression was reported with a lower expression rate (16.7%) in another cohort of LN-negative breast carcinomas.21 This difference could be attributed to a different cutoff for the definition of Sox2 positivity or, more reasonably, is the inclusion of LN-positive breast cancers causing more Sox2-positive cases in the current series. In keeping with previous studies,21,22 the present study reported an association between Sox2 expression and poor prognostic features of breast cancer such as high grade, advanced nodal spread, a trend of large tumor size, poor NPI, negative hormone receptor status and the triple-negative phenotype. The same significant association with the triple-negative phenotype was further confirmed on multivariate analysis, suggesting that Sox2 expression is particularly enhanced with this molecular subtype. However, Lengerke et al could not verify any correlation between Sox2 expression and the triple-negative phenotype.22 They attributed their finding to the underrepresentation of this particular phenotype in their cohort of postmenopausal patients. Sox2-mediated induction of tumor invasiveness and spread may be a common event in different malignancies.22,33-35 To further investigate this issue in breast cancer, we evaluated Sox2 expression in primary tumors relative to their corresponding MLNs. We found that all Sox2-positive primary tumors kept their Sox2-expressing phenotype in their corresponding LNMs. This finding may suggest that Sox2-bearing cancer cells have more probability and ability to metastasize to the LNs and supports the notion that Sox2 plays a crucial role in breast cancer invasiveness and spread. Interestingly, a significantly higher Sox2 expression in LNMs compared with their respective primary tumors was found. A similar finding

Clinical Breast Cancer February 2014

was also reported in a previous study,22 reflecting an increase in Sox2-expressing cells in these metastases. It is worthwhile to mention that bias induced by varied tissue processing, immunohistochemical procedures, or interobserver variations cannot be an explanation for the higher Sox2 expression seen in metastatic LN disease compared with the primary tumors because both primary tumor and matched LN specimens were uniformly fixed, processed, archived, immunohistochemically stained, and then evaluated by the same pathologists. A recent cell line study provided evidence that overexpression of EpCAM promotes tumorigenesis through the upregulation of reprogramming factors (Oct-4, Nanog, and Sox2) in HCT116 colon cancer cells, SAS oral carcinoma cells, FaDu pharynx carcinoma cells, and Hep3B hepatoma cells, whereas its downregulation inhibited the expression of these factors, thereby suppressing tumor initiation, progression, and invasiveness.9 Therefore, it might be assumed that both EpCAM and Sox2 drive the malignant progression in various tumors. In this context, the current study hypothesized the presence of a possible correlation between EpCAM and Sox2 during different phases of breast cancer progression. Interestingly, strong positive correlations between EpCAM overexpression and Sox2 positivity were found in DCIS, IBC, and LNM.

Conclusion This study highlighted the important role of EpCAM and Sox2 in breast carcinogenesis and demonstrated their involvement in breast cancer progression and the development of LNM. In the longer term, detailed characterization of EpCAM and Sox2 in a larger scale study of breast cancer is likely to provide powerful diagnostic and prognostic markers and attractive therapeutic targets as well.

Clinical Practice Points  This study emphasized the role of EpCAM and Sox2 in breast

carcinogenesis and revealed their implication in breast cancer progression and LN metastases.  In the longer term, detailed characterization of EpCAM and Sox2 in a larger scale of breast cancer is warranted to validate these findings.  This study may also, suggest the role of EpCAM and Sox2 in breast cancer as powerful diagnostic and prognostic markers and attractive therapeutic targets as well.

References 1. Siegel R, Naishadham D, Jemal A. Cancer statistics, 2012. CA Cancer J Clin 2012; 62:10-29. 2. Mokhtar N, Gouda I, Adel I, eds. Cancer pathology registry 2003-2004 and time trend analysis. Department of Pathology, National Cancer Institute, 2007. 3. El-Bolkainy MN, Nouh MA, El-Bolkainy TN. Breast cancer. In: Topographic Pathology of Cancer. 3rd ed. El-Asdekaa Press, Cairo, Egypt: 2005:87. 4. Tavassoli FA, Devillee P. World Health Organization Classification of Tumours. Pathology and Genetics of Tumours of the Breast and Female Genital Organs. Lyon, France: IARC Press; 2003. 5. Al-Hajj M, Wicha MS, Benito-Hernandez A, et al. Prospective identification of tumorigenic breast cancer cells. Proc Natl Acad Sci U S A 2003; 100:3983-8. 6. Campbell LL, Polyak K. Breast tumor heterogeneity: cancer stem cells or clonal evolution? Cell Cycle 2007; 6:2332-8. 7. Al-Ejeh F, Smart CE, Morrison BJ, et al. Breast cancer stem cells: treatment resistance and therapeutic opportunities. Carcinogenesis 2011; 32:650-8.

Nehad M.R. Abd El-Maqsoud, Dalia M. Abd El-Rehim 8. Imrich S, Hachmeister M, Gires O. EpCAM and its potential role in tumorinitiating cells. Cell Adh Migr 2012; 6:30-8. 9. Lin CW, Liao MY, Lin WW, et al. Epithelial cell adhesion molecule regulates tumor initiation and tumorigenesis via activating reprogramming factors and epithelial-mesenchymal transition gene expression in colon cancer. J Biol Chem 2012; 287:39449-59. 10. Chen Y, Shi L, Zhang L, et al. The molecular mechanism governing the oncogenic potential of SOX2 in breast cancer. J Biol Chem 2008; 283:17969-78. 11. Leis O, Eguiara A, Lopez-Arribillaga E, et al. Sox2 expression in breast tumours and activation in breast cancer stem cells. Oncogene 2012; 31:1354-65. 12. Spizzo G, Fong D, Wurm M, et al. EpCAM expression in primary tumour tissues and metastases: an immunohistochemical analysis. J Clin Pathol 2011; 64:415-20. 13. Patriarca C, Macchi RM, Marschner AK, et al. Epithelial cell adhesion molecule expression (CD326) in cancer: a short review. Cancer Treat Rev 2012; 38:68-75. 14. Schmidt M, Scheulen ME, Dittrich C, et al. An open-label, randomized phase II study of adecatumumab, a fully human anti- EpCAM antibody, as monotherapy in patients with metastatic breast cancer. Ann Oncol 2010; 21:275-82. 15. Trzpis M, McLaughlin PMJ, de Leij LMFH, et al. Epithelial cell adhesion molecule: more than a carcinoma marker and adhesion molecule. Am J Pathol 2007; 171:386-95. 16. Winter MJ, Nagelkerken B, Mertens AE, et al. Expression of Ep-CAM shifts the state of cadherin-mediated adhesions from strong to weak. Exp Cell Res 2003; 285: 50-8. 17. Gires O. Lessons from common markers of tumor-initiating cells in solid cancers. Cell Mol Life Sci 2011; 68:4009-22. 18. Saigusa S, Mohri Y, Ohi M, et al. Podoplanin and SOX2 expression in esophageal squamous cell carcinoma after neoadjuvant chemo-radiotherapy. Oncol Rep 2011; 26:1069-74. 19. Chen S, Xu Y, Chen Y, et al. SOX2 gene regulates the transcriptional network of oncogenes and affects tumorigenesis of human lung cancer cells. PLoS ONE 2012; 7:e36326. 20. Ruan J, Wei B, Xu Z, et al. Predictive value of Sox2 expression in transurethral resection specimens in patients with T1bladder cancer. Med Oncol 2013; 30: 445-53. 21. Rodriguez-Pinilla SM, Sarrio D, Moreno-Bueno G, et al. Sox2: a possible driver of the basal-like phenotype in sporadic breast cancer. Mod Pathol 2007; 20:474-81.

22. Lengerke C, Fehm T, Kurth R, et al. Expression of the embryonic stem cell marker SOX2 in early-stage breast carcinoma. BMC Cancer 2011; 11:42-51. 23. Blamey RW, Ellis IO, Pinder SE, et al. Survival of invasive breast cancer according to the Nottingham Prognostic Index in cases diagnosed in 1990-1999. Eur J Cancer 2007; 43:1548-55. 24. Spizzo G, Went P, Dirnhofer S, et al. High Ep-CAM expression is associated with poor prognosis in node-positive breast cancer. Breast Cancer Res Treat 2004; 86: 207-13. 25. Niemiec JA, Adamczyk A, Małecki K, et al. Relationships between immunophenotype, Ki-67 index, microvascular density, Ep-CAM/P-cadherin, and MMP-2 expression in early-stage invasive ductal breast cancer. Appl Immunohistochem Mol Morphol 2012; 20:550-60. 26. Münz M, Kieu C, Mack B, et al. The carcinoma-associated antigen EpCAM upregulates c-myc and induces cell proliferation. Oncogene 2004; 23:5748-58. 27. Osta WA, Chen Y, Mikhitarian K, et al. EpCAM is overexpressed in breast cancer and is a potential target for breast cancer gene therapy. Cancer Res 2004; 64: 5818-24. 28. Schmidt M, Hasenclever D, Schaeffer M, et al. Prognostic effect of epithelial cell adhesion molecule overexpression in untreated node-negative breast cancer. Clin Cancer Res 2008; 14:5849-55. 29. Agboola AJ, Paish EC, Rakha EA, et al. EpCAM expression is an indicator of recurrence in basal-like breast cancer. Breast Cancer Res Treat 2012; 133:575-82. 30. Cimino A, Halushka M, Illei P, et al. Epithelial cell adhesion molecule (EpCAM) is overexpressed in breast cancer metastases. Breast Cancer Res Treat 2010; 123:701-8. 31. Martowicz A, Spizzo G, Gastl G, et al. Phenotype dependent effects of EpCAM expression on growth and invasion of human breast cancer cell lines. BMC Cancer 2012; 12:501-15. 32. Brunello A, Borgato L, Basso U, et al. Targeted approaches to triple-negative breast cancer: current practice and future directions. Curr Med Chem 2013; 20:605-12. 33. Bae KM, Parker NN, Dai Y, et al. E-cadherin plasticity in prostate cancer stem cell invasion. Am J Cancer Res 2011; 1:71-84. 34. Neumann J, Bahr F, Horst D, et al. SOX2 expression correlates with lymphnode metastases and distant spread in right-sided colon cancer. BMC Cancer 2011; 11:518. 35. Girouard SD, Laga AC, Mihm MC, et al. SOX2 contributes to melanoma cell invasion. Lab Invest 2012; 92:362-70.

Clinical Breast Cancer February 2014

- e9