- Email: [email protected]
Cytokeratin 7 and cytokeratin 20 expression in colorectal adenocarcinomas
Pathology – Research and Practice 207 (2011) 156–160
Contents lists available at ScienceDirect
Pathology – Research and Practice journal homepage: www.elsevier.de/prp
Original Article
Cytokeratin 7 and cytokeratin 20 expression in colorectal adenocarcinomas Reyhan Bayrak ∗ , Sibel Yenidünya, Hacer Haltas Department of Pathology, School of Medicine, Fatih University, 06540 Emek, Ankara, Turkey
a r t i c l e
i n f o
Article history: Received 21 July 2010 Received in revised form 28 September 2010 Accepted 15 December 2010 Keywords: Colon carcinoma Cytokeratin 7 Cytokeratin 20 Immunohistochemistry
a b s t r a c t Cytokeratin 7 (CK7) and cytokeratin 20 (CK20) are low molecular weight cytokeratins. The expressions of CK7 and CK20 have been studied in various primary and metastatic carcinomas. Their expression patterns may help to distinguish the site of origin of metastatic carcinomas. We investigated the expressions of CK7 and CK20 in 196 cases of colorectal carcinoma. Paraffin sections of 196 colonic adenocarcinomas were randomly selected, retrieved, and immunostained for CK7 and CK20 with a standard avidin–biotin complex method. CK7 was expressed in 34/196 (17.3%) and CK20 in 159/196 (81.1) cases of colorectal adenocarcinoma. CK7−/CK20+ had the greatest proportion (65.8%) in colorectal carcinomas. The CK7+/CK20+ immunophenotype was identified in 30/196 (15.3%), CK7−/CK20− in 33/196 (16.9%), and CK7+/CK20− in 4/196 (2%) colon adenocarcinomas. The CK7 and CK20 expression patterns were different in colorectal carcinomas according to histological grade, location of the tumor, and lymph node metastasis. CK20 positivity was more common in low grade carcinomas than in high grade carcinomas (85.1% versus 47.6%) and in rectal and sigmoid carcinomas than in proximal colon carcinomas (88.2% versus 63.2% and 88.9% versus 63.2%, respectively). Furthermore, CK7 expression was more common in tumors with lymph node metastasis than in non-metastatic tumors (25.3% versus 11%). In conclusion, a considerable number of colorectal carcinomas showed reactivity to CK7 (17.3%) or no reactivity to CK20 (18.9%). Therefore, CK7 positivity or CK20 negativity does not rule out a colorectal origin of metastatic carcinoma. © 2011 Published by Elsevier GmbH.
Introduction A metastatic lesion can be the first clinical presentation of a neoplastic process, probably accounting for more than 10% of all new tumor diagnoses [1]. Although modern imaging technology has resulted in improvements in the identification of primary tumors, in most patients, the origin remains unknown. Therefore, the pathologist has acquired an increasingly important diagnostic role in characterizing the site of origin of these tumors [2]. Immunohistochemistry has been shown to be a useful tool in the identification of the primary site. It is the least invasive diagnostic technique and is certainly less expensive than imaging techniques. The generation and availability of new antibodies that are specific for a wide range of antigens and tissues that work on routinely fixed, paraffin-embedded material will ensure that this technique will continue to offer useful information in the search for elusive primary sites. Cytokeratins are among a group of approximately 20 cytoskeletal structural proteins present in normal epithelia, and their expression is maintained during malignant transformation [3].
∗ Corresponding author at: Fatih Üniversitesi Tıp Fakültesi, Alparslan Türkes¸ Cad. No. 57, 06510 Emek, Ankara, Turkey. Tel.: +90 3122035584; fax: +90 3122035460. E-mail address: [email protected] (R. Bayrak). 0344-0338/$ – see front matter © 2011 Published by Elsevier GmbH. doi:10.1016/j.prp.2010.12.005
Because CK expression is usually preserved in neoplastic cells, the use of specific antibodies is of value in determining the origin of metastatic carcinomas. In normal tissue, CK7 is expressed in simple epithelia of breast, lung, mesothelium, female genital tract, and urinary bladder, with limited expression in gastric and intestinal mucosa [4,5]. CK20 is a major cellular protein of mature enterocytes and goblet cells in the intestinal tract, urothelium, and Merkel cells in the skin [6]. The relative expression of CK20 and CK7 has been observed to vary among different epithelial tumors, and these markers are currently used as diagnostic tools to help determine the site of origin of metastatic carcinomas [7–11]. For example, the CK20+/CK7− immunoprofile has been considered characteristic for colonic adenocarcinoma, and has been used to distinguish it from adenocarcinomas of other origins, such as female genital tract, lung, breast, and liver. Because cytoplasmic CK7 expression in colorectal adenocarcinoma is considered rare, CK7 positivity in metastatic carcinoma is often considered to exclude a colorectal primary [9,12,13]. However, not all colorectal carcinomas show the CK20+/CK7− expression pattern; a substantial proportion of colorectal carcinomas are CK20− or CK7+ [14–16]. In the present study, we investigated the expression profile of CK7 and CK20 in 196 primary colorectal adenocarcinomas in the light of the potential applicability of these markers in the clinical context of metastatic adenocarcinomas.
R. Bayrak et al. / Pathology – Research and Practice 207 (2011) 156–160
157
Materials and methods Case selection and tissue samples Paraffin-embedded tissue sections were obtained from archival material from 196 patients who underwent resection for primary colorectal carcinoma between January 2005 and December 2009. All cases were selected from patient files at the Department of Pathology, Fatih University Hospital. Pathological findings, including histological type, histological differentiation, depth of invasion, and lymph node status, were obtained from hematoxylin and eosinstained sections. All cases were reviewed to confirm the diagnosis. WHO criteria were used for histological typing [17]. Postoperative pathological staging was performed according to the American Joint Committee on Cancer (AJCC) TNM staging system [18]. One paraffin block with the maximum amount of tumor and proper fixation was selected from each case for immunohistochemical (IHC) studies. Among the 196 cases of colorectal carcinoma, 174 were classified as adenocarcinoma, 14 as mucinous carcinoma, and 8 as signet-ring cell carcinoma. Eight cases were categorized as T1, 16 as T2, 148 as T3, and 24 as T4 according to the American Joint Committee on Cancer staging system. Based on histological grading, the carcinomas were classified as 175 low-grade carcinomas (well-moderately differentiated) and 21 high-grade carcinomas (poorly differentiated and undifferentiated carcinomas). According to their location, tumors were classified as rectal carcinomas (n = 85), sigmoid carcinomas (n = 54), and carcinomas proximal to the rectosigmoid region (n = 57). Lymph node metastasis was observed in 87 patients. Immunohistochemistry Four-m-thick sections were cut from blocks of paraffinembedded tissue, deparaffinized, and rehydrated as usual. To reduce non-specific background staining due to endogenous peroxidase, slides were incubated in Hydrogen Peroxide Block for 15 min. Before immunostaining, antigen retrieval was performed by incubating the slides for 15 min with pepsin (LabVision; catalog no. AP-9007) at a concentration of 1 mg/ml for CK20. Slides were microwaved in 10 mM of citric acid at pH 6.0 for 20 min for CK7. The slides were incubated for 60 min with primary antibodies to CK7 (clone OV-TL 12/30, LabVision/NeoMarkers; 1:50) and CK20 (clone Ks 20.8, Dako; 1:50) at room temperature. The standard avidin–biotin–peroxidase complex (ABC) technique was performed using the LabVision Secondary Detection Kit (UltraVision Detection System Anti-polyvalent, HRP). AEC and DAB were used as chromogen. All slides were counterstained with Mayer’s hematoxylin. Positive immunostaining for CK7 and CK20 was identified in the cytoplasm, cell membrane, or both tumor cell components. The immunostaining result was assessed semiquantitatively. The percentage of positive cells was recorded as follows: less than 5%, 5–25%, 26–50%, 51–75%, and more than 75%. Cases that stained less than 5% were considered negative, and all others as positive for statistical purpose. Normal colonic mucosal tissue was used as a CK20-positive control, and breast carcinoma tissue was used as a CK7-positive control. We used the normal colorectal tissue of resected specimens to evaluate the normal expression patterns of CK7 and CK20. Statistical analysis The results were evaluated using Pearson’s chi-square test. p values less than 0.05 were considered significant. SPSS 13.0 for Windows was used for all statistical analyses.
Fig. 1. Normal colonic epithelium with cytokeratin 20 (A) and cytokeratin 7 (B) immunoreactivity (Note: The lack of CK7 immunoreactivity in neoplastic gland) (100×).
Results Histologically normal colonic epithelium stained positively for CK20. Cytoplasmic CK20 immunoreactivity was more prominent in the surface epithelium and tended to diminish toward the crypts (Fig. 1A). We also observed CK7 immunoreactivity in normal colonic epithelium, and this immunoreactivity tended to be focal in the surface and the crypt epithelium (Fig. 1B). CK20 expression was detected in 159 of 196 (81.1%), and CK7 expression was detected in 34 of 196 (17.3%) colorectal carcinomas. CK20 reactivity was diffuse in the majority of colon adenocarcinomas (more than 50% of cells were positive) in 54% of the cases (Fig. 2). Among the 34 cases of colorectal carcinoma expressing CK7, 22 were focally positive, and 12 were diffusely positive (Fig. 3). In diffusely CK7-positive cases, 4 were CK20-negative (Fig. 4). The results are summarized in Table 1. In terms of the combined expression of CK7 and CK20, the proportion of CK7−/CK20+ was highest, accounting for 65.8% (129 of 196) of colorectal carcinomas. The CK7+/CK20+ immunophenotype was expressed in 30 (15.3%), and CK7−/CK20− in 33 (16.9%) cases. The CK7+/CK20− expression pattern was observed in only 2% (4 of 196) of colorectal carcinomas (Table 2). CK7 and CK20 expression in colorectal carcinoma was compared with the clinicopathological characteristics (Table 3). No association between CK7 expression and anatomical location of carcinomas, tumor type, tumor stage (pT), and grade was found. There was a correlation between CK7 expression and lymph node metastasis, as CK7 expression was more common in tumors with lymph node metastasis than in non-metastatic tumors (25.3% versus
158
R. Bayrak et al. / Pathology – Research and Practice 207 (2011) 156–160
Fig. 2. Diffuse CK20 immunopositivity in low grade (A) and high grade (B) colorectal adenocarcinomas (100×).
Fig. 3. Diffuse (A) and focal (B) CK7 immunopositivity in colorectal adenocarcinomas (100×).
11%, 2 = 6879; p = 0.009). CK7+/CK20+ expression pattern was also more common in metastatic tumors than in non-metastatic tumors (21.8% versus 10.1%, 2 = 5151; p = 0.023). No association was observed among CK20 expression and tumor type, tumor stage (pT), and nodal status. The association between CK20 expression and anatomical location of tumors was statistically highly significant. CK20 positivity was more common in rectal and sigmoid carcinomas than in proximal colon carcinomas (88.2% versus 63.2%, 2 = 12.574; p < 0.001 and 88.9% versus 63.2%, 2 = 9.974; p = 0.002, respectively). CK20 expression was similar in rectal and sigmoid carcinomas (88.2% and 88.9%, respectively. 2 = 0.014; p = 0.906). Furthermore, the proportion of CK7−/CK20+ was higher in rectal and sigmoid carcinomas than in proximal colon carcinomas (69.4% versus 54.4% and 72.2% versus 54.4%, respectively). This result was not statistically significant (2 = 3.319, p = 0.068 and 2 = 3.787, p = 0.052, respectively). CK20 positivity was more common in low grade carcinomas than in high grade carcinomas (85.1% versus 47.6%, 2 = 14,876; p < 0.001). In terms of combined expression patterns, the proportion of CK7−/CK20+ was higher in low grade carcinomas than in high grade carcinomas (68.6% versus 42.9%, 2 = 5511; p = 0.019). The CK7−/CK20− expression pattern was more common in high grade tumors than in low grade tumors as expected (42.9% versus 13.7%, 2 = 11,373; p < 0.001).
Discussion Cytokeratins (CKs) are cytoplasmic intermediate filaments found in the epithelial cells, and different cytokeratin proteins are distributed in a tissue-specific fashion [3]. Carcinomas usually have a characteristic cytokeratin profile, similar to that of its native epithelium that can still be maintained in its metastasis [8,11,19]. This consistency of cytokeratin expression profile in specific epithelial neoplasms has been used to identify the origin of a metastatic tumor [9,11–13,19–22]. The identification of the primary site of metastatic tumor could lead to more effective treatment, improving the overall outcome. Cytokeratin 7 is expressed in many ductal and glandular epithelia, including respiratory epithelium, mammary ducts and asini, müllerian epithelium, apocrine and ecrine glands, biliary tract epithelium, urothelium, and mesothelium [3,4]. The presence of CK7 reactivity in an adenocarcinoma generally favors an origin from breast, urinary bladder, female genital tract, lung, or pancreatobiliary tract [5,7,8,11,19]. In contrast, CK20 has been found to be restricted in colonic epithelium, urothelium, and Merkel cells, and the neoplasms derived from these epithelia and cells [3,6–8]. Their differential expression, maintained in malignant transformation, has been the subject of several investigations that have delineated
Table 1 Percentages of CK20- and CK7-positive cells in colorectal carcinomas.
CK20 CK7
<5%
5–25%
26–50%
51–75%
76–100%
37/196 (18.9%) 162/196 (82.7%)
20/196 (10.2%) 14/196 (7.1%)
33/196 (16.9%) 8/196 (4.1%)
56/196 (28.5%) 4/196 (2%)
50/196 (25.5%) 8/196 (4.1%)
R. Bayrak et al. / Pathology – Research and Practice 207 (2011) 156–160
159
Table 2 CK20 and CK7 expression patterns in our studied groups.
Fig. 4. A case of colorectal adenocarcinoma showing CK7+/CK20− staining pattern. (A) CK7 (100×); (B) CK20 (100×).
a specific pattern of CK7/CK20 reactivity in primary and metastatic carcinomas [7–10]. The CK7−/CK20+ pattern is considered specific for colorectal epithelial tumors, and was identified in up to 95% of the cases in var-
CK7+
CK7−
Total
CK20+ CK20−
30 (15.3%) 4 (2%)
129 (65.8%) 33 (16.9%)
159 (81.1%) 37 (18.9%)
Total
34 (17.3%)
162 (82.7%)
196 (100%)
ious studies [7,8,15,16,23,24]. In comparison with the CK7−/CK20+ immunoprofile, the CK7+/CK20+ immunoprofile is less specific in the diagnosis of colorectal glandular neoplasia, and is commonly present in urothelial carcinomas and less common in adenocarcinomas of the stomach and the pancreatobiliary tract [7,8,10]. CK7 positivity was variable, ranging from 0% to 22%, and the immunostaining score or the percentage of CK7-positive cells was generally low in previous studies [5,7,8,15,16,23,24]. Chu et al. [8] reported CK20 positivity in 95% and CK7 in 5% of their colorectal adenocarcinomas, with more than 5% positive cells as a cutoff. Ramalingam et al. [15] reported 13% of their cases as CK7+/CK20+. Saad et al. [16] reported CK20 positivity in 89% and CK7 positivity in 22% of their colorectal adenocarcinoma cases. In their 12 cases of colorectal carcinoma expressing CK7 (12 of 55), 8 were diffusely positive (more than 50% of cells were positive). Park et al. [24] reported 5% of their cases as CK7+/CK20+ and 4% as CK7+/CK20−. In their 21 cases of colorectal carcinoma expressing CK7 (21 of 225), 6 were focally positive, and 15 were diffusely positive. We also found that the CK7−/CK20+ expression pattern was most common in colorectal carcinomas (65.8%) in the present study. CK20 reactivity was diffuse in the majority of colon carcinomas (more than 50% of cells were positive) in 54% of the cases. However, a considerable number of colorectal carcinomas showed reactivity to CK7 (17.3%) or no reactivity to CK20 (18.9%). In our 34 cases of colorectal carcinoma expressing CK7, 22 were focally positive, and 12 were diffusely positive. Among the diffusely CK7positive cases, 4 were CK20-negative. Therefore, CK7 positivity or CK20 negativity does not rule out a colorectal origin of metastatic carcinoma. In contrast, a recent study has reported a high percentage for CK7 positivity in rectal carcinoma, up to 74% [14]. The reasons for this discrepancy are unclear. However, this may be
Table 3 Expression of CK20 and CK7 in colorectal cancer tissue by histopathologic and clinical characteristics.
Grade Low High Anatomic site Rectum Sigmoid Proximal Histology Adeno Mucinous Signet ring Tumor stage T1 T2 T3 T4 Nodal stage N0 N1–2 * # ␥ § ¶ §§ †
CK20+
CK7+
CK20+/CK7−
CK20+/CK7+
CK20−/CK7+
CK20−/CK7−
149/175 (85.1%)* 10/21 (47.6%)*
31/175 (17.7%) 3/21 (14.3%)
120/175 (68.6%)¶ 9/21 (42.9%)¶
29/175 (16.6%) 1/21 (4.8%)
2/175 (1.1%) 2/21 (9.5%)
24/175 (13.7%)† 9/21 (42.9%)†
75/85 (88.2%)# 48/54 (88.9%) 36/57 (63.2%)# ,
16/85 (18.8%) 10/54 (18.5%) 8/57 (14%)
59/85 (69.4%) 39/54 (72.2%) 31/57 (54.4%)
16/85 (18.8%) 9/54 (16.7%) 5/57 (8.8%)
0/85 (0%) 1/54 (1.9%) 3/57 (5.3%)
10/85 (11.8%) 5/54 (9.3%) 18/57 (31.6%)
143/174 (82.2%) 10/14 (71.4%) 6/8 (75%)
31/174 (17.8%) 3/14 (21.4%) 0/8 (0%)
116/174 (66.7%) 7/14 (50%) 6/8 (75%)
27/174 (15.5%) 3/14 (21.4%) 0/8 (0%)
4/174 (2.3%) 0/14 (0%) 0/8 (0%)
27/174 (15.5%) 4/14 (28.6%) 2/8 (25%)
5/8 (62.5%) 14/16 (87.5%) 121/148 (81.8%) 19/24 (79.2%)
1/8 (12.5%) 4/16 (25%) 25/148 (16.9%) 4/24 (16.7%)
5/8 (62.5%) 10/16 (62.5%) 98/148 (66.2%) 16/24 (66.7%)
0/8 (0%) 4/16 (25%) 23/148 (15.5%) 3/24 (12.5%)
1/8 (12.5%) 0/16 (0%) 2/148 (1.4%) 1/24 (4.1%)
2/8 (25%) 2/16 ((12.5%) 25/148 (16.9%) 4/24 (16.7%)
85/109 (78%) 74/87 (85.1%)
12/109 (11%)§ 22/87 (25.3%)§
74/109 (67.9%) 55/87 (63.2%)
11/109 (10.1%)§§ 19/87 (21.8%)§§
1/109 (0.9%) 3/87 (3.4%)
23/109 (21.1%) 10/87 (11.5%)
2 = 14.876; p < 0.001. CK20 in rectal versus proximal Ca, 2 = 12.574; p < 0.001. CK20 in sigmoid versus proximal Ca, 2 = 9.974; p = 0.002. 2 = 6.879; p = 0.009. 2 = 5.511; p = 0.019. 2 = 5.151; p = 0.023. 2 = 11.373; p < 0.001.
160
R. Bayrak et al. / Pathology – Research and Practice 207 (2011) 156–160
the result of differences in the studied population and technical variations in the immunohistochemical procedure or difficulty in separating CK7+/CK20+ anal gland adenocarcinomas arising in the anal region from conventional rectal adenocarcinomas on a purely morphologic basis. Factors that are associated with CK7 and CK20 positivity in colorectal carcinomas were studied by a small number of groups [23,24]. Park et al. [24] reported that CK7 and CK20 positivities were associated with histologic grade and tumor location. They found that the CK7−/CK20+ pattern was more frequent in low-grade and left-sided colorectal carcinomas, and the CK7+/CK20− pattern was more frequent in high-grade and right-sided colonic carcinomas. Hernandez et al. [23] reported more CK7 expression in advancedstage colorectal adenocarcinomas than in early stage. Although the study did not address the relationship between CK7 positivity and degree of histological differentiation, one can presume that higher stage tumors were also likely to be less differentiated. The present study showed that CK20 positivity is associated with histologic grade and tumor location, CK7 positivity is associated with lymph node metastasis. CK7 expression was more common in tumors with lymph node metastasis. No association between CK7 expression and anatomical location of carcinomas, tumor type, tumor stage (pT), or grade was found. There was an association between CK20 expression and anatomical location of tumors, which was statistically highly significant. CK20 positivity was more common in rectal and sigmoid carcinomas than in proximal colon carcinomas. Furthermore, CK20 positivity was more common in low grade carcinomas than in high grade carcinomas. No association was established among CK20 expression and tumor type, tumor stage (pT), or nodal status. Our results underscore that CK7 positivity is present in a significant number of colorectal carcinomas, and that its presence does not exclude a primary tumor from this site. Positive CK7 reactivity should be interpreted with caution, even in the light of negative CK20 reactivity, and should not be used as the sole basis for excluding a colorectal primary. Acknowledgements The results of this study were partially presented in the poster session of the 21st European Congress of Pathology, 8–13 September 2007, I˙ stanbul, Turkey. References [1] L. Pecciarini, M.G. Cangi, C. Doglioni, Identifying the primary sites of metastatic carcinoma: the increasing role of immunohistochemistry, Curr. Diag. Pathol. 7 (2001) 168–175. [2] G.R. Varadhachary, J.L. Abbruzzese, R. Lenzi, Diagnostic strategies for unknown primary cancer, Cancer 100 (2004) 1776–1785.
[3] R. Moll, W. Werner, F.D. Schiller, The catalog of human cytokeratins: patterns of expression in normal epithelia, tumors and cultured cells, Cell 31 (1982) 11–24. [4] C.C. Van Niekerk, P.H. Jap, F.C. Ramaekers, et al., Immunohistochemical demonstration of keratin 7 in routinely fixed parafin embedded human tissues, J. Pathol. 165 (1991) 145–152. [5] F.C. Ramaekers, C.C. Van Niekerk, L. Poels, et al., Use of monoclonal antibodies to keratin 7 in differential diagnosis of adenocarcinomas, Am. J. Pathol. 136 (1990) 641–655. [6] R. Moll, A. Lowe, J. Laufer, et al., Cytokeratin 20 in human carcinomas. A new histodiagnostic marker detected by monoclonal antibodies, Am. J. Pathol. 140 (1992) 427–447. [7] N.P. Wang, S. Zee, R.J. Zarbo, et al., Coordinate expression of cytokeratins 7 and 20 defines unique subsets of carcinomas, Appl. Immunohistochem. Mol. Morphol. 3 (1995) 99–107. [8] P. Chu, E. Wu, L.M. Weiss, Cytokeratin 7 and cytokeratin 20 expression in epithelial neoplasms: a survey of 435 cases, Mod. Pathol. 13 (2000) 962–972. [9] C.C. Wauters, F. Smedts, L.G. Gerrits, et al., Keratins 7 and 20 as diagnostic markers of carcinomas metastatic to the ovary, Hum. Pathol. 26 (1995) 850–855. [10] A. Rullier, B. Le Bail, R. Fawaz, et al., Cytokeratin 7 and 20 expression in cholangiocarcinomas varies along the biliary tract but still differs from that in colorectal carcinomas metastasis, Am. J. Surg. Pathol. 24 (2000) 870–876. [11] S.Y. Park, B.H. Kim, J.H. Kim, et al., Panels of immunohistochemical markers help determine primary sites of metastatic adenocarcinoma, Arch. Pathol. Lab. Med. 131 (2007) 1561–1567. [12] T. Maeda, K. Kajiyama, E. Adachi, et al., The expression of cytokeratins 7, 19, 20 in primary and metastatic carcinomas of the liver, Mod. Pathol. 9 (1996) 901–909. [13] T.S. Loy, R.D. Calaluce, Utility of cytokeratin immunostaining in separating pulmonary adenocarcinomas from colonic adenocarcinomas, Am. J. Clin. Pathol. 102 (1994) 764–767. [14] P.J. Zhang, M. Shah, G.W. Spiegel, et al., Cytokeratin 7 immunoreactivity in rectal adenocarcinomas, Appl. Immunohistochem. Mol. Morphol. 11 (2003) 306–310. [15] P. Ramalingam, W.R. Hart, J.R. Goldblum, Cytokeratin subset immunostaining in rectal adenocarcinoma and normal anal glands, Arch. Pathol. Lab. Med. 125 (2001) 1074–1077. [16] R.S. Saad, J.F. Silverman, M.A. Khalifa, et al., CDX2, cytokeratins 7 and 20 immunoreactivity in rectal adenocarcinoma, Appl. Immunohistochem. Mol. Morphol. 17 (2009) 196–201. [17] S.R. Hamilton, B. Vogelstein, S. Kudo, et al., Carcinoma of the colon and rectum, in: S.R. Hamilton, L.A. Aaltonen (Eds.), World Health Organization Classification of Tumours: Pathology and Genetics of Tumours of the Digestive System, IARC Press, Lyon, France, 2000, pp. 103–143. [18] S.B. Edge, D.R. Byrd, M.A. Carducci, C.C. Compton (Eds.), AJCC Cancer Staging Manual, 7th ed., Springer, New York, NY, 2009. [19] J.H. Lagendijk, H. Mullink, P.J. Van Diest, et al., Tracing the origin of adenocarcinomas with unknown primary using immunohistochemistry: differential diagnosis between colonic and ovarian carcinomas as primary sites, Hum. Pathol. 29 (1998) 491–497. [20] T.S. Loy, R.D. Calauce, G.L. Keeney, Cytokeratin immunostaining in differentiating primary ovarian carcinoma from metastatic colonic adenocarcinoma, Mod. Pathol. 9 (1996) 1040–1044. [21] A. Perry, T.T. Parisi, P.J. Kurtin, Metastatic adenocarcinoma to the brain: an immunohistcochemical approach, Hum. Pathol. 28 (1997) 938–943. [22] S. Ikeda, M. Fujimori, S. Shibala, et al., Combined immunohistochemistry of beta-catenin, cytokeratin 7 and cytokeratin 20 is useful in discriminating primary lung from metastatic colorectal carcinoma, BMC Cancer 6 (2006) 31–36. [23] B.Y. Hernandez, M.F. Frierson, C.A. Moskaluk, et al., CK20 and CK7 protein expression in colorectal carcinoma: demonstration of the utility of a population-based tissue microarray, Hum. Pathol. 36 (2005) 275–281. [24] S.Y. Park, H.S. Kim, E.K. Hong, et al., Expression of cytokeratins 7 and 20 in primary carcinomas of the stomach and colorectum and their value in the differential diagnosis of metastatic carcinomas to the ovary, Hum. Pathol. 29 (1998) 723–728.
Contents lists available at ScienceDirect
Pathology – Research and Practice journal homepage: www.elsevier.de/prp
Original Article
Cytokeratin 7 and cytokeratin 20 expression in colorectal adenocarcinomas Reyhan Bayrak ∗ , Sibel Yenidünya, Hacer Haltas Department of Pathology, School of Medicine, Fatih University, 06540 Emek, Ankara, Turkey
a r t i c l e
i n f o
Article history: Received 21 July 2010 Received in revised form 28 September 2010 Accepted 15 December 2010 Keywords: Colon carcinoma Cytokeratin 7 Cytokeratin 20 Immunohistochemistry
a b s t r a c t Cytokeratin 7 (CK7) and cytokeratin 20 (CK20) are low molecular weight cytokeratins. The expressions of CK7 and CK20 have been studied in various primary and metastatic carcinomas. Their expression patterns may help to distinguish the site of origin of metastatic carcinomas. We investigated the expressions of CK7 and CK20 in 196 cases of colorectal carcinoma. Paraffin sections of 196 colonic adenocarcinomas were randomly selected, retrieved, and immunostained for CK7 and CK20 with a standard avidin–biotin complex method. CK7 was expressed in 34/196 (17.3%) and CK20 in 159/196 (81.1) cases of colorectal adenocarcinoma. CK7−/CK20+ had the greatest proportion (65.8%) in colorectal carcinomas. The CK7+/CK20+ immunophenotype was identified in 30/196 (15.3%), CK7−/CK20− in 33/196 (16.9%), and CK7+/CK20− in 4/196 (2%) colon adenocarcinomas. The CK7 and CK20 expression patterns were different in colorectal carcinomas according to histological grade, location of the tumor, and lymph node metastasis. CK20 positivity was more common in low grade carcinomas than in high grade carcinomas (85.1% versus 47.6%) and in rectal and sigmoid carcinomas than in proximal colon carcinomas (88.2% versus 63.2% and 88.9% versus 63.2%, respectively). Furthermore, CK7 expression was more common in tumors with lymph node metastasis than in non-metastatic tumors (25.3% versus 11%). In conclusion, a considerable number of colorectal carcinomas showed reactivity to CK7 (17.3%) or no reactivity to CK20 (18.9%). Therefore, CK7 positivity or CK20 negativity does not rule out a colorectal origin of metastatic carcinoma. © 2011 Published by Elsevier GmbH.
Introduction A metastatic lesion can be the first clinical presentation of a neoplastic process, probably accounting for more than 10% of all new tumor diagnoses [1]. Although modern imaging technology has resulted in improvements in the identification of primary tumors, in most patients, the origin remains unknown. Therefore, the pathologist has acquired an increasingly important diagnostic role in characterizing the site of origin of these tumors [2]. Immunohistochemistry has been shown to be a useful tool in the identification of the primary site. It is the least invasive diagnostic technique and is certainly less expensive than imaging techniques. The generation and availability of new antibodies that are specific for a wide range of antigens and tissues that work on routinely fixed, paraffin-embedded material will ensure that this technique will continue to offer useful information in the search for elusive primary sites. Cytokeratins are among a group of approximately 20 cytoskeletal structural proteins present in normal epithelia, and their expression is maintained during malignant transformation [3].
∗ Corresponding author at: Fatih Üniversitesi Tıp Fakültesi, Alparslan Türkes¸ Cad. No. 57, 06510 Emek, Ankara, Turkey. Tel.: +90 3122035584; fax: +90 3122035460. E-mail address: [email protected] (R. Bayrak). 0344-0338/$ – see front matter © 2011 Published by Elsevier GmbH. doi:10.1016/j.prp.2010.12.005
Because CK expression is usually preserved in neoplastic cells, the use of specific antibodies is of value in determining the origin of metastatic carcinomas. In normal tissue, CK7 is expressed in simple epithelia of breast, lung, mesothelium, female genital tract, and urinary bladder, with limited expression in gastric and intestinal mucosa [4,5]. CK20 is a major cellular protein of mature enterocytes and goblet cells in the intestinal tract, urothelium, and Merkel cells in the skin [6]. The relative expression of CK20 and CK7 has been observed to vary among different epithelial tumors, and these markers are currently used as diagnostic tools to help determine the site of origin of metastatic carcinomas [7–11]. For example, the CK20+/CK7− immunoprofile has been considered characteristic for colonic adenocarcinoma, and has been used to distinguish it from adenocarcinomas of other origins, such as female genital tract, lung, breast, and liver. Because cytoplasmic CK7 expression in colorectal adenocarcinoma is considered rare, CK7 positivity in metastatic carcinoma is often considered to exclude a colorectal primary [9,12,13]. However, not all colorectal carcinomas show the CK20+/CK7− expression pattern; a substantial proportion of colorectal carcinomas are CK20− or CK7+ [14–16]. In the present study, we investigated the expression profile of CK7 and CK20 in 196 primary colorectal adenocarcinomas in the light of the potential applicability of these markers in the clinical context of metastatic adenocarcinomas.
R. Bayrak et al. / Pathology – Research and Practice 207 (2011) 156–160
157
Materials and methods Case selection and tissue samples Paraffin-embedded tissue sections were obtained from archival material from 196 patients who underwent resection for primary colorectal carcinoma between January 2005 and December 2009. All cases were selected from patient files at the Department of Pathology, Fatih University Hospital. Pathological findings, including histological type, histological differentiation, depth of invasion, and lymph node status, were obtained from hematoxylin and eosinstained sections. All cases were reviewed to confirm the diagnosis. WHO criteria were used for histological typing [17]. Postoperative pathological staging was performed according to the American Joint Committee on Cancer (AJCC) TNM staging system [18]. One paraffin block with the maximum amount of tumor and proper fixation was selected from each case for immunohistochemical (IHC) studies. Among the 196 cases of colorectal carcinoma, 174 were classified as adenocarcinoma, 14 as mucinous carcinoma, and 8 as signet-ring cell carcinoma. Eight cases were categorized as T1, 16 as T2, 148 as T3, and 24 as T4 according to the American Joint Committee on Cancer staging system. Based on histological grading, the carcinomas were classified as 175 low-grade carcinomas (well-moderately differentiated) and 21 high-grade carcinomas (poorly differentiated and undifferentiated carcinomas). According to their location, tumors were classified as rectal carcinomas (n = 85), sigmoid carcinomas (n = 54), and carcinomas proximal to the rectosigmoid region (n = 57). Lymph node metastasis was observed in 87 patients. Immunohistochemistry Four-m-thick sections were cut from blocks of paraffinembedded tissue, deparaffinized, and rehydrated as usual. To reduce non-specific background staining due to endogenous peroxidase, slides were incubated in Hydrogen Peroxide Block for 15 min. Before immunostaining, antigen retrieval was performed by incubating the slides for 15 min with pepsin (LabVision; catalog no. AP-9007) at a concentration of 1 mg/ml for CK20. Slides were microwaved in 10 mM of citric acid at pH 6.0 for 20 min for CK7. The slides were incubated for 60 min with primary antibodies to CK7 (clone OV-TL 12/30, LabVision/NeoMarkers; 1:50) and CK20 (clone Ks 20.8, Dako; 1:50) at room temperature. The standard avidin–biotin–peroxidase complex (ABC) technique was performed using the LabVision Secondary Detection Kit (UltraVision Detection System Anti-polyvalent, HRP). AEC and DAB were used as chromogen. All slides were counterstained with Mayer’s hematoxylin. Positive immunostaining for CK7 and CK20 was identified in the cytoplasm, cell membrane, or both tumor cell components. The immunostaining result was assessed semiquantitatively. The percentage of positive cells was recorded as follows: less than 5%, 5–25%, 26–50%, 51–75%, and more than 75%. Cases that stained less than 5% were considered negative, and all others as positive for statistical purpose. Normal colonic mucosal tissue was used as a CK20-positive control, and breast carcinoma tissue was used as a CK7-positive control. We used the normal colorectal tissue of resected specimens to evaluate the normal expression patterns of CK7 and CK20. Statistical analysis The results were evaluated using Pearson’s chi-square test. p values less than 0.05 were considered significant. SPSS 13.0 for Windows was used for all statistical analyses.
Fig. 1. Normal colonic epithelium with cytokeratin 20 (A) and cytokeratin 7 (B) immunoreactivity (Note: The lack of CK7 immunoreactivity in neoplastic gland) (100×).
Results Histologically normal colonic epithelium stained positively for CK20. Cytoplasmic CK20 immunoreactivity was more prominent in the surface epithelium and tended to diminish toward the crypts (Fig. 1A). We also observed CK7 immunoreactivity in normal colonic epithelium, and this immunoreactivity tended to be focal in the surface and the crypt epithelium (Fig. 1B). CK20 expression was detected in 159 of 196 (81.1%), and CK7 expression was detected in 34 of 196 (17.3%) colorectal carcinomas. CK20 reactivity was diffuse in the majority of colon adenocarcinomas (more than 50% of cells were positive) in 54% of the cases (Fig. 2). Among the 34 cases of colorectal carcinoma expressing CK7, 22 were focally positive, and 12 were diffusely positive (Fig. 3). In diffusely CK7-positive cases, 4 were CK20-negative (Fig. 4). The results are summarized in Table 1. In terms of the combined expression of CK7 and CK20, the proportion of CK7−/CK20+ was highest, accounting for 65.8% (129 of 196) of colorectal carcinomas. The CK7+/CK20+ immunophenotype was expressed in 30 (15.3%), and CK7−/CK20− in 33 (16.9%) cases. The CK7+/CK20− expression pattern was observed in only 2% (4 of 196) of colorectal carcinomas (Table 2). CK7 and CK20 expression in colorectal carcinoma was compared with the clinicopathological characteristics (Table 3). No association between CK7 expression and anatomical location of carcinomas, tumor type, tumor stage (pT), and grade was found. There was a correlation between CK7 expression and lymph node metastasis, as CK7 expression was more common in tumors with lymph node metastasis than in non-metastatic tumors (25.3% versus
158
R. Bayrak et al. / Pathology – Research and Practice 207 (2011) 156–160
Fig. 2. Diffuse CK20 immunopositivity in low grade (A) and high grade (B) colorectal adenocarcinomas (100×).
Fig. 3. Diffuse (A) and focal (B) CK7 immunopositivity in colorectal adenocarcinomas (100×).
11%, 2 = 6879; p = 0.009). CK7+/CK20+ expression pattern was also more common in metastatic tumors than in non-metastatic tumors (21.8% versus 10.1%, 2 = 5151; p = 0.023). No association was observed among CK20 expression and tumor type, tumor stage (pT), and nodal status. The association between CK20 expression and anatomical location of tumors was statistically highly significant. CK20 positivity was more common in rectal and sigmoid carcinomas than in proximal colon carcinomas (88.2% versus 63.2%, 2 = 12.574; p < 0.001 and 88.9% versus 63.2%, 2 = 9.974; p = 0.002, respectively). CK20 expression was similar in rectal and sigmoid carcinomas (88.2% and 88.9%, respectively. 2 = 0.014; p = 0.906). Furthermore, the proportion of CK7−/CK20+ was higher in rectal and sigmoid carcinomas than in proximal colon carcinomas (69.4% versus 54.4% and 72.2% versus 54.4%, respectively). This result was not statistically significant (2 = 3.319, p = 0.068 and 2 = 3.787, p = 0.052, respectively). CK20 positivity was more common in low grade carcinomas than in high grade carcinomas (85.1% versus 47.6%, 2 = 14,876; p < 0.001). In terms of combined expression patterns, the proportion of CK7−/CK20+ was higher in low grade carcinomas than in high grade carcinomas (68.6% versus 42.9%, 2 = 5511; p = 0.019). The CK7−/CK20− expression pattern was more common in high grade tumors than in low grade tumors as expected (42.9% versus 13.7%, 2 = 11,373; p < 0.001).
Discussion Cytokeratins (CKs) are cytoplasmic intermediate filaments found in the epithelial cells, and different cytokeratin proteins are distributed in a tissue-specific fashion [3]. Carcinomas usually have a characteristic cytokeratin profile, similar to that of its native epithelium that can still be maintained in its metastasis [8,11,19]. This consistency of cytokeratin expression profile in specific epithelial neoplasms has been used to identify the origin of a metastatic tumor [9,11–13,19–22]. The identification of the primary site of metastatic tumor could lead to more effective treatment, improving the overall outcome. Cytokeratin 7 is expressed in many ductal and glandular epithelia, including respiratory epithelium, mammary ducts and asini, müllerian epithelium, apocrine and ecrine glands, biliary tract epithelium, urothelium, and mesothelium [3,4]. The presence of CK7 reactivity in an adenocarcinoma generally favors an origin from breast, urinary bladder, female genital tract, lung, or pancreatobiliary tract [5,7,8,11,19]. In contrast, CK20 has been found to be restricted in colonic epithelium, urothelium, and Merkel cells, and the neoplasms derived from these epithelia and cells [3,6–8]. Their differential expression, maintained in malignant transformation, has been the subject of several investigations that have delineated
Table 1 Percentages of CK20- and CK7-positive cells in colorectal carcinomas.
CK20 CK7
<5%
5–25%
26–50%
51–75%
76–100%
37/196 (18.9%) 162/196 (82.7%)
20/196 (10.2%) 14/196 (7.1%)
33/196 (16.9%) 8/196 (4.1%)
56/196 (28.5%) 4/196 (2%)
50/196 (25.5%) 8/196 (4.1%)
R. Bayrak et al. / Pathology – Research and Practice 207 (2011) 156–160
159
Table 2 CK20 and CK7 expression patterns in our studied groups.
Fig. 4. A case of colorectal adenocarcinoma showing CK7+/CK20− staining pattern. (A) CK7 (100×); (B) CK20 (100×).
a specific pattern of CK7/CK20 reactivity in primary and metastatic carcinomas [7–10]. The CK7−/CK20+ pattern is considered specific for colorectal epithelial tumors, and was identified in up to 95% of the cases in var-
CK7+
CK7−
Total
CK20+ CK20−
30 (15.3%) 4 (2%)
129 (65.8%) 33 (16.9%)
159 (81.1%) 37 (18.9%)
Total
34 (17.3%)
162 (82.7%)
196 (100%)
ious studies [7,8,15,16,23,24]. In comparison with the CK7−/CK20+ immunoprofile, the CK7+/CK20+ immunoprofile is less specific in the diagnosis of colorectal glandular neoplasia, and is commonly present in urothelial carcinomas and less common in adenocarcinomas of the stomach and the pancreatobiliary tract [7,8,10]. CK7 positivity was variable, ranging from 0% to 22%, and the immunostaining score or the percentage of CK7-positive cells was generally low in previous studies [5,7,8,15,16,23,24]. Chu et al. [8] reported CK20 positivity in 95% and CK7 in 5% of their colorectal adenocarcinomas, with more than 5% positive cells as a cutoff. Ramalingam et al. [15] reported 13% of their cases as CK7+/CK20+. Saad et al. [16] reported CK20 positivity in 89% and CK7 positivity in 22% of their colorectal adenocarcinoma cases. In their 12 cases of colorectal carcinoma expressing CK7 (12 of 55), 8 were diffusely positive (more than 50% of cells were positive). Park et al. [24] reported 5% of their cases as CK7+/CK20+ and 4% as CK7+/CK20−. In their 21 cases of colorectal carcinoma expressing CK7 (21 of 225), 6 were focally positive, and 15 were diffusely positive. We also found that the CK7−/CK20+ expression pattern was most common in colorectal carcinomas (65.8%) in the present study. CK20 reactivity was diffuse in the majority of colon carcinomas (more than 50% of cells were positive) in 54% of the cases. However, a considerable number of colorectal carcinomas showed reactivity to CK7 (17.3%) or no reactivity to CK20 (18.9%). In our 34 cases of colorectal carcinoma expressing CK7, 22 were focally positive, and 12 were diffusely positive. Among the diffusely CK7positive cases, 4 were CK20-negative. Therefore, CK7 positivity or CK20 negativity does not rule out a colorectal origin of metastatic carcinoma. In contrast, a recent study has reported a high percentage for CK7 positivity in rectal carcinoma, up to 74% [14]. The reasons for this discrepancy are unclear. However, this may be
Table 3 Expression of CK20 and CK7 in colorectal cancer tissue by histopathologic and clinical characteristics.
Grade Low High Anatomic site Rectum Sigmoid Proximal Histology Adeno Mucinous Signet ring Tumor stage T1 T2 T3 T4 Nodal stage N0 N1–2 * # ␥ § ¶ §§ †
CK20+
CK7+
CK20+/CK7−
CK20+/CK7+
CK20−/CK7+
CK20−/CK7−
149/175 (85.1%)* 10/21 (47.6%)*
31/175 (17.7%) 3/21 (14.3%)
120/175 (68.6%)¶ 9/21 (42.9%)¶
29/175 (16.6%) 1/21 (4.8%)
2/175 (1.1%) 2/21 (9.5%)
24/175 (13.7%)† 9/21 (42.9%)†
75/85 (88.2%)# 48/54 (88.9%) 36/57 (63.2%)# ,
16/85 (18.8%) 10/54 (18.5%) 8/57 (14%)
59/85 (69.4%) 39/54 (72.2%) 31/57 (54.4%)
16/85 (18.8%) 9/54 (16.7%) 5/57 (8.8%)
0/85 (0%) 1/54 (1.9%) 3/57 (5.3%)
10/85 (11.8%) 5/54 (9.3%) 18/57 (31.6%)
143/174 (82.2%) 10/14 (71.4%) 6/8 (75%)
31/174 (17.8%) 3/14 (21.4%) 0/8 (0%)
116/174 (66.7%) 7/14 (50%) 6/8 (75%)
27/174 (15.5%) 3/14 (21.4%) 0/8 (0%)
4/174 (2.3%) 0/14 (0%) 0/8 (0%)
27/174 (15.5%) 4/14 (28.6%) 2/8 (25%)
5/8 (62.5%) 14/16 (87.5%) 121/148 (81.8%) 19/24 (79.2%)
1/8 (12.5%) 4/16 (25%) 25/148 (16.9%) 4/24 (16.7%)
5/8 (62.5%) 10/16 (62.5%) 98/148 (66.2%) 16/24 (66.7%)
0/8 (0%) 4/16 (25%) 23/148 (15.5%) 3/24 (12.5%)
1/8 (12.5%) 0/16 (0%) 2/148 (1.4%) 1/24 (4.1%)
2/8 (25%) 2/16 ((12.5%) 25/148 (16.9%) 4/24 (16.7%)
85/109 (78%) 74/87 (85.1%)
12/109 (11%)§ 22/87 (25.3%)§
74/109 (67.9%) 55/87 (63.2%)
11/109 (10.1%)§§ 19/87 (21.8%)§§
1/109 (0.9%) 3/87 (3.4%)
23/109 (21.1%) 10/87 (11.5%)
2 = 14.876; p < 0.001. CK20 in rectal versus proximal Ca, 2 = 12.574; p < 0.001. CK20 in sigmoid versus proximal Ca, 2 = 9.974; p = 0.002. 2 = 6.879; p = 0.009. 2 = 5.511; p = 0.019. 2 = 5.151; p = 0.023. 2 = 11.373; p < 0.001.
160
R. Bayrak et al. / Pathology – Research and Practice 207 (2011) 156–160
the result of differences in the studied population and technical variations in the immunohistochemical procedure or difficulty in separating CK7+/CK20+ anal gland adenocarcinomas arising in the anal region from conventional rectal adenocarcinomas on a purely morphologic basis. Factors that are associated with CK7 and CK20 positivity in colorectal carcinomas were studied by a small number of groups [23,24]. Park et al. [24] reported that CK7 and CK20 positivities were associated with histologic grade and tumor location. They found that the CK7−/CK20+ pattern was more frequent in low-grade and left-sided colorectal carcinomas, and the CK7+/CK20− pattern was more frequent in high-grade and right-sided colonic carcinomas. Hernandez et al. [23] reported more CK7 expression in advancedstage colorectal adenocarcinomas than in early stage. Although the study did not address the relationship between CK7 positivity and degree of histological differentiation, one can presume that higher stage tumors were also likely to be less differentiated. The present study showed that CK20 positivity is associated with histologic grade and tumor location, CK7 positivity is associated with lymph node metastasis. CK7 expression was more common in tumors with lymph node metastasis. No association between CK7 expression and anatomical location of carcinomas, tumor type, tumor stage (pT), or grade was found. There was an association between CK20 expression and anatomical location of tumors, which was statistically highly significant. CK20 positivity was more common in rectal and sigmoid carcinomas than in proximal colon carcinomas. Furthermore, CK20 positivity was more common in low grade carcinomas than in high grade carcinomas. No association was established among CK20 expression and tumor type, tumor stage (pT), or nodal status. Our results underscore that CK7 positivity is present in a significant number of colorectal carcinomas, and that its presence does not exclude a primary tumor from this site. Positive CK7 reactivity should be interpreted with caution, even in the light of negative CK20 reactivity, and should not be used as the sole basis for excluding a colorectal primary. Acknowledgements The results of this study were partially presented in the poster session of the 21st European Congress of Pathology, 8–13 September 2007, I˙ stanbul, Turkey. References [1] L. Pecciarini, M.G. Cangi, C. Doglioni, Identifying the primary sites of metastatic carcinoma: the increasing role of immunohistochemistry, Curr. Diag. Pathol. 7 (2001) 168–175. [2] G.R. Varadhachary, J.L. Abbruzzese, R. Lenzi, Diagnostic strategies for unknown primary cancer, Cancer 100 (2004) 1776–1785.
[3] R. Moll, W. Werner, F.D. Schiller, The catalog of human cytokeratins: patterns of expression in normal epithelia, tumors and cultured cells, Cell 31 (1982) 11–24. [4] C.C. Van Niekerk, P.H. Jap, F.C. Ramaekers, et al., Immunohistochemical demonstration of keratin 7 in routinely fixed parafin embedded human tissues, J. Pathol. 165 (1991) 145–152. [5] F.C. Ramaekers, C.C. Van Niekerk, L. Poels, et al., Use of monoclonal antibodies to keratin 7 in differential diagnosis of adenocarcinomas, Am. J. Pathol. 136 (1990) 641–655. [6] R. Moll, A. Lowe, J. Laufer, et al., Cytokeratin 20 in human carcinomas. A new histodiagnostic marker detected by monoclonal antibodies, Am. J. Pathol. 140 (1992) 427–447. [7] N.P. Wang, S. Zee, R.J. Zarbo, et al., Coordinate expression of cytokeratins 7 and 20 defines unique subsets of carcinomas, Appl. Immunohistochem. Mol. Morphol. 3 (1995) 99–107. [8] P. Chu, E. Wu, L.M. Weiss, Cytokeratin 7 and cytokeratin 20 expression in epithelial neoplasms: a survey of 435 cases, Mod. Pathol. 13 (2000) 962–972. [9] C.C. Wauters, F. Smedts, L.G. Gerrits, et al., Keratins 7 and 20 as diagnostic markers of carcinomas metastatic to the ovary, Hum. Pathol. 26 (1995) 850–855. [10] A. Rullier, B. Le Bail, R. Fawaz, et al., Cytokeratin 7 and 20 expression in cholangiocarcinomas varies along the biliary tract but still differs from that in colorectal carcinomas metastasis, Am. J. Surg. Pathol. 24 (2000) 870–876. [11] S.Y. Park, B.H. Kim, J.H. Kim, et al., Panels of immunohistochemical markers help determine primary sites of metastatic adenocarcinoma, Arch. Pathol. Lab. Med. 131 (2007) 1561–1567. [12] T. Maeda, K. Kajiyama, E. Adachi, et al., The expression of cytokeratins 7, 19, 20 in primary and metastatic carcinomas of the liver, Mod. Pathol. 9 (1996) 901–909. [13] T.S. Loy, R.D. Calaluce, Utility of cytokeratin immunostaining in separating pulmonary adenocarcinomas from colonic adenocarcinomas, Am. J. Clin. Pathol. 102 (1994) 764–767. [14] P.J. Zhang, M. Shah, G.W. Spiegel, et al., Cytokeratin 7 immunoreactivity in rectal adenocarcinomas, Appl. Immunohistochem. Mol. Morphol. 11 (2003) 306–310. [15] P. Ramalingam, W.R. Hart, J.R. Goldblum, Cytokeratin subset immunostaining in rectal adenocarcinoma and normal anal glands, Arch. Pathol. Lab. Med. 125 (2001) 1074–1077. [16] R.S. Saad, J.F. Silverman, M.A. Khalifa, et al., CDX2, cytokeratins 7 and 20 immunoreactivity in rectal adenocarcinoma, Appl. Immunohistochem. Mol. Morphol. 17 (2009) 196–201. [17] S.R. Hamilton, B. Vogelstein, S. Kudo, et al., Carcinoma of the colon and rectum, in: S.R. Hamilton, L.A. Aaltonen (Eds.), World Health Organization Classification of Tumours: Pathology and Genetics of Tumours of the Digestive System, IARC Press, Lyon, France, 2000, pp. 103–143. [18] S.B. Edge, D.R. Byrd, M.A. Carducci, C.C. Compton (Eds.), AJCC Cancer Staging Manual, 7th ed., Springer, New York, NY, 2009. [19] J.H. Lagendijk, H. Mullink, P.J. Van Diest, et al., Tracing the origin of adenocarcinomas with unknown primary using immunohistochemistry: differential diagnosis between colonic and ovarian carcinomas as primary sites, Hum. Pathol. 29 (1998) 491–497. [20] T.S. Loy, R.D. Calauce, G.L. Keeney, Cytokeratin immunostaining in differentiating primary ovarian carcinoma from metastatic colonic adenocarcinoma, Mod. Pathol. 9 (1996) 1040–1044. [21] A. Perry, T.T. Parisi, P.J. Kurtin, Metastatic adenocarcinoma to the brain: an immunohistcochemical approach, Hum. Pathol. 28 (1997) 938–943. [22] S. Ikeda, M. Fujimori, S. Shibala, et al., Combined immunohistochemistry of beta-catenin, cytokeratin 7 and cytokeratin 20 is useful in discriminating primary lung from metastatic colorectal carcinoma, BMC Cancer 6 (2006) 31–36. [23] B.Y. Hernandez, M.F. Frierson, C.A. Moskaluk, et al., CK20 and CK7 protein expression in colorectal carcinoma: demonstration of the utility of a population-based tissue microarray, Hum. Pathol. 36 (2005) 275–281. [24] S.Y. Park, H.S. Kim, E.K. Hong, et al., Expression of cytokeratins 7 and 20 in primary carcinomas of the stomach and colorectum and their value in the differential diagnosis of metastatic carcinomas to the ovary, Hum. Pathol. 29 (1998) 723–728.