Prognostic value of E-cadherin, [beta]-catenin, CD44v6, and HER2/neu in metastatic cutaneous adenocarcinoma.
Abstract: * Context.--Our recent experience with a patient developing cutaneous metastases within 3 months of diagnosis of esophageal adenocarcinoma suggests that altered expression of the cellular adhesion molecules, E-cadherin and CD44v6, may have had a role to play in the rapid onset of metastases.

Objective.--To corroborate these findings, we designed a cross-sectional study to investigate the expression of select molecules involved in the metastatic cascade.

Design.--E-cadherin, [beta]-catenin, CD44v6, and HER2/neu immunohistochemical stains were performed on archival materials of metastatic adenocarcinoma to the skin from 27 patients and the available corresponding primary tumors in 10 patients. The primary sites included breast (n = 10; 37%), gastrointestinal tract (n = 10; 37%), ovary (n = 1; 4%), thyroid (n = 2; 7%), lung (n = 1; 4%), and unknown primary (n = 3; 11%).

Results.--Expression of all markers was noted with the most significant increases observed in [beta]-catenin (26 of 27 cases; 96%), followed by CD44v6 (24 of 27 cases; 89%), E-cadherin (22 of 27 cases; 82%), and HER2/neu (11 of 27 cases; 41%). Contrasting expression of these molecules in the primary versus the metastatic tumors, enhanced expression of CD44v6 was observed in the cutaneous metastases relative to the primary in 6 of 10 (60%) cases. Of interest, 2 of these 6 cases (33%) also showed reduction in E-cadherin--a member of the cadherin family functioning as an invasion suppressor molecule.

Conclusions.--These findings reinforce the complexities of the metastatic cascade and imply that the variation in adhesive properties of tumor cells is, perhaps, a consequence of the difference in density of the molecules mediating this process.

(Arch Pathol Lab Med. 2009;133:1285-1290)
Article Type: Report
Subject: Adenocarcinoma (Development and progression)
Adenocarcinoma (Care and treatment)
Metastasis (Development and progression)
Metastasis (Care and treatment)
Cadherins (Properties)
Cellular signal transduction (Genetic aspects)
Gene expression (Physiological aspects)
Authors: Pozdnyakova, Olga
Hoang, Mai M.P.
Dresser, Karen A.
Mahalingam, Meera
Pub Date: 08/01/2009
Publication: Name: Archives of Pathology & Laboratory Medicine Publisher: College of American Pathologists Audience: Academic; Professional Format: Magazine/Journal Subject: Health Copyright: COPYRIGHT 2009 College of American Pathologists ISSN: 1543-2165
Issue: Date: August, 2009 Source Volume: 133 Source Issue: 8
Geographic: Geographic Scope: United States Geographic Code: 1USA United States
Accession Number: 230247119
Full Text: The growing incidence of adenocarcinoma has spurred interest in the identification of molecules or markers that delineate patients at a higher risk of metastases. The weakening of the cell-cell adhesion mechanisms is a basic prerequisite for tumor metastasis to occur. This weakening involves changes in homotypic cell-cell adhesion, heterotypic cell-cell adhesion, and interactions of cells with the extracellular matrix at the primary tumor site. (1) Each of the molecules selected in the current study contribute to some of the above steps, all of which appear crucial in the metastatic cascade. Briefly, HER2/neu is functionally integrated with the epithelial-mesenchymal transformation that the tumor cells have to undergo to acquire migratory capabilities; the lymphocyte homing receptor CD44v6, an ubiquitous cell surface adhesion molecule, is involved in cell-cell and cell-matrix organization; the physical association of E-cadherin and [beta]-catenin is believed to be critical to homotypic cell-cell adhesion. (2,3)

Conflicting data exist regarding the prognostic significance of these molecules. For example, although the link between overexpression of HER2/ neu and poor prognosis in breast cancer and, albeit in select studies, esophageal adenocarcinoma, is a well-established one, others have shown that its utility in the latter is limited to defining a particular histologic subtype of esophageal adenocarcinomas. (4) Also, although reduction or loss of expression of E-cadherin has been documented in neoplasms originating in the gastrointestinal tract, redistribution has been noted in others, including those originating in the thyroid and the breast. (5) These contradictory results encouraged us to perform a study to determine the expression of select proto-oncogenes and cellular adhesion molecules, with a view to defining a marker that helped identify a biologically aggressive neoplasm. A comparative study with the primary tumor, in those cases in which it was available, was also performed in an effort to better understand the development of a metastatic phenotype.


The study was approved by the UMass Medical Center (Worcester, Massachusetts) institutional review boards. Archival materials from 27 patients with a histologic diagnosis of cutaneous metastatic adenocarcinoma were identified and retrieved from the pathology files of UMass Medical Center. These included 10 cases in which biopsy specimens from the primary tumor were available for a comparative review. Briefly, the 2 main groups included adenocarcinoma originating in the breast (n = 10; 37%) and gastrointestinal tract (n = 10; 37%). The other 7 cases included 3 with an unknown primary (11%), 1 metastatic ovarian papillary serous adenocarcinoma (4%), 2 metastatic adenocarcinomas from the thyroid (7%), and 1 metastatic bronchial adenocarcinoma (4%). Adenocarcinomas originating in the gastrointestinal tract included primaries originating in the rectum (n = 6; 60%), pancreas (n = 3; 30%), and esophagus (n = 1; 10%). Histologic sections of all cases were rereviewed, and the diagnoses were confirmed independently by 2 dermatopathologists (M.M. and M.M.P.H.). All patient data had identities removed.

Immunohistochemical studies were performed on 5-[micro]m-thick sections of formalin-fixed (10% buffered formalin), paraffin-embedded tissue. Antigen retrieval was carried out with heat-induced epitope retrieval buffer in an 800-W microwave oven for 15 minutes for all stains. Immunohistochemical staining was performed using antibodies to E-cadherin (clone, CDH1; 1:3200; Dako, Carpinteria, California), [beta]-catenin (clone, 14; 1:100; BD Biosciences, San Jose, California), CD44 (CD44v6 was used; clone, 2F10; 1:200; R&D Systems, Minneapolis, Minnesota), and HER2/ neu (polyclonal rabbit; used neat; Dako). Positive controls varied with the antibody used and included the following: infiltrating lobular carcinoma of the breast for E-cadherin, papillary carcinoma of the thyroid for [beta]-catenin, benign breast tissue for CD44v6, and a positive control slide included in the US Food and Drug Administration-approved staining kit for HER2/neu. For negative controls, the primary antibody was replaced with antibody diluent. Staining patterns were interpreted relative to the pattern and intensity of staining in the positive control slide. The immunostains were reviewed independently by 2 dermatopathologists (M.M.P.H., M.M.), and disagreements were reviewed together to achieve a consensus score. Staining was scored as being diffuse, when more than 50% of the tumor cells were positive; as focal, when 10% to 49% of the tumor cells were positive; and as 1+/negative, when less than 10% of the cells were positive. A semiquantitative assessment of the density of antigen expression in primary tumors versus the cutaneous metastases was obtained by grading the intensity of staining on a gradient of 3+ to 1+.


The detailed results of our study are presented in Tables 1 and 2. Representative examples of hematoxylin-eosin staining of primary adenocarcinoma originating in the breast (Figures 1, A, and 4, A), in the esophagus (Figure 2, A), and in the rectum (Figure 3, A), and their corresponding metastases (Figures 1, E; 2, E; and 3, E, respectively) are shown.


Cytoplasmic and membranous staining was observed in 22 of 27 cases (81%), with diffuse staining in 18 cases (67%) and focal staining in 4 of 27 cases (15%; Table 1). No differences in expression were observed in primary versus metastatic tumors in 9 of 10 cases (90%; Table 2; Figures 2, B and F, and 3, B and F). Complete loss of expression in the metastases, relative to staining in the primary tumor, was observed in one case (case 1; Figure 1, B and F). In 7 cases, E-cadherin expression was significantly lower than that of [beta]-catenin (cases 5, 7, 17, 19, 20, 22, and 25; Figure 4, B and C).


Cytoplasmic and membranous staining was observed in 26 of 27 cases (96%), with diffuse staining in 25 (93%) and focal staining in 1 of 27 cases (4%; Table 1). No differences in expression were observed in either the primary tumor or the metastases in 9 of 10 cases (90%; Table 2; Figures 2, C and G, and 3, C and G). Complete loss of expression in the metastases, relative to staining in the primary tumor, was observed in one case (case 1) (Figure 1, C and G).


Cytoplasmic and membranous staining was observed in 24 of 27 cases (89%), with diffuse staining in 18 (67%) and focal staining in 6 of 27 cases (22%; Table 1). Relative to the primary, 6 of 10 cases (60%) demonstrated an increased staining in the cutaneous metastasis (cases 1, 2, 11, 12, 20, and 21/22) , whereas the rest showed no difference between the primary tumor and the metastases (Table 2; Figures 1, D and H; 2, D and H; and 3, D and H).


Cytoplasmic and membranous staining was observed in 11 of 27 cases (41%), with diffuse staining in 5 (19%) and focal staining in 6 of 27 cases (22%; Table 1). Relative to the primary tumor, 1 of 10 cases (10%; case 1) demonstrated downregulation in the cutaneous metastasis (Table 2).



The CD44 family, generated from a single gene by alternative splicing and differential glycosylation, is important in a variety of physiologic and pathologic processes, including the development of metastasis. (6) Splice variants of the lymphocyte homing receptor CD44 first attracted attention when it was discovered that its expression conferred a metastatic phenotype. (7) Briefly, CD44, containing a segment encoded by variant exon 6, was found to confer full metastatic potential on a rat carcinoma cell line. Furthermore, overexpression of a variant of CD44v6 in nonmetastatic tumor cell lines was found to initiate metastatic behavior. Homologous variants of CD44v6 overexpressed in human tumors suggest that these molecules may also have a role to play in tumor progression and metastases in humans, corroborated by several reports indicating that expression of CD44v6 is associated with poor prognosis in several human tumors, particularly those originating in the gastrointestinal tract. (8) Our findings support this role because we found expression of CD44v6 in 90% of the cutaneous metastases originating in the gastrointestinal tract. However, we also found expression of CD44v6 in 90% of cutaneous metastases originating in the breast and in 85% of tumors originating in sites other than the breast and gastrointestinal tract. Of interest, enhanced expression of CD44v6 in the cutaneous metastases relative to the primary was detected in 60% (6 of 10 cases), 2 of which showed reduction in E-cadherin. Enhanced expression of CD44v6 in the metastases translates to a reduction in the ability of cells to metastasize, whereas loss of E-cadherin, a transmembrane glycoprotein localized in the adherens junction, has been shown to correlate with an increased ability to invade and metastasize. (9) Contrary as these results might seem, early studies have shown that not only do differences exist between cells of varying metastatic potential, but individual cells isolated from a primary tumor also show considerable diversity in expression of molecules on the cell surface. (10) Furthermore, our findings on CD44v6 expression contrast with the findings from another study (11) in which significantly higher levels of CD44v6 expression were found in the primary tumor relative to the metastases. However, in the study by Bendardarf et al, (11) the antibody used was different (20E6), raising the possibility that it may have been directed against a different epitope. Also, 42 of the 46 metastases (91%) were noncutaneous and even in the remaining 4 (labeled other; 9%) skin was not a specified site. This suggests that differences in CD44v6 expression might not occur until the circulating tumor cells have established a viable colony at the site of metastasis and implicates involvement of genes active in the final stages of tumor colonization. Favoring the importance of milieu is a previous study showing that genetic alterations of the metastatic human melanoma cell line C8161 suppress metastasis at a step subsequent to tumor cells entering the bloodstream. Briefly, these cells lodge in the lungs but fail to proliferate. Of interest, cells isolated from the lung formed tumors when injected into the skin, indicating that they were still viable, but dormant. (12)




Cadherins are thought to be the prime mediators of cell-cell adhesion in normal cells. (9) Reduction in expression of E-cadherin promotes epithelial cell invasiveness, dedifferentiation, and metastasis supporting a role for this protein as an "invasion suppressor molecule." In our study, 18 of 27 cases (67%) showed diffuse expression of E-cadherin. This conflicts with findings from several other studies (13-17) in which reduction or loss of expression in E-cadherin, associated with an increased ability to metastasize, has been noted in adenocarcinomas originating in the gastrointestinal tract, breast, ovary, and even the thyroid gland. The discrepancy may be a function of the usage of different antibodies with different sensitivities (clone CDH1 used in the current study versus clones 5H9 or NCH38 used in others), clonal differences in antigen expression, differences in tumor population studied (adenocarcinoma originating from heterogeneous primaries in the current study versus adenocarcinoma originating from primarily a single organ), or even a consequence of the grade of the malignancy studied (in extramammary Paget disease cases, documenting loss of E-cadherin, were all invasive or metastatic lymph node samples). (15,16,18) Variability in E-cadherin expression has also been attributed to the influence of the microenvironment. (19) Favoring this concept, in one study on nevomelanocytic proliferations, a gradient in expression of E-cadherin was observed from junctional to deep dermal localization of both nevi and melanomas. An obvious shortcoming of our study is that immunohistochemistry is, at best, a semiquantitative technique. Although grading the intensity of staining in primary tumors versus metastases offers a relative estimate of the densities of antigen expression, it does not really distinguish between an increase in adhesion molecule expression consequent to an increase in cell numbers versus that from an actual increase in density of antigen expression. (20) The latter might partly explain the lack of a difference in expression of E-cadherin between primary tumors and metastases noted in all but one of our cases. The cytoplasmic domain of E-cadherin is closely associated with a [beta]-catenin, which plays an important role in tethering E-cadherin to the cytoskeleton. (1) Our findings indicate that despite this physical linkage, changes in expression of E-cadherin and [beta]-catenin do not necessarily parallel each other (Figure 4, B and C). In 7 of 27 cases (26%) of metastatic adenocarcinoma, E-cadherin expression was significantly lower than that of [beta]-catenin (cases 5, 7, 17, 19, 20, 22, and 25). This implies the recruitment of other factors in adherens junction disassembly. For example, one could speculate that [beta]-catenin binds to the cytoplasmic domain of other molecules, such as N-cadherin, in the same way it binds to E-cadherin in normal skin.

Direct experimental evidence for a causal link between overexpression of HER2/ neu, a molecule with intrinsic tyrosine kinase activity, and tumor progression has come from several experimental studies showing that cells transfected with HER2/ neu are invasive in vitro and metastatic in vivo and that HER2/neu-expressing tumors in transgenic mice are capable of metastasis. (1,4) The association between gene amplification, overexpression of HER2/ neu, and poor prognosis in breast cancers is a well-established one. (21) In our cohort of cases, no similar conclusion could be drawn, although the obvious limiting factor is that there were only 3 cases in which the primary was available for a comparative review. In the 2 cases (cases 1 and 2) with amplification and overexpression of HER2/ neu in the primary tumor relative to the metastases, time to onset of cutaneous metastasis varied greatly (3 versus 26 months, respectively).

The analyses of expression of CD44v6, E-cadherin, and [beta]-catenin in both the primary tumor and the corresponding cutaneous metastases offer a unique opportunity to understand dynamic changes in expression of these molecules. Of the molecules in the current study, CD44v6 appears to hold promise because its expression on lesional cells translates to a tumor with an increased propensity to metastasize as a result of its ability to transcend "anchoring" forces. Enhanced expression of CD44v6, relative to the primary tumor in 6 of 10 cutaneous metastases (60%), suggests that genes acting at the final stages of the metastatic cascade may affect the ability of the cells to survive and proliferate at a specific secondary site. That 4 of these 6 cases (67%) did not exhibit significant differences in expression of molecules of the cadherin family, which typically function as invasion suppressor molecules, reinforces the complexities of the metastatic cascade. An additional study analyzing expression of these molecules in cutaneous versus noncutaneous metastases might provide a better insight into the definitive role of these molecules and development of the metastatic phenotype.


(1.) Nair KS, Naidoo R, Chetty R. Expression of cell adhesion molecules in esophageal carcinoma and its prognostic value. J Clin Pathol. 2005;58(4):343-351.

(2.) Guarino M. Epithelial-mesenchymal transition and tumour invasion. Int J Biochem Cell Biol. 2007;39:2153-2160.

(3.) Hirohashi S, Kanai Y. Cell adhesion system and human cancer morphogenesis. Cancer Sci. 2003;94(7):575-581.

(4.) Spencer KS, Graus-Porta D, Leng J, Hynes NE, Klemke RL. ErbB2 is necessary for induction of carcinoma cell invasion by ErbB family receptor tyrosine kinases. J Cell Biol. 2000;148(2):385-397.

(5.) Okegawa T, Pong RC, Li Y, Hsieh JT. The role of cell adhesion molecule in cancer progression and its application in cancer therapy. Acta Biochim Pol. 2004; 51(2):445-457.

(6.) Marhaba R, Zoller M. CD44 in cancer progression: adhesion, migration and growth regulation. J Mol Histol. 2004;35(3):21 1-231.

(7.) Hanley WD, Napier SL, Burdick MM, Schnaar RL, Sackstein R, Konstantopoulos K. Variant isoforms of CD44 are P- and L-selectin ligands on colon carcinoma cells. FASEB J. 2006;20(2):337-339.

(8.) Vizoso FJ, Fernandez JC, Corte MD, et al. Expression and clinical significance of CD44V5 and CD44V6 in resectable colorectal cancer. J Cancer Res Clin Oncol. 2004;130(11):679-686.

(9.) Yasui W, Oue N, Aung PP, Matsumura S, Shutoh M, Nakayama H. Molecular-pathological prognostic factors of gastric cancer: a review. Gastric Cancer. 2005;8(2):86-94.

(10.) Fidler IJ, Gersten DM, Hart IR. The biology of cancer invasion and metastasis. Adv Cancer Res. 1978;28:149-250.

(11.) Bendardaf R, Algars A, Elzagheid A, et al. Comparison of CD44 expression in primary tumors and metastases of colorectal cancer. Oncol Rep. 2006;16(4): 741-746.

(12.) Goldberg SF, HarmsJF,Quon K, Welch DR. Metastasis-suppressed C8161 melanoma cells arrest in lung but fail to proliferate. Clin Exp Metastasis. 1999; 17(7):601-607.

(13.) Nair KS, Naidoo R, Chetty R. Expression of cell adhesion molecules in oesophageal carcinoma and its prognostic value. JClin Pathol. 2005;58(4):343 351.

(14.) Ozawa M, Baribault H, Kemler R. The cytoplasmic domain of the cell adhesion molecule uvomorulin associates with three independent proteins structurally related in different species. EMBO J. 1989;8(6):1711-1717.

(15.) Dorudi S, Sheffield J, Faulsom R, et al. E-cadherin expression in colorectal cancer. Am J Pathol. 1993;142:981-986.

(16.) Gamallo C, Palacios J, Suarez A, et al. Correlation of E-cadherin expression with differentiation grade and histological type in breast carcinoma. Am J Pathol. 1993;142(4):987-993.

(17.) Vleminckx K, Vakaet LJ, Mareel M, et al. Genetic manipulation of E-cadherin expression epithelial tumour cells reveals an invasion suppressor role. Cell. 1991;66:107-119.

(18.) Liu H, Urabe K, Uchi H, et al. Expression and prognostic significance of Stat5a and E-cadherin in extramammary Paget's disease. J Cutan Pathol. 2007; 34(1):33-38.

(19.) Krengel S, Groteluschen F, Bartsch S, Tronnier M. Cadherin expression pattern in melanocytic tumors more likely depends on the melanocyte environment than on tumor cell progression. J Cutan Pathol. 2004;31(1):1-7.

(20.) Mahalingam M. Analysis of surface density of expression of molecules by flow cytometry. Cytometry. 1 996;24(2):1 90.

(21.) Eccles SA. The role of c-erbB-2/HER2/neu in breast cancer progression and metastasis. J Mammary Gland Biol Neoplasia. 2001;6(4):393-406.

Olga Pozdnyakova, MD, PhD; Mai M. P. Hoang, MD; Karen A. Dresser, BS; Meera Mahalingam, MD, PhD, FRCPath

Accepted for publication October 16, 2008.

From the Department of Pathology, University of Massachusetts Medical School, Worcester (Dr Pozdnyakova and Ms Dresser); the Department of Pathology, Massachusetts General Hospital and Harvard Medical School, Boston (Dr Hoang); and the Dermatopathology Section, Department of Dermatology, Boston University School of Medicine, Boston, Massachusetts (Dr Mahalingam).

The authors have no relevant financial interest in the products or companies described in this article.

Reprints: Meera Mahalingam, MD, PhD, FRCPath, Department of Dermatology, Boston University School of Medicine, 609 Albany St, J-301, Boston, MA 02118 (e-mail:
Table 1. Expression of Molecules in Metastatic Adenocarcinoma

Case No.   Diagnosis                E-Cadherin   [beta]-Catenin

   1       Breast primary (a)         -            -
   2       Breast primary (a)       D + + +      D + + +
   3       Breast primary (a)       D + + +      D + + +
   4       Breast primary           D + +        D + +
   5       Breast primary             -          D + +
   6       Breast primary           D + + +      D + +
   7       Breast primary             -          D + + +
   8       Breast primary           D + +        D + + +
   9       Breast primary           D + +        D + + +
   10      Breast primary           D + + +      D + + +
   11      Colorectal primary (a)   D +          D + + +
   12      Colorectal primary (a)   D + +        D + + +
   13      Colorectal primary (a)   D + +        D + + +
   14      Colorectal primary       D + +        D + + +
   15      Colorectal primary       D + +        D + + +
   16      Colorectal primary       D +          D + + +
   17      Pancreatic primary (a)   F + / + +    D + + +
   18      Pancreatic primary       D + +        D + + +
   19      Pancreatic primary         -          F + +
   20      Esophageal primary (a)   F +          D + + +
   21      Thyroid primary (a)      F + / + +    D + + +
   22      Thyroid primary (a)      F +          D + +
   23      Ovarian primary (a)      D + + +      D + + +
   24      Lung primary (a)         D + + +      D + + +
   25      Unknown primary            -          D + +
   26      Unknown primary          D + + +      D + + +
   27      Unknown primary          D +          D + + +

Case No.   CD44v6      HER2/neu

   1       D + +         -
   2       D + + +     D + + +
   3       -           F + +
   4       D + +       F +
   5       D + +       D +
   6       D + +       D +
   7       D + +         -
   8       D + + +     F +
   9       D + + +     F +
   10      D + + +       -
   11      D + -         -
   12      D + / + +   D +
   13      F + -         -
   14        -           -
   15      D + -         -
   16      D +         D +
   17      F + +         -
   18      D + + +       -
   19      D + -         -
   20      F + +         -
   21      D + + +       -
   22      D + + +       -
   23      F + / + +   F +/+ +
   24      F + -         -
   25      D + +         -
   26        -         F +
   27      F + -         -

Abbreviations: D, diffuse; F, focal; -, negative.

(a) Case in which the primary tumor was available for comparison.

Table 2.   Comparison of Expression in Primary Tumor Versus Metastases

                       Onset of Cutaneous
Case No.   Site          Metastasis, mo                  E-Cadherin

    1      Breast               3           Primary      D + + +
                                            Metastasis     -
    2      Breast              26           Primary      D + + +
                                            Metastasis   D + + +
    3      Breast               0           Primary      D + +
                                            Metastasis   D + + +
   11      Rectum              17           Primary      D + +
                                            Metastasis   D +
   12      Rectum              71           Primary      D +
                                            Metastasis   D + +
   13                                       Metastasis   D + +
   17      Pancreas             7           Primary      D + + +
                                            Metastasis   F + /+ +
   20      Esophagus            4           Primary      D +
                                            Metastasis   F +
   16      Thyroid             34           Primary      F +
                                            Metastasis   F + /+ +
   17                                       Metastasis   F +
   23      Ovary                0           Primary      D +
                                            Metastasis   D + + +
   24      Lung                 1           Primary      D + +
                                            Metastasis   D + + +

Case No.   [beta]-catenin   CD44v6     HER2/neu

    1      D + + +          F +        D +/+ +
             -              D + +        -
    2      D + + +          D + +      D + + +
           D + + +          D + + +    D + + +
    3      D + + +          F +        F +
           D + + +            -        F + +
   11      D + + +          F + /+ +     -
           D + + +          D +          -
   12      D + + +          F + /+ +   D +
           D + + +          D + /+ +   D +
   13      D + + +          F +          -
   17      D + + +          F + +        -
           D + + +          F + +        -
   20      D + + +            -          -
           D + + +          F + +        -
   16      D + +            D + +      F +
           D + + +          D + + +      -
   17      D + +            D + + +      -
   23      D + + +          F + +        -
           D + + +          F + /+ +   F +/+ +
   24      D + + +          F +          -
           D + + +          F +          -

Abbreviations: D, diffuse; F, focal; -, negative.
Gale Copyright: Copyright 2009 Gale, Cengage Learning. All rights reserved.