Breast Cancer Res Treat (2013) 142:311–322 DOI 10.1007/s10549-013-2756-y
PRECLINICAL STUDY
Expression and clinical significance of carcinoembryonic antigenrelated cell adhesion molecule 6 in breast cancers Julia Y. S. Tsang • Ying Kin Kwok • Kit Wing Chan • Yun-Bi Ni • Wan Ning Vanessa Chow • Kwok Fai Lau • Mu-Min Shao • Siu Ki Chan Puay-Hoon Tan • Gary M. Tse
•
Received: 17 September 2013 / Accepted: 25 October 2013 / Published online: 2 November 2013 Ó Springer Science+Business Media New York 2013
Abstract Carcino-embryonic antigen-related cell adhesion molecule 6 (CEACAM6), one of the members of human carcino-embryonic antigens, is a multifunctional regulatory protein involved in various cellular processes in cancers. Its role in malignant transformation and the clinical significance has been extensively studied in colonic and pancreatic cancers. However, relatively few studies have been done on breast cancers. In the current study, CEACAM6 expression in two independent cohorts of invasive breast cancers were evaluated immunohistochemically and correlated with clinico-pathological features, biomarker profiles and patient survival. In the
Electronic supplementary material The online version of this article (doi:10.1007/s10549-013-2756-y) contains supplementary material, which is available to authorized users. J. Y. S. Tsang Y. K. Kwok K. W. Chan Y.-B. Ni G. M. Tse (&) Department of Anatomical and Cellular Pathology, Prince of Wales Hospital, Ngan Shing Street, Shatin, NT, Hong Kong e-mail:
[email protected] W. N. V. Chow K. F. Lau School of Life Sciences, The Chinese University of Hong Kong, Hong Kong, Hong Kong M.-M. Shao Department of Pathology, Shenzhen Affiliated Hospital, Guangzhou University of Traditional, Chinese Medicine, Shenzhen, China S. K. Chan Department of Pathology, Kwong Wah Hospital, Hong Kong, Hong Kong P.-H. Tan Department of Pathology, Singapore General Hospital, Singapore, Singapore
primary cohort, CEACAM6 expression was detected in 37.1 % (312/840) of primary invasive cancers. It was positively correlated with HER2 (p \ 0.001). Concordantly, HER2-OE subtype showed the highest CEACAM6 expression (62.7 %) among all molecular subtypes; whereas, other subtypes also showed substantial CEACAM6 expression (21.8–37.5 %). Interestingly, a significantly worse overall survival was found in high pN stage HER2 positive cancers with CEACAM6 positivity (logrank = 4.452, p = 0.035) and this could be validated in an independent cohort. Additionally, HER2 signaling was found to induce SMAD3 phosphorylation and CEACAM6 expression in a cell line model. Likewise, in the primary tumors, a positive association was found between HER2 and SMAD3 phosphorylation in CEACAM6 positive cancers (p = 0.012). Overall, CEACAM6 was widely expressed in different molecular subtypes, but highest and significantly in HER2-OE breast cancer. Within this group, CEACAM6 was associated with adverse high nodal stage patient outcome. Given the wide expression of CEACAM6 in all breast cancers, its roles as prognostic marker and therapeutic target warrant further evaluation. Keywords Breast cancer CEACAM6 HER2 Immunohistochemistry Nodal stages
Introduction Carcino-embryonic antigen-related cell adhesion molecule 6 (CEACAM6) is one of the members of human carcinoembryonic antigens (CEA), belonging to a group of highly glycosylated membrane proteins of the immunoglobulin superfamily. CEACAMs function as intercellular adhesion molecules mediating homotypic and heterotypic cell–cell
123
312
Breast Cancer Res Treat (2013) 142:311–322
interaction. They play important roles in numerous cellular processes, ranging from organization of tissue architecture, regulation of signal transduction, and cellular differentiation to microbial receptors. These multifunctional regulatory proteins have also been shown to be involved in malignant transformation. CEACAM6 is expressed in granulocytes and epithelia from various organs [1, 2]. Its role in malignant transformation has been extensively studied in colonic [3–6] and pancreatic cancers [3, 7–11]. CEACAM6 is also overexpressed in other type of cancers, including breast cancer [3, 12–15]. It has generally been identified as a marker for decreased survival in patients with colonic cancers [5], intrahepatic cholangiocarcinomas [12], and gastric cancers [14]. CEACAM6 functions through homotypic or heterotypic binding with other members of the CEACAM family [16]. In addition, CEACAM6 also shows heterotypic interactions with integrin receptors [17]. Through these interactions, it modulates cancer cell differentiation, apoptosis, cell growth, and therapeutic response [4, 8, 10, 18, 19]. CEACAM6 silencing in pancreatic cancer cells has been reported to increase their susceptibility to caspase mediated anoikis (anchorage-dependent apoptosis) and down-regulation of AKT cell survival pathway under anchorage-independent conditions [8, 18]. CEACAM6 can act as an inducer of cellular proliferation in lung adenocarcinoma cells by interfering with the contact-inhibiting signal triggered by CEACAM1-4L, thus inducing undifferentiated anchorage independent cell growth [19]. In addition to cellular survival and growth, over-expression of CEACAM6 has been found to promote in vivo metastatic ability, treatment resistance, and cellular invasiveness via a c-SRC-dependent increase of AKT activity [7, 10]. Pancreatic cancer cell lines stably over-expressing CEACAM6 show markedly increased chemo-resistance to gemcitabine [8]. In the breast, CEACAM6 is expressed in cancer cells [3, 20] but not normal breast tissue [21]. Expression of
CEACAM6 in atypical duct hyperplasia is strongly associated with the development of invasive breast cancer, suggesting a contributory role in breast cancer progression [22]. CEACAM6 is also implicated in breast cancer invasion and treatment resistance; antibodies inhibiting the binding of CEACAM6 positive breast cancer cells to endothelial cells reduce tumor invasion [20]. CEACAM6 has been shown to be up-regulated in Tamoxifen-resistant or estrogen-deprived breast cancer cell lines [23, 24]. These cell lines are not only refractory to hormonal therapy, but also show enhanced invasiveness, increased migratory capacity [24, 25], and anchorage independent growth [23]. Furthermore, CEACAM6 expression predicts breast cancer recurrence following adjuvant tamoxifen [23]. Although cumulative evidence suggested significant roles of CEACAM6 in breast cancers, the data were derived from relatively small cohorts. Breast cancer is heterogeneous, and different histotypes show significantly different CEACAM6 expression [3]. Molecular classification basing on gene expression profiling has classified breast cancers into distinct biological subgroups with different prognoses [26]. It is possible that CEACAM6 expression may differ among these molecular groups. In this study, CEACAM6 expression in different breast cancer molecular subgroups was examined and its association with different clinic-pathological features, biomarker expression, and patient outcome was evaluated. The impact on survival was assessed according to REMARK criteria [27].
Materials and methods Patients data The histologic files of the three involved institutions were searched for breast carcinoma. The primary cohort included all consecutive cases with excision specimens from a
Table 1 Antibodies for immunohistochemical analysis of TMA Markers
Company
Clone
Dilution
Antigen retrieval
Incubation condition (min, °C)
Assessment
Cut-off
ER
Ventana
SP1
Pre-diluted
EDTA pH8
32, 37
N
C1 %
PR
Ventana
1E2
Pre-diluted
EDTA pH8
32, 37
N
C1 %
Ki67 HER2
Ventana Ventana
30–9 4B5
Pre-diluted Pre-diluted
EDTA pH8 EDTA pH8
32, 37 16, 37
N M
[14 % 3?
CK5/6
Dako
D5/6 B4
1:40
EDTA pH8
32, 37
C, M
C5 %
CK14
Neomarker
LL002
1:100
EDTA pH8
32, 37
C, M
C5 %
c-Kit
Dako
104D2
1:300
EDTA pH8
32, 37
C, M
C5 %
EGFR
Ventana
3C6
Pre-diluted
EDTA pH8
32, 37
M
C5 %
P63
Ventana
4A4
Pre-diluted
EDTA pH8
32, 37
N
C5 %
CEACAM6
Abcam
9A6
1:100
EDTA pH8
32, 37
C, M
[25 %
pSMAD3
Abcam
EP823Y
1:200
EDTA pH8
32, 37
N
[20
123
Breast Cancer Res Treat (2013) 142:311–322
single institution collected over a period of 7 (2003–2009) years while the independent cohort was from the other two involved institutions with cases collected over a period of 2 (2003–2004) and 4 (2002–2005) years. All the specimens were formalin fixed, paraffin embedded, and routinely processed. The 4 micron slides were stained with hematoxylin and eosin (H&E) and reviewed with the diagnoses confirmed. The tumors were graded using modified Bloom and Richardson grading [28] and the histologic diagnosis and histotype confirmed (WHO [29]). In addition, lymphovascular invasion (LVI) and the presence of extensive in situ components (EIC) (in situ component occupying more than 25 % of the total tumor volume) were also evaluated as present or absent. Patient details and clinical information including patients’ age, tumor size, lymph node involvement (presence or absence of lymph node metastasis), pN stage, pT stage, treatment received and patient outcome data were retrieved from the medical records. For the outcome data, overall survival (OS) was defined as the time interval from the date of initial diagnosis to the date of breast cancer related death. Disease free survival (DFS) was defined as the duration from the date of initial diagnosis to the first detection of breast cancerspecific relapse or death.
313 Table 2 Clinical features of the primary and validation cohorts Clinico-pathological features
Primary cohort N (%)
Validation cohort N (%)
Grade
1
134 (16.0)
41 (13.7)
2
342 (40.7)
114 (38.0)
3
364 (43.3)
145 (48.3)
Total
840
300
Absent
577 (70.9)
156 (74.3)
Present
237 (29.1)
54 (25.7)
Total Absent
814 662 (78.8)
210 210 (82.4)
Present
178 (21.2)
45 (17.6)
Total
840
255
Negative
421 (51.7)
139 (47.8)
Positive
393 (48.3)
152 (52.2)
Total
814
291
1
360 (43.1)
106 (37.5)
2
423 (50.6)
150 (53.0)
3
43 (5.1)
18 (6.4)
4
10 (1.2)
9 (3.1)
LVI
EIC
LN status
pT
pN
Tissue microarray construction Cellular areas of the tumors on H&E stained slides were chosen and the corresponding areas were taken from the paraffin blocks for tissue microarray (TMA) construction. Two 1.5 mm tissue cores were obtained from each case. The TMAs were assembled with a tissue arrayer (Beecher Instruments, Silver Springs, MD). Twenty-seven composite TMA blocks, each containing 54 tissue cores, were constructed. Serial 4 micron sections were cut and transferred to Superfrost Plus glass slides (Menzel-Glaser, Germany). One section from each tissue array block was stained with H&E to confirm the presence of representative tumors in the TMA blocks. Immunohistochemistry and scoring The TMA slides were stained for the following antibodies: estrogen receptor (ER), progesterone receptor (PR), human epidermal growth factor receptor 2 (HER2), Ki67, p63, c-kit, epidermal growth factor receptor (EGFR), CK14, CK5/6, pSMAD3, and CEACAM6. Immunohistochemical staining of all markers was performed by BenchMark XT automated slide-staining instrument (Ventana, Arizona, USA) with Ultraview Universal DAB Detection Kit (Ventana, Arizona, USA) after deparaffinization, rehydration, and antigen retrieval. After primary antibody incubation, the sections
Histotypes
Treatmentb
CEACAM6
Age
Size
Total
836
283
0
415 (51.4)
139 (47.8)
1
238 (29.5)
85 (29.2)
2
95 (11.8)
38 (13.1)
3 Total
59 (7.3) 807
29 (10.0) 291
IDC
740 (88.1)
256 (85.3)
ILC
19 (2.3)
10 (3.3)
Othersa
81 (9.6)
34 (11.4)
Total
840
300
No treatment
85 (11.0)
2 (2.5)
HT
198 (25.6)
15 (18.8)
CT
135 (17.5)
23 (28.7)
HT ? CT
355 (45.9)
40 (50.0)
Total
773
80
Low
528 (62.9)
285 (65.0)
High
312 (37.1)
105 (35.0)
Total
840
300
Mean
55.5
51.9
SD
12.4
13.5
Range Mean
23–97 2.61
22–94 2.82
SD
1.42
1.78
Range
0.3–10.2
0.2–13.9
IDC invasive ductal carcinoma, ILC invasive lobular carcinoma, HT hormonal therapy, CT chemotherapy a
Other histotypes including mucinous carcinoma, metaplastic carcinomas, medullary carcinomas, and other miscellaneous subtypes
b
Treatment data were only available from one institute for the validation cohort
123
314
were incubated with anti-mouse horseradish peroxidaselabeled polymer (Roche, Arizona, USA) for 30 min at room temperature, and then developed with diaminobenzidine. All slides were counterstained with hematoxylin. Details of the antibodies, antigen retrieval, and staining conditions were listed in Table 1. The TMA slides were scored for the intensity of staining in the nucleus, cytoplasm, or membrane according to different antibodies by three of the authors blinded to the clinical information and the staining results of other markers. For ER, PR, Ki67, pSMAD3, and p63, the reactivity assessed was nuclear. For CEACAM6, CK5/6, CK14, and c-kit, the reactivity was both cytoplasmic and membranous. For EGFR and HER2, the reactivity assessed was membranous. The staining was considered positive when there was moderate or strong nuclear immunoreactivity in C1 % of tumor cells for ER and PR [30]. HER2 immunostaining was evaluated as 0, 1?, 2?, and 3? according to the ASCO guideline [31]. Tumors with 3? staining were considered positive. For Ki67, high expression was defined as C14 % of tumor cells showing moderate to strong immunoreactivity [32]. The mean CEACAM6 expression was 27 % and this was used as a reference to define cutoff, and this cutoff value was similar to previously described [25 % of tumor cells showing moderate to strong immunoreactivity as cutoff [6]. Positivity of pSMAD3 was defined by score [20 (the mean of
Breast Cancer Res Treat (2013) 142:311–322
staining score generated by multiplication of intensity (0— no staining, 1—weak, 2—moderate, 3—strong staining) and percentage of positive cells (0–100 %), the score theoretically ranged from 0 to 300). For all other markers, a C5 % cut-off was used to define positivity. Any discrepancies were resolved by reviewing at a multi-head microscope and a consensus reached. The tumors were also classified into the five different molecular subtypes by immunohistochemical expression as surrogate as follows [33, 34]: Luminal A: ER? and/or PR?, HER2-, CK5/6±, and Ki67 \ 14 % Luminal B: ER? and/or PR?, CK5/6±, HER2?, or Ki67 C 14 % HER2 over-expressed (HER2-OE): ER-, PR-, HER2?, CK5/6 ± Basal like breast cancers (BLBC): ER-, PR-, HER2-, (triple negative), CK5/6?, and/or EGFR? Unclassified: ER-, PR-, HER2-, (triple negative), CK5/6-, and EGFRCell culture and treatment Human breast cancer cell lines SK-BR-3 and MCF7 were cultured in RPMI1640 supplemented with 10 % fetal bovine serum (FBS). For transforming growth factor beta
Fig. 1 Representative staining of CEACAM6 in breast cancers a, b (9400) and EIC components c, d (9100)
123
Breast Cancer Res Treat (2013) 142:311–322
315
Table 3 Correlation of CEACAM6 expression with clinico-pathological features in primary cohort
Table 3 continued CEACAM 6 N (%)
CEACAM 6 N (%)
EIC
LN status
Histotypes
Treatment
Mean
55.6
55.4
55.5
0.717
SD
12.1
12.8
12.4
Range
23–97
27–91
Mean
2.58
2.67
2.61
SD
1.39
1.48
1.42
Range
0.3–9.5
0.3–10.2
Invasive cancer
528
312 (37.1)
840
0.218
DCIS
29
11 (27.5)
40
Total
557
323
880
1
88
46 (34.3)
134
2
206
136 (39.8)
342
pM stage was not included as no distant metastasis was observed in the initial diagnosis
3
234
130 (35.7)
364
Bold value is statistically significant (p \ 0.05)
Size
528
312
840
Absent
370
207 (35.9)
577
Present
145
92 (30.8)
237
Total
515
299
814
Absent
428
234 (35.3)
662
Present Total
100 528
78 (43.8) 312
178 840
Negative
273
148 (35.1)
421
154 (39.1)
393
1
239 512 230
302 130 (36.1)
0.887a
a
Total
Total
pN
p value
p value
Positive
pT
Total
Total
Invasive cancers
LVI
High
High
Age
Grade
Low Low
0.429
Spearman correlation
Table 4 Correlation of CEACAM6 expression with molecular subtypes in primary cohort CEACAM 6 N (%) Low
0.038 Lum A
0.235
Lum B
HER2-OE
814 360
a
0.642
TNBC
247
153 (38.3)
400
277
157 (36.2)
434
Total Negative
524 401
310 236 (37.1)
834 637
Positive
123
74 (37.6)
197
Total
524
310
834
Negative
496
263 (34.7)
759
Positive
28
47 (62.7)
75
Total
524
310
834
Negative
428
278 (39.4)
706
160 (37.8)
423
3
27
16 (37.2)
43
4
6
4 (40.0)
10
Total
526
310
836
0
269
415 238
Positive Total
Positive
1
142
2
57
38 (40.0)
95
3
39
20 (33.9)
59
Total
507
300
807
IDC
462
278 (37.6)
740
ILC
13
6 (31.5)
19
Others
53
28 (34.6)
81
Total
528
312
840
No treatment
54
31 (36.4)
85
HT
122
76 (38.4)
198
CT HT ? CT
90 213
45 (33.3) 142 (40.0)
135 355
Total
479
294
773
0.578a Unclassified
Total
Negative
263
BLBC
High
Positive
2
146 (35.2) 96 (40.3)
0.499
32 (25.0)
128
Total
524
96
310
834
Negative
481
298 (38.3)
779
Positive
43
12 (21.8)
55
Total
524
310
834
Negative
471
290 (38.1)
761
53 524
20 (27.4) 310
p value 0.536
0.896