Printer Friendly

Cytokeratin and epithelial membrane antigen expression in angiosarcomas: an immunohistochemical study of 33 cases.

* Context.--Expression of epithelial cell markers can occur in mesenchymal tumors and has been reported in angiosarcomas with variable frequency. In these situations, establishing the diagnosis becomes problematic.

Objective.--To determine the expression of cytokeratin and epithelial membrane antigen in angiosarcoma.

Design.--To address this issue, 33 well-documented cases of angiosarcomas were retrieved from the archival material of Shands Hospital at the University of Florida, Gainesville, and Jackson Memorial Hospital at the University of Miami, Miami, Florida. These cases were all reviewed and studied using a cytokeratin cocktail (CAM 5.2 and AE1/AE3) and epithelial membrane antigen using standard immunohistochemical techniques. All 33 cases had available material for cytokeratin analysis; however, only 20 cases had enough material for epithelial membrane antigen staining.

Results.--In the 33 cases studied, the age range of the patients was 2 to 88 years (mean, 63 years). There were 23 (70%) men and 10 (30%) women. One (3%) of 33 was cytokeratin-immunoreactive and 2 (10%) of 20 were epithelial membrane antigen-immunoreactive.

Conclusion.--Cytokeratin and epithelial membrane antigen immunoreactivity in angiosarcomas is infrequent but may be encountered. Interpretation of such expression should be done with caution and in conjunction with the characteristic clinical and morphologic features of the tumor as well as the expression of endothelial cell antigens.

(Arch Pathol Lab Med. 2007;131:288-292)

Angiosarcomas are vascular tumors arising from the endothelial cells of blood vessels and lymphatics. Although their morphologic and immunohistochemical features are characteristic, occasionally surgical pathologists may encounter problematic cases. Because these tumors arise from endothelial cells, they usually express factor VIII-related antigen and/or ulex europaeus agglutinin-binding protein. Some of these tumors may exhibit an epithelial morphology mimicking epithelial tumors, and hence are given the term epithelioid angiosarcoma. Immunoreactivity of angiosarcoma to epithelial markers has been subject to scattered reports that described expression of epithelial markers mainly in the epithelioid variant of angiosarcoma. (1-6) Expression of these markers in angiosarcoma was found to depend on their morphology and was more common in the epithelioid variants. (7,8) Using immunohistochemical stains, Miettinen and Fetsch (9) have found that normal endothelial cells and a spectrum of vascular neoplasms can express different types of keratins, some of which exhibit consistent focal immunoreactivity. The same finding was also replicated in cultured human fibroblast, endothelial cells, and their sarcomas by Katagata et al (10) using 2-dimensional gel electrophoresis and electron microscopy techniques. The aim of our study was to evaluate the expression of cytokeratin (CK) and epithelial membrane antigen (EMA) in angiosarcomas regardless of their morphologic appearance. A total of 33 well-documented cases of angiosarcoma were studied. Well-controlled standard immunohistochemical stains for CK and EMA expression were performed.


Formalin-fixed, paraffin-embedded tissue blocks and unstained slides from 33 well-documented cases of angiosarcoma were retrieved from the pathology archives of the University of Florida-affiliated hospitals in Gainesville, and the University of Miami-affiliated hospitals in Miami, Florida, during a period of 8 years. All cases were reviewed; none were reclassified. All the cases exhibited the morphologic features of angiosarcoma and at least 2 vascular immunohistochemical markers were expressed by all but one case (Table, case 11) that was positive for CD34 only. To the best of our knowledge, all cases were considered primary at the time of diagnosis. Well-controlled immunohistochemical stains for CK and EMA (DakoCytomation, Carpinteria, Calif) were performed on all cases with sufficient histologic material according to established protocols using the diaminobenzidine detection with microwave antigen retrieval in a citrate buffer. To increase the sensitivity of CK expression, a cocktail of antibodies against AE1/AE3 (Dako) and CAM 5.2 (Becton Dickinson, San Jose, Calif) was used. Sufficient histologic material for CK immunostaining was available for all cases. However, because some of these cases had been referred for consultation and therefore had limited material for review and immunohistochemical stains, only 20 cases had enough material for EMA immunostaining. Positive staining was considered if more than 10% of tumor cells showed obvious strong immunoreactivity. All cases were also stained with antibodies for factor VIII-related antigen (Dako), ulex europaeus agglutinin (Dako), and CD34 (BioGenex, San Ramon, Calif) using the same standard technique.


The clinical and pathologic features of all our cases are shown in the Table. Immunoreactivity for CK was found in 1 (3%) of the 33 cases; 2 (10%) of 20 cases showed immunoreactivity for EMA (Table; Figures 1 through 3). Immunoreactivity for CK and EMA was noticed in the neoplastic cells with epithelioid as well as spindle morphology (Figure 1, C; Figure 2, C; Figure 3, B). As shown in the figures, few tumor cells were noticed to be positive in Figure 1, C and Figure 2, C but there were relatively more cells in the third case (Figure 3, B). That case was a liver needle biopsy and the volume of tumor present was limited. All cases, except one (case 11), were immunoreactive for at least 2 additional vascular markers (factor VIII-related antigen, ulex europaeus agglutinin, and CD34).


Case 1

The first case was a 66-year-old woman who had an ulcerating lesion in her buttock region. Low-power examination revealed hemorrhagic vascular neoplasm with anastomosing vascular channels lined by atypical plump endothelial cells (Figure 1, A and B). Focal areas were relatively more compact (left side of Figure 1, A) with cells that were more polygonal and epithelioid. The tumor cells were clearly positive for cytoplasmic CK expression (Figure 1, C).

Case 2

Case 2 was an 86-year-old woman with a scalp lesion. At low-power examination, the tumor was located in the upper and deep dermis and was composed of atypical vascular spaces containing red blood cells and lined by atypical plump endothelial cells (Figure 2, A and B). The tumor cells were immunoreactive for EMA (Figure 2, C).

Case 3

The third case was a 74-year-old man who had a history of thorotrast exposure with subsequent development of angiosarcoma in the liver. The diagnosis was made on a core needle biopsy where the tumor cells occupied an area that was surrounded by the remaining liver parenchyma (Figure 3, A). All the cells in this case were EMA positive (Figure 3, B). In this case, the volume of tumor was small and our evaluation was limited to the findings in this biopsy.


The expression of epithelial markers such as CK by mesenchymal cells and their tumors has been subject to many reports. (1,2,4,5,7-22) As part of the intermediate filaments family, different CKs were shown to be expressed in both benign and malignant mesenchymal tissue specimens. (10,14) Given the fact that immunohistochemistry is now part of the usual workup in the diagnosis of neoplasms, caution must be exercised in interpreting the findings. When encountering a vascular tumor in which the morphology appears to be consistent with angiosarcoma, confirmation by standard immunohistochemical stains for the usual endothelial cell markers is essential. It is not uncommon for additional stains that sometimes include epithelial cell markers to be ordered as part of a diagnostic panel, or because the tumor exhibits focal epithelioid morphology. When CK and/or EMA immunoreactivity is found, the differential diagnosis should be expanded to include carcinomas, particularly those with prominent vascular pattern. In these situations, thorough evaluation of the neoplastic cells and their morphology in conjunction with their immunoreactivity for vascular and epithelial markers is very critical. All our cases were reviewed and the diagnosis of angiosarcoma was made on the basis of morphology and immunoreactivity of the neoplastic cells to a minimum of 2 endothelial cell markers (except one case). We opted to evaluate the CK and EMA immunoreactivity in our angiosarcoma cases regardless of their epithelioid morphology for the following reasons: (1) determination of the epithelioid morphology is sometimes subjective, (2) those areas are often focal, and (3) we wanted to see whether the immunoreactivity for the epithelial antigens is seen in tumor cells with a particular pattern.

Whether the presence of immunoreactivity is real or represents an unusual phenomena is still controversial. (21,23,24) Gown et al (14) proposed some possible answers including cross-reactivity, accessibility of different epitopes, contamination with epithelial cell proteins, and a proliferation phenomenon. The cross-reactivity hypothesis was disputed and could not be confirmed when immunoblotting techniques failed to demonstrate similar findings. (14) In addition, the possibility of cross-reactivity with a protein or proteins sharing some homology with the CKs by other mesenchymal markers such as desmin and vimentin was also found to be inaccurate. Carefully extracting mesenchymal tissue such as myometrial smooth muscles and using double-labeling experiments argued against true epithelial contamination. (14) Although it was noticed that similar expressions were found in fetal mesenchymal tissue, more in the first trimester than the second and third, it was speculated that the anomalous CK expression may represent a proliferation phenomenon. In later studies, investigations on vascular neoplasms revealed focal keratin expression that was preferentially present in the epithelioid areas. (9,25)

The number of lesions and neoplasms that express mesenchymal and epithelial markers continues to grow, making the interpretation of results more challenging. The expression of epithelial markers in angiosarcomas can occur and has been reported in up to 35% of the cases. (25) In fact, Meis-Kindblom and Kindblom, (25) in their ultrastructural analysis of certain poorly differentiated areas of their angiosarcoma cases, encountered whorls of abundant intermediate filaments and tonofilament-like structures, which may explain the keratin expression.

Battifora (26) has alluded to the fact that our current definition of epithelia may be too rigid. Endothelial cells are the origin of angiosarcoma and share features with mesothelial and synovial cells that produce biphasic tumors with coexpression of both keratin and vimentin. Evidence that the neoplastic endothelial cells may produce epithelial markers is not surprising. However, whether this phenomenon is anarchic or atavistic in nature is still uncertain. (26) The explanation of this phenomenon is still controversial and in the hypothesis/theory phase. (21,23,25)

The authors thank M. Nadji, MD, and M. Nassiri, MD, for their contributions.


(1.) Gray MH, Rosenberg AE, Dickersin GR, Bhan AK. Cytokeratin expression in epithelioid vascular neoplasms. Hum Pathol. 1990;21:212-217.

(2.) Falk S, Krishnan J, Meis JM. Primary angiosarcoma of the spleen: a clinicopathologic study of 40 cases. Am J Surg Pathol. 1993;17:959-970.

(3.) Mills SE, Gaffey MJ, Watts JC, et al. Angiomatoid carcinoma and 'angiosarcoma' of the thyroid gland: a spectrum of endothelial differentiation. Am J Clin Pathol. 1994;102:322-330.

(4.) Christensen L, Schiodt T, Blichert-Toft M. Sarcomatoid tumours of the breast in Denmark from 1977 to 1987: a clinicopathological and immunohistochemical study of 100 cases. Eur J Cancer. 1993;29A:1824-1831.

(5.) Ben-Izhak O, Auslander L, Rabinson S, Lichtig C, Sternberg A. Epithelioid angiosarcoma of the adrenal gland with cytokeratin expression: report of a case with accompanying mesenteric fibromatosis. Cancer. 1992;69:1808-1812.

(6.) Wenig BM, Abbondanzo SL, Heffess CS. Epithelioid angiosarcoma of the adrenal glands: a clinicopathologic study of nine cases with a discussion of the implications of finding ''epithelial-specific'' markers. Am J Surg Pathol. 1994;18: 62-73.

(7.) Meis-Kindblom JM, Kindblom LG. Angiosarcoma of soft tissue: a study of 80 cases. Am J Surg Pathol. 1998;22:683-697.

(8.) Hasegawa T, Fujii Y, Seki K, et al. Epithelioid angiosarcoma of bone. Hum Pathol. 1997;28:985-989.

(9.) Miettinen M, Fetsch JF. Distribution of keratins in normal endothelial cells and a spectrum of vascular tumors: implications in tumor diagnosis. Hum Pathol. 2000;31:1062-1067.

(10.) Katagata Y, Takeda H, Ishizawa T, Hozumi Y, Kondo S. Occurrence and comparison of the expressed keratins in cultured human fibroblasts, endothelial cells and their sarcomas. J Dermatol Sci. 2002;30:1-9.

(11.) Anthony PP, Ramani P. Endothelial markers in malignant vascular tumours of the liver: superiority of QB-END/10 over von Willebrand factor and Ulex europaeus agglutinin 1. J Clin Pathol. 1991;44:29-32.

(12.) Fletcher CD, Beham A, Bekir S, Clarke AM, Marley NJ. Epithelioid angiosarcoma of deep soft tissue: a distinctive tumor readily mistaken for an epithelial neoplasm. Am J Surg Pathol. 1991;15:915-924.

(13.) Fletcher CD, Beham A, Schmid C. Spindle cell haemangioendothelioma: a clinicopathological and immunohistochemical study indicative of a non-neoplastic lesion. Histopathology. 1991;18:291-301.

(14.) Gown AM, Boyd HC, Chang Y, Ferguson M, Reichler B, Tippens D. Smooth muscle cells can express cytokeratins of ''simple'' epithelium: immunocytochemical and biochemical studies in vitro and in vivo. Am J Pathol. 1988;132:223-232.

(15.) Granger JK, Houn HY, Collins C. Massive hemorrhagic functional adrenal adenoma histologically mimicking angiosarcoma: report of a case with immunohistochemical study. Am J Surg Pathol. 1991;15:699-704.

(16.) Hojo H, Newton WA Jr, Hamoudi AB, et al. Pseudosarcomatous myofibroblastic tumor of the urinary bladder in children: a study of 11 cases with review of the literature--an Intergroup Rhabdomyosarcoma Study. Am J Surg Pathol. 1995;19:1224-1236.

(17.) Litzky LA, Brooks JJ. Cytokeratin immunoreactivity in malignant fibrous histiocytoma and spindle cell tumors: comparison between frozen and paraffin-embedded tissues. Mod Pathol. 1992;5:30-34.

(18.) Marrogi AJ, Hunt SJ, Cruz DJ. Cutaneous epithelioid angiosarcoma. Am J Dermatopathol. 1990;12:350-356.

(19.) McCluggage WG, Clarke R, Toner PG. Cutaneous epithelioid angiosarcoma exhibiting cytokeratin positivity. Histopathology. 1995;27:291-294.

(20.) O'Connell JX, Kattapuram SV, Mankin HJ, Bhan AK, Rosenberg AE. Epithelioid hemangioma of bone: a tumor often mistaken for low-grade angiosarcoma or malignant hemangioendothelioma. Am J Surg Pathol. 1993;17:610-617.

(21.) Penneys NS, Nadji M, Ziegels-Weissman J, Morales AR. Prekeratin in spindle cell tumors of the skin. Arch Dermatol. 1983;119:476-479.

(22.) Perez-Atayde AR, Achenbach H, Lack EE. High-grade epithelioid angiosarcoma of the scalp: an immunohistochemical and ultrastructural study. Am J Dermatopathol. 1986;8:411-418.

(23.) Norton AJ, Thomas JA, Isaacson PG. Cytokeratin-specific monoclonal antibodies are reactive with tumours of smooth muscle derivation: an immunocytochemical and biochemical study using antibodies to intermediate filament cytoskeletal proteins. Histopathology. 1987;11:487-499.

(24.) Ganjei P, Giraldo KA, Lampe B, Nadji M. Vulvar Paget's disease: is immunocytochemistry helpful in assessing the surgical margins? J Reprod Med. 1990;35:1002-1004.

(25.) Meis-Kindblom JM, Kindblom LG. Angiosarcoma of soft tissue: a study of 80 cases. Am J Surg Pathol. 1998;22:683-97.

(26.) Battifora H. Expression of keratins by endothelial cells: atavism or anarchy? Ultrastruct Pathol. 1990;14:iii-v.

Accepted for publication August 2, 2006.

Mousa A. Al-Abbadi, MD; Nidal M. Almasri, MD; Samer Al-Quran, MD; Edward J. Wilkinson, MD

From the Department of Pathology, Immunology and Laboratory Medicine, University of Florida, Gainesville. Dr Almasri is currently in the Department of Pathology, Jordan University of Science & Technology, Irbid, Jordan. Dr Al-Abbadi is currently in the Department of Pathology, Wayne State University/Detroit Medical Center, Harper University Hospital, Detroit, Mich.

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

Reprints: Mousa A. Al-Abbadi, MD, Wayne State University/Detroit Medical Center, Harper University Hospital, Department of Pathology, 3990 John R St, Detroit, MI 48201 (e-mail:
Clinical and Pathologic Features of Cases *

Case Age, Factor
No. y/Sex Organ VIII Ulex CD34

1 66/F Buttock + + +
2 86/F Scalp + + +
3 74/M Liver + + +
4 65/M Scalp + + w
5 61/M Scalp + w +
6 31/M Mandible + + +
7 2/M Liver + + +
8 73/M Knee - + +
9 77/M Scalp + + +
10 62/F Wrist + + +
11 79/M Skin - - +
12 66/M Skin + + +
13 24/M Atrium + + +
14 70/M Scalp + w +
15 72/M Liver + + +
16 30/F Axilla + + w
17 88/M Subcutaneous + + w
18 53/M Bone + + +
19 40/M Neck soft tissue + + +
20 58/F Breast + - +
21 61/F Rib w + +
22 36/M Liver + + +
23 77/M Skin + + w
24 57/F Small intestine + + w
25 69/M Larynx + + +
26 79/M Neck soft tissue + + +
27 82/F Scalp + + w
28 66/M Skin + + -
29 68/F Atrium - + +
30 84/M Scalp + w +
31 86/F Breast + + +
32 71/M Skin + + w
33 73/M Scalp + + +

Case Age,
No. y/Sex EMA CK Comments

1 66/F - + ...
2 86/F + - ...
3 74/M + - Thorotrast exposure
4 65/M - - Post RT
5 61/M - - ...
6 31/M - - ...
7 2/M - - ...
8 73/M - - ...
9 77/M - - Post RT
10 62/F - - Postmastectomy
11 79/M - - ...
12 66/M - - ...
13 24/M - - ...
14 70/M - - Post RT
15 72/M - - Vinyl chloride
16 30/F - - Mafucci syndrome
17 88/M - - ...
18 53/M - - ...
19 40/M - - ...
20 58/F - - ...
21 61/F NA - ...
22 36/M NA - Hepatitis C
23 77/M NA - ...
24 57/F NA - Post RT
25 69/M NA - ...
26 79/M NA - ...
27 82/F NA - ...
28 66/M NA - ...
29 68/F NA - ...
30 84/M NA - ...
31 86/F NA - ...
32 71/M NA - ...
33 73/M NA - ...

* EMA indicates epithelial membrane antigen; CK, cytokeratin;
+, positive; -, negative; w, weak; RT, radiotherapy; and NA, not
COPYRIGHT 2007 College of American Pathologists
No portion of this article can be reproduced without the express written permission from the copyright holder.
Copyright 2007 Gale, Cengage Learning. All rights reserved.

Article Details
Printer friendly Cite/link Email Feedback
Author:Abbadi, Mousa A. Al-; Almasri, Nidal M.; Quran, Samer Al-; Wilkinson, Edward J.
Publication:Archives of Pathology & Laboratory Medicine
Article Type:Clinical report
Geographic Code:1USA
Date:Feb 1, 2007
Previous Article:Follicular lymphoma in staging bone marrow specimens: correlation of histologic findings with the results of flow cytometry immunophenotypic analysis.
Next Article:Malakoplakia outside the urinary tract.

Terms of use | Privacy policy | Copyright © 2019 Farlex, Inc. | Feedback | For webmasters