APMIS 100: 901-907, 1992
Irnmunohistochemical evaluation of the cytoarchitecture of benign and malignant breast lesions YLERMI SOINI' and MARKKU MIETTINEN2 'Department of Pathology, University of Oulu, Finland and 'Thomas Jefferson University, Philadelphia, Pennsylvania, U S A .
Soini, Y. 8t Miettinen, M. Immunohistochemical evaluation of the cytoarchitecture of benign and malignant breast lesions. APMIS 100: 901-907, 1992. Fifty-three breast lesions, which had been fixed in formalin and embedded in paraffin, were immunohistochemically analyzed with monoclonal antibodies to cytokeratin subtypes 1, 5 , 10, 14 (34BE12), muscle-specific actins (HHF35) and antiserum to SlOO protein, all of which have been used as markers for myoepithelial cells. With these antibodies, a continuous myoepithelial cell layer could generally be seen around the benign ducts and acini. In in situ carcinomas, such a layer could still be observed, though it was usually discontinuous and sometimes absent. In infiltrating carcinomas, no myoepithelial cell layer could be observed. In intraductal hyperplasias, scattered HHF35,34BE12 and S 100-positive cells could be seen amongst the proliferating intraductal cells. In in situ and infiltrating carcinomas, however, such cells could also be observed. This was seen especially with antibodies 34BE12 and S100, and to a lesser extent also with HHF35. Morphologically these cells seemed to belong to the malignant cell population. Although myoepithelial cell preservation is an important morphological parameter in the histological evaluation of breast lesions, the results suggest that the myoepithelial cell markers 34BE12, HHF35 and SlOO cannot be used in the differential diagnosis between benign and malignant breast lesions in a straightforward manner. This is because in situ carcinomas have a more or less preserved myoepithelial cell layer, and because many infiltrating and in situ carcinomas contain a subpopulation of neoplastic cells expressing these markers, possibly signifying myoepithelial cell differentiation. Key words: Cytokeratin; muscle actin; S 100-protein; breast carcinoma. Ylermi Soini, Department of Pathology, University of Oulu, Kajaanintie 52 D, SF-90220 Oulu, Finland.
In breast tissues, myoepithelial cells are normally found around ductal and acinar structures. In benign ductal hyperplasia, myoepithelial cells can also be seen intraductally among the proliferating ductal cells. The absence of these cells in ductal proliferations has been used as a morphological criterion in the evaluation of malignancy (3, 19). In immunohistochemical investigations, benign ductal proliferations have been found to contain a subpopulation of intraductal cells re-
Received January 5 , 1992. Accepted May 25, 1992.
active for medium and high molecular weight cytokeratins, muscle-specific actins and S 100protein (1 1, 16). In in situ and infiltrating carcinomas a monotonous cell proliferation positive for lower molecular weight cytokeratins 8, 18 and 19 has been found, although a minority of tumors have also expressed medium and high molecular weight cytokeratins (1, 2, 7 , 8, 12, 14, 16). Occasional intraductal positivity for medium and high molecular weight cytokeratins in in situ carcinomas has been interpreted to signify a pagetoid spread of the tumor cells among the myoepithelial cell layer (1 1, 16). Positivity for HHF35 has not usually been reported in in situ or infiltrating carcinomas and
90 1
SOINI & MIETTINEN
an occasional positivity found in some in situ lesions has been interpreted to represent crossreactive phenomena with non-muscle actins (1 1, 16). However, positivity for CALLA and GFAP has been reported in some carcinomas, possibly signifying myoepithelial cell differentiation (7, 9)Positivity for S1 00-protein has been reported in breast carcinomas ( 5 , 20). In spite of this, SlOO protein has been used as a myoepithelial cell marker in breast tissues. In this immunohistochemical investigation we analyzed fifty-three breast lesions with antibodies 34BE12, HHF35 and SlOO in order to evaluate the distribution of myoepithelial cells in various pathological states of the breast, and how these antibodies could be used in the differential diagnosis of various breast lesions. The three antibodies were selected because they all function in formalin-fixed and paraffin-embedded tissues. MATERIALS AND METHODS Fifty-three cases of various breast lesions were collected from the files of the Department of Pathology at the Thomas Jefferson University Hospital between 1988 and 1990. All material had been fixed in neutral formalin and embedded in paraffin. The diagnosis of the cases was based on a light microscopic evaluation using conventional hematoxylin-eosin staining. The material consisted of seventeen infiltrating ductal carcinomas with an accompanying in situ component, eight ductal in situ carcinomas, six intraductal papillomas, three atypical intraductal hyperplasias, three intraductal hyperplasias, three intraductal papillomatoses, eight fibrocystic mastopathies and four sclerosing adenoses. A polyclonal rabbit antibody for SlOO and monoclonal mouse antibodies for muscle-specific actin
(HHF35) and cytokeratins I , 5, 10, 14 (34BE12) were used in the immunostainings. The antibody to 5100protein was purchased from Dakopatts (Carpinteria, Ca) and the other two from Enzo Biochemicals (New York, NY). Sections, 5 pm thick, were deparaffinized in xylene and rehydrated. The endogenous peroxidase was blocked with 0.3% hydrogen peroxide in methanol for 30 min. For 34BE12, a pepsin treatment (0.05% pepsin, 2500 U/g, Merck, Darmstadt, pH 1.8, for 30 min at 37°C) was used. The immunostaining was performed by the ABC technique (10) using a Vectastain ABC kit. The sections were first incubated in normal serum (goat for S100, horse for 34BE12 and HHF35) for 20 min at room temperature, followed by the primary antibody at a dilution of 1:800 for S100, 1:8000 for 34BE12 and 1:12800 for HHF35 for 30 min at room temperature, except for the slides stained with 34BE12, which were incubated overnight at 4°C. This was followed by incubations with the secondary biotinylated antibody and the ABC complex. Rinses with PBS were carried out between the various steps of the staining procedure. The color was developed using diaminobenzidine as a chromogen. The slides were then counterstained lightly with hematoxylin and mounted with Permount. The negative control consisted of omitting the primary antibody in each case.
RESULTS Benign breast ducts and acini were surrounded by a layer of HHF35-, 34BE 12- and S I OO-positive cells. In cystically dilated ducts, this layer, however, was attenuated and sometimes discontinuous. In smaller breast ducts and acini, a subgroup of luminal cells were positive for 34BE12 and S100. In benign hyperplasias, a myoepithelial cell layer could be seen which was positive with all three antibodies. In some hyperplastic lesions,
Fig. f. In this ductal carcinoma (case 2) the in situ component contains 34HB12-cytokeratin-positive neoplastic cells intraductally. The ducts are surrounded by 34HB 12-positivemyoepithelial cells. ( x 130, immunoperoxidase stain). Fig. 2. The same case stained with SIOO. Positive intraductal cells can also be seen with this antibody. ( x 130, immunoperoxidase stain). Fig. 3. The same case stained with HHF35. The periductal myoepithelial cells can be clearly seen. However, some positively stained neoplastic appearing cells can also be seen intraductally. ( x 130, immunoperoxidase stain). Fig. 4. This is an example of a ductal infiltrating carcinoma (case 1) expressing widespread immunoreactivity for 34BE12-cytokeratin. ( x 130, immunoperoxidase stain).
902
CYTOARCHITECTURE OF BREAST LESIONS
SOINl & MIETTINEN
Fig. 5. Another case of infiltrating ductal carcinoma (case 6 ) revealing S100-positive tumor cells in the immunostaining. ( x 130, immunoperoxidase stain).
Fig. 6. The same case stained with HHF35. Positivity for muscle-specific actins can be observed in tumor cells. ( x 130, immunoperoxidase stain).
the myoepithelial cell layer was sometimes attenuated and slightly discontinuous. Scattered HHF35-, 34BE12- and S100-positive cells could be seen intraductally. In intraductal papillomas, a myoepithelial cell layer positive for all three antibodies was seen around the proliferating luminal epithelium. The vascular structures of the papillary cores were HHF35 positive. Scattered S100- and 34BE 12-positive luminal cells could be seen intraductally, while all the luminal cells appeared to be negative for HHF35. In the sclerosing adenoses, a rich network of 34BE12-, S loo-, and HHF35-positive cells could be seen around the proliferating ducts. In atypical hyperplasias, scattered cells reactive for all three antibodies could be observed intraductally. The myoepithelial cell layer was largely preserved in these lesions. 904
In in situ carcinomas, a HHF35-, 34BE12and S100-positive cell layer could still be observed around the ducts, though it was discontinuous and focally absent (Figs. 1-3). The discontinuity was evaluated in in situ carcinomas and infiltrating carcinomas with a separate in situ component, and was found to be 44% for 34BE12,62% for HHF35 and 11% for S100. The in situ carcinoma component often contained neoplastic cells which showed variable positivity for 34BE12 and SlOO (Figs. 1 & 2). In some cases also cells positive for HHF35 were observed (Fig. 3). With 34BE12 a majority of the carcinomas analyzed contained positive tumor cells (18/25 cases positive) (Figs. 1 & 4). A substantial number of carcinomas were also SlOO positive (14/ 25 cases) (Figs. 2 & 5). In seven carcinomas scattered HHF35-positive tumor cells could be
CYTOARCHITECTURE OF BREAST LESIONS
TABLE 1. The results of the immunohistochemical stainings in carcinomas Case Diagnosis Percentage of positive tumor number cells 34BE12 HHF35 SlOO 1. IFC + ISC 95%) 2. 10% IFC ISC 10% ISC 3. 20% 10% 4. ISC 1OYO 10% IFC + ISC 20% 5. 30% 6. IFC + ISC 6Oy0 10% 10% 7. ISC 5% IFC + ISC 8. IFC + ISC 9. 5% 10. ISC 30% IFC + ISC 5% 11. 10% IFC + ISC 12. 5'1/0 5% IFC ISC 13. 10% IFC + ISC 14. 5% 5Yn 15. IFC ISC 10% 5% ISC 16. 5% 5% 17. IFC ISC ISC 18. 50% 40% IFC + ISC 5% 19. 5% ISC 20. 21. 5Y" ISC ISC 22. 23. IFC + ISC 5% IFC + ISC 24. 25. IFC ISC 10% Positive 18 7 14 (total) Abbreviations: IFC = Infiltrating ductal carcinoma ISC = In situ carcinoma.
+
+ + +
+
seen (Figs. 3 & 6). Positive cells were found both intraductally and in infiltrating areas of the breast carcinomas studied (see Table 1). In invasive carcinomas there was no positivity around the tumor cell islands with 34BE12 or S100. With HHF35 positive fusiform cells were often observed around these islands though morphologically they represented myofibroblastic cells. They were also seen in the stroma of the infiltrating carcinomas. DISCUSSION In this study fifty-three benign and malignant breast lesions were analyzed immunohistochemically with antibodies to cytokeratins 1, 5 , 10, 14 (34BE12), muscle-specific actin (HHF35) and
SlOO protein. These antibodies have been used to evaluate the preservation of the myoepithelial cell component, which is used as a criterion in the evaluation of breast malignancy (3, 19). While our findings concerning the benign breast lesions are in line with those reported by others (1, 2, 11, 14, 16), in carcinomas some results were at variance with previous reports. In our material a majority of carcinomas expressed variable positivity for 34BE12. Some were also positive for SlOO and a few even for HHF35. The findings suggest that breast carcinomas are immunohistochemically heterogeneous. Based on the results it seems possible that some tumor cells have differentiated towards myoepithelial cells. In some earlier immunohistochemical reports, only a minority of breast carcinomas were found to be positive for medium and high molecular weight cytokeratins (1, 2, 11, 14, 16). With antibodies KA1 and KA4 recognizing cytokeratins 5 and 14, and cytokeratins 14, 15, 16 and 19, respectively, only a few invasive carcinomas were positive with KA1, while all were positive for KA4 (1 1, 14). In another study evaluating the distribution of keratin 19 immunoreactivity in benign and malignant breast lesions, carcinoma cells seemed to express keratin 19, while benign proliferative lesions also contained a substantial number of keratin 19-negative cells (2). Some other authors, however, have found more positivity for medium and high molecular weight cytokeratins in breast carcinomas (7). Thus the different results may partly be due to use of different antibodies specific for different cytokeratin subtypes and possibly different epitopes of the molecules. Whilst in a recent report only a minority of ductal in situ carcinomas were positive for 34BE12-cytokeratin antibody, the authors found positive intraductal cells in ten additional cases, which they interpreted to represent nonneoplastic cells (1 6). The epithelium of the terminal ductal lobular units is generally considered to be the site of origin of most mammary carcinomas (14). The epithelium of these units has been found to be immunoreactive for both high and low molecular weight cytokeratins (14, 17). Our results are thus in accordance with the presumed origin of breast carcinomas, reflecting their capability also to express high and medium molecular 905
SOINI & MIETTINEN
weight cytokeratins. Because of the existence of high and medium molecular weight cytokeratin reactivity in nonmyoepithelial cells in terminal ductal and lobular units (17), the expression of these cytokeratins need not necessarily indicate true myoepithelial cell differentiation in tumor cells. In a minority of breast carcinomas (both in situ and invasive), however, reactivity for muscle-specific actins could be seen. This reactivity was located in the neoplastic cells, indicating that some malignant cells may express true myoepithelial cell differentiation. This possibility is also suggested by other immunohistochemical investigations which reveal CALLA (9), GFAP (7) or vimentin (4, 7) immunoreactivity in a minority of breast carcinomas. In some ultrastructural studies myoepithelial cell differentiation has also been described (6). Finally, myoepithelial cell differentiation can be seen in SV40-transformed breast epithelial cell lines (1 8). A substantial number of tumors were positive for S100. SlOO positivity has been described in myoepithelial cells and may be considered, in a restricted sense, to indicate myoepithelial cell differentiation (15). However, in benign nonmyoepithelial ductal cells in breast tissues, reactivity for SlOO can also be observed ( 5 ) . SlOO protein reactivity has, in addition, been shown in various types of mesenchymal and epithelial tumors, including breast carcinomas (15 , 20). In benign ductal hyperplasias and in situ carcinomas we evaluated the continuity of the myoepithelial cell layer around the ducts. In benign lesions this ranged from 70 to 100% depending on the antibody used (data not shown). In in situ carcinomas it ranged from 0 to 100% with mean values of 44%, 62% and 11% for 34BE12, HHF35 and S100, respectively (data not shown). The partial loss of the myoepithelial cell layer in intraductal hyperplasias and cystically dilated glands is possibly due to the expanded forces generated by the dilated ducts. Although in in situ carcinomas this may also play a part, most of the effect is probably due to different collagenases and proteases produced by the neoplastic cells which disrupt the basement membrane on which the myoepithelial cells rest. Production of such collagenases by the tumor cells has been shown immunohistochemically, and immunoreactivity for collagenase IV in the pro906
liferating cells has even been suggested as a marker for malignant breast disease (1 3). Another point to be noted is the difference between the immunoreactivities with the different antibodies. Since myoepithelial cells are most certainly responsible for the staining around the ductal structures, the ability of the antibodies to recognize myoepithelial cells can be tested in this way. According to the results, HHF35 is the most sensitive of the antibodies, while SlOO seems to be the least sensitive. This also suggests that not all 34BE12 and SlOO positivity found in carcinomas necessarily indicates myoepithelial cell differentiation. In conclusion, the results show that none of the antibodies employed in this investigation can be used to distinguish between benign and malignant breast epithelial proliferations because immunoreactivities for all antibodies can be shown in in situ and invasive carcinomas in addition to benign lesions. We are grateful to Mr A1 Kovatich, the Immunopathology Laboratory, Thomas Jefferson University Hospital, for assistance in the performance of the immunostainings. This study was supported by a grant from the European Cancer Society (UICC).
REFERENCES 1. Altsmannsberger, M., Dirk, T , Droese, M., Weber, K. & Osborn, M.: Keratin polypeptide distribution in benign and malignant breast tumors: subdivision of ductal carcinomas using monoclonal antibodies. Virchows Arch. B. 51: 265-275, 1986. 2. Bartek, J., Taylor-Papadimitriou, J,, Miller, N. & Millis, R. :Patterns of expression of keratin 19 as detected with monoclonal antibodies in human breast tissues and tumors. Int. J. Cancer 36: 299-306, 1985. 3. Carter, D.:Interpretation of breast biopsies. Biopsy interpretation series. Raven Press, New York 1984, pp. 25-75. 4. Domagalu, W , Lasota, J., Dukowicz, A , , Markiewski, M., Striker, G., Weber, K, & Osborn, M.: Vimentin expression appears to be associated with poor prognosis in node-negative ductal NOS breast carcinomas. Am. J. Pathol. 137: 1299-1304, 1990. 5 . Gillett, C. E., Bobrow, L. G. & Millis, R. R.: SlOO protein in human mammary tissue-immunoreactivity in breast carcinoma, including Paget’s disease of the nipple, and value as a marker of myoepithelial cells. J. Pathol. 160: 19-24, 1990.
CYTOARCHITECTURE OF BREAST LESIONS
6. Gould, K E., Juo, W & Buttiforu, H.: Ultrastructural analysis in the differential diagnosis of breast tumors. The significance of myoepithelial cells, basal lamina, intracytoplasmic lumina and secretory granules. Pathol. Res. Pract. 167; 45-70, 1980. 7. Gould, !l E., Koukoulis, G. K., Junsson, D. S., Nugle, R. B., Frunke, W W & Moll, R.: Coexpression patterns of vimentin and glial filament protein with cytokeratins in the normal, hyperplastic, and neoplastic breast. Am. J. Pathol. 137: 1143-1 155, 1990. 8. Gown, A . M. & Vogel, A. M.: Monoclonal antibodies to human intermediate filament proteins. 111. Analysis of tumors. Am. J. Clin. Pathol. 84: 413424, 1985. 9. Gusterson, B. A , , Monughun, I! M., Muhendrun, R., Ellis, J. & O’Hure, M. J.: Identification of myoepithelial cells in human and rat breasts by anti-common acute lymphoblastic leukemia antigen antibody A12. J. Natl. Cancer Inst. 77: 343-349, 1986. 10. Hsu, S. M., Ruine. L. & Funger, H.: Use of avidinbiotin-peroxidase complex (ABC) in immunoperoxidase techniques: a comparison between ABC and unlabelled antibody (PAP) procedures. J. Histochem. Cytochem. 29: 577-580, 1981. I 1. Jurusch, E.-D., Nugle, R. B., Kuufmann, M., Maurer, C. & Bocker, W J.: Differential diagnosis of benign epithelial proliferations and carcinomas of the breast using antibodies to cytokeratins. Hum. Pathol. 19: 276-289, 1988. 12. Moll, R., Frunke, W W , Schiller, D. L., Geiger, B. & Krepler, R.: The catalog of human cytokeratins: patterns of expression in normal epithelia, tumors and cultured cells. Cell 31: 11-24, 1982. 13. Monteugudo, C., Merino, M . J., Sun-Juan, J., Liottu, L. A . & Stetler-Stevenson, W G.:Immuno-
histochemical distribution of type IV collagenase in normal, benign and malignant breast tissue. Am. J. Pathol. 136: 585-592, 1990. 14. Nugle, R. B., Bocker, W, Davis, J. R., Heid, H. W , Kuufmunn, M., Lucus, D. & Jarusch, E.-D.: Characterization of breast carcinomas by two monoclonal antibodies distinguishing myoepithelial from luminal epithelial cells. J. Histochem. Cytochem. 34: 869-881, 1986. 15. Nukujimu, T , Watunube, S., Suto, II, Kumeyu, T , Hirotu, T & Shimosuto, II; An immunoperoxidase study of S-100 protein distribution in normal and neoplastic tissues. Am. J. Surg. Pathol. 6: 715-727, 1982. 16. Ruju, U., Crissmun, J. D., Zurbo, R. J. & Gottlieb, C.: Epitheliosis of the breast. An immunohistochemical characterization and comparison to malignant intraductal proliferations of the breast. Am. J. Surg. Pathol. 14: 939-947, 1990. 17. Rudlund, El S. & Hughes, C. H.: Immunocytochemical identification of cell types in human mammary gland: variations in cellular markers are dependent on glandular topography and differentiation. J. Histochem. Cytochem. 37: 1087-1 100, 1989. 18. Rudlund, l? S., Ollerheud, G.& Burruclough. R.: Isolation of simian virus 40-transformed human mammary epithelial stem cell lines that can differentiate to myoepithelial-like cells in culture and in vivo. Develop. Biol. 136: 167-180, 1989. 19. Sloune, J. El; Biopsy pathology of the breast. First Edition. Chapman & Hall, London 1985, pp. 93-123. 20. Stroup, R. M. & Pinkus, G. S.: S-100 immunoreactivity in primary and metastatic carcinoma of the breast: a potential source of error in immunodiagnosis. Hum. Pathol. 19: 949-953, 1988.
907
Irnmunohistochemical evaluation of the cytoarchitecture of benign and malignant breast lesions YLERMI SOINI' and MARKKU MIETTINEN2 'Department of Pathology, University of Oulu, Finland and 'Thomas Jefferson University, Philadelphia, Pennsylvania, U S A .
Soini, Y. 8t Miettinen, M. Immunohistochemical evaluation of the cytoarchitecture of benign and malignant breast lesions. APMIS 100: 901-907, 1992. Fifty-three breast lesions, which had been fixed in formalin and embedded in paraffin, were immunohistochemically analyzed with monoclonal antibodies to cytokeratin subtypes 1, 5 , 10, 14 (34BE12), muscle-specific actins (HHF35) and antiserum to SlOO protein, all of which have been used as markers for myoepithelial cells. With these antibodies, a continuous myoepithelial cell layer could generally be seen around the benign ducts and acini. In in situ carcinomas, such a layer could still be observed, though it was usually discontinuous and sometimes absent. In infiltrating carcinomas, no myoepithelial cell layer could be observed. In intraductal hyperplasias, scattered HHF35,34BE12 and S 100-positive cells could be seen amongst the proliferating intraductal cells. In in situ and infiltrating carcinomas, however, such cells could also be observed. This was seen especially with antibodies 34BE12 and S100, and to a lesser extent also with HHF35. Morphologically these cells seemed to belong to the malignant cell population. Although myoepithelial cell preservation is an important morphological parameter in the histological evaluation of breast lesions, the results suggest that the myoepithelial cell markers 34BE12, HHF35 and SlOO cannot be used in the differential diagnosis between benign and malignant breast lesions in a straightforward manner. This is because in situ carcinomas have a more or less preserved myoepithelial cell layer, and because many infiltrating and in situ carcinomas contain a subpopulation of neoplastic cells expressing these markers, possibly signifying myoepithelial cell differentiation. Key words: Cytokeratin; muscle actin; S 100-protein; breast carcinoma. Ylermi Soini, Department of Pathology, University of Oulu, Kajaanintie 52 D, SF-90220 Oulu, Finland.
In breast tissues, myoepithelial cells are normally found around ductal and acinar structures. In benign ductal hyperplasia, myoepithelial cells can also be seen intraductally among the proliferating ductal cells. The absence of these cells in ductal proliferations has been used as a morphological criterion in the evaluation of malignancy (3, 19). In immunohistochemical investigations, benign ductal proliferations have been found to contain a subpopulation of intraductal cells re-
Received January 5 , 1992. Accepted May 25, 1992.
active for medium and high molecular weight cytokeratins, muscle-specific actins and S 100protein (1 1, 16). In in situ and infiltrating carcinomas a monotonous cell proliferation positive for lower molecular weight cytokeratins 8, 18 and 19 has been found, although a minority of tumors have also expressed medium and high molecular weight cytokeratins (1, 2, 7 , 8, 12, 14, 16). Occasional intraductal positivity for medium and high molecular weight cytokeratins in in situ carcinomas has been interpreted to signify a pagetoid spread of the tumor cells among the myoepithelial cell layer (1 1, 16). Positivity for HHF35 has not usually been reported in in situ or infiltrating carcinomas and
90 1
SOINI & MIETTINEN
an occasional positivity found in some in situ lesions has been interpreted to represent crossreactive phenomena with non-muscle actins (1 1, 16). However, positivity for CALLA and GFAP has been reported in some carcinomas, possibly signifying myoepithelial cell differentiation (7, 9)Positivity for S1 00-protein has been reported in breast carcinomas ( 5 , 20). In spite of this, SlOO protein has been used as a myoepithelial cell marker in breast tissues. In this immunohistochemical investigation we analyzed fifty-three breast lesions with antibodies 34BE12, HHF35 and SlOO in order to evaluate the distribution of myoepithelial cells in various pathological states of the breast, and how these antibodies could be used in the differential diagnosis of various breast lesions. The three antibodies were selected because they all function in formalin-fixed and paraffin-embedded tissues. MATERIALS AND METHODS Fifty-three cases of various breast lesions were collected from the files of the Department of Pathology at the Thomas Jefferson University Hospital between 1988 and 1990. All material had been fixed in neutral formalin and embedded in paraffin. The diagnosis of the cases was based on a light microscopic evaluation using conventional hematoxylin-eosin staining. The material consisted of seventeen infiltrating ductal carcinomas with an accompanying in situ component, eight ductal in situ carcinomas, six intraductal papillomas, three atypical intraductal hyperplasias, three intraductal hyperplasias, three intraductal papillomatoses, eight fibrocystic mastopathies and four sclerosing adenoses. A polyclonal rabbit antibody for SlOO and monoclonal mouse antibodies for muscle-specific actin
(HHF35) and cytokeratins I , 5, 10, 14 (34BE12) were used in the immunostainings. The antibody to 5100protein was purchased from Dakopatts (Carpinteria, Ca) and the other two from Enzo Biochemicals (New York, NY). Sections, 5 pm thick, were deparaffinized in xylene and rehydrated. The endogenous peroxidase was blocked with 0.3% hydrogen peroxide in methanol for 30 min. For 34BE12, a pepsin treatment (0.05% pepsin, 2500 U/g, Merck, Darmstadt, pH 1.8, for 30 min at 37°C) was used. The immunostaining was performed by the ABC technique (10) using a Vectastain ABC kit. The sections were first incubated in normal serum (goat for S100, horse for 34BE12 and HHF35) for 20 min at room temperature, followed by the primary antibody at a dilution of 1:800 for S100, 1:8000 for 34BE12 and 1:12800 for HHF35 for 30 min at room temperature, except for the slides stained with 34BE12, which were incubated overnight at 4°C. This was followed by incubations with the secondary biotinylated antibody and the ABC complex. Rinses with PBS were carried out between the various steps of the staining procedure. The color was developed using diaminobenzidine as a chromogen. The slides were then counterstained lightly with hematoxylin and mounted with Permount. The negative control consisted of omitting the primary antibody in each case.
RESULTS Benign breast ducts and acini were surrounded by a layer of HHF35-, 34BE 12- and S I OO-positive cells. In cystically dilated ducts, this layer, however, was attenuated and sometimes discontinuous. In smaller breast ducts and acini, a subgroup of luminal cells were positive for 34BE12 and S100. In benign hyperplasias, a myoepithelial cell layer could be seen which was positive with all three antibodies. In some hyperplastic lesions,
Fig. f. In this ductal carcinoma (case 2) the in situ component contains 34HB12-cytokeratin-positive neoplastic cells intraductally. The ducts are surrounded by 34HB 12-positivemyoepithelial cells. ( x 130, immunoperoxidase stain). Fig. 2. The same case stained with SIOO. Positive intraductal cells can also be seen with this antibody. ( x 130, immunoperoxidase stain). Fig. 3. The same case stained with HHF35. The periductal myoepithelial cells can be clearly seen. However, some positively stained neoplastic appearing cells can also be seen intraductally. ( x 130, immunoperoxidase stain). Fig. 4. This is an example of a ductal infiltrating carcinoma (case 1) expressing widespread immunoreactivity for 34BE12-cytokeratin. ( x 130, immunoperoxidase stain).
902
CYTOARCHITECTURE OF BREAST LESIONS
SOINl & MIETTINEN
Fig. 5. Another case of infiltrating ductal carcinoma (case 6 ) revealing S100-positive tumor cells in the immunostaining. ( x 130, immunoperoxidase stain).
Fig. 6. The same case stained with HHF35. Positivity for muscle-specific actins can be observed in tumor cells. ( x 130, immunoperoxidase stain).
the myoepithelial cell layer was sometimes attenuated and slightly discontinuous. Scattered HHF35-, 34BE12- and S100-positive cells could be seen intraductally. In intraductal papillomas, a myoepithelial cell layer positive for all three antibodies was seen around the proliferating luminal epithelium. The vascular structures of the papillary cores were HHF35 positive. Scattered S100- and 34BE 12-positive luminal cells could be seen intraductally, while all the luminal cells appeared to be negative for HHF35. In the sclerosing adenoses, a rich network of 34BE12-, S loo-, and HHF35-positive cells could be seen around the proliferating ducts. In atypical hyperplasias, scattered cells reactive for all three antibodies could be observed intraductally. The myoepithelial cell layer was largely preserved in these lesions. 904
In in situ carcinomas, a HHF35-, 34BE12and S100-positive cell layer could still be observed around the ducts, though it was discontinuous and focally absent (Figs. 1-3). The discontinuity was evaluated in in situ carcinomas and infiltrating carcinomas with a separate in situ component, and was found to be 44% for 34BE12,62% for HHF35 and 11% for S100. The in situ carcinoma component often contained neoplastic cells which showed variable positivity for 34BE12 and SlOO (Figs. 1 & 2). In some cases also cells positive for HHF35 were observed (Fig. 3). With 34BE12 a majority of the carcinomas analyzed contained positive tumor cells (18/25 cases positive) (Figs. 1 & 4). A substantial number of carcinomas were also SlOO positive (14/ 25 cases) (Figs. 2 & 5). In seven carcinomas scattered HHF35-positive tumor cells could be
CYTOARCHITECTURE OF BREAST LESIONS
TABLE 1. The results of the immunohistochemical stainings in carcinomas Case Diagnosis Percentage of positive tumor number cells 34BE12 HHF35 SlOO 1. IFC + ISC 95%) 2. 10% IFC ISC 10% ISC 3. 20% 10% 4. ISC 1OYO 10% IFC + ISC 20% 5. 30% 6. IFC + ISC 6Oy0 10% 10% 7. ISC 5% IFC + ISC 8. IFC + ISC 9. 5% 10. ISC 30% IFC + ISC 5% 11. 10% IFC + ISC 12. 5'1/0 5% IFC ISC 13. 10% IFC + ISC 14. 5% 5Yn 15. IFC ISC 10% 5% ISC 16. 5% 5% 17. IFC ISC ISC 18. 50% 40% IFC + ISC 5% 19. 5% ISC 20. 21. 5Y" ISC ISC 22. 23. IFC + ISC 5% IFC + ISC 24. 25. IFC ISC 10% Positive 18 7 14 (total) Abbreviations: IFC = Infiltrating ductal carcinoma ISC = In situ carcinoma.
+
+ + +
+
seen (Figs. 3 & 6). Positive cells were found both intraductally and in infiltrating areas of the breast carcinomas studied (see Table 1). In invasive carcinomas there was no positivity around the tumor cell islands with 34BE12 or S100. With HHF35 positive fusiform cells were often observed around these islands though morphologically they represented myofibroblastic cells. They were also seen in the stroma of the infiltrating carcinomas. DISCUSSION In this study fifty-three benign and malignant breast lesions were analyzed immunohistochemically with antibodies to cytokeratins 1, 5 , 10, 14 (34BE12), muscle-specific actin (HHF35) and
SlOO protein. These antibodies have been used to evaluate the preservation of the myoepithelial cell component, which is used as a criterion in the evaluation of breast malignancy (3, 19). While our findings concerning the benign breast lesions are in line with those reported by others (1, 2, 11, 14, 16), in carcinomas some results were at variance with previous reports. In our material a majority of carcinomas expressed variable positivity for 34BE12. Some were also positive for SlOO and a few even for HHF35. The findings suggest that breast carcinomas are immunohistochemically heterogeneous. Based on the results it seems possible that some tumor cells have differentiated towards myoepithelial cells. In some earlier immunohistochemical reports, only a minority of breast carcinomas were found to be positive for medium and high molecular weight cytokeratins (1, 2, 11, 14, 16). With antibodies KA1 and KA4 recognizing cytokeratins 5 and 14, and cytokeratins 14, 15, 16 and 19, respectively, only a few invasive carcinomas were positive with KA1, while all were positive for KA4 (1 1, 14). In another study evaluating the distribution of keratin 19 immunoreactivity in benign and malignant breast lesions, carcinoma cells seemed to express keratin 19, while benign proliferative lesions also contained a substantial number of keratin 19-negative cells (2). Some other authors, however, have found more positivity for medium and high molecular weight cytokeratins in breast carcinomas (7). Thus the different results may partly be due to use of different antibodies specific for different cytokeratin subtypes and possibly different epitopes of the molecules. Whilst in a recent report only a minority of ductal in situ carcinomas were positive for 34BE12-cytokeratin antibody, the authors found positive intraductal cells in ten additional cases, which they interpreted to represent nonneoplastic cells (1 6). The epithelium of the terminal ductal lobular units is generally considered to be the site of origin of most mammary carcinomas (14). The epithelium of these units has been found to be immunoreactive for both high and low molecular weight cytokeratins (14, 17). Our results are thus in accordance with the presumed origin of breast carcinomas, reflecting their capability also to express high and medium molecular 905
SOINI & MIETTINEN
weight cytokeratins. Because of the existence of high and medium molecular weight cytokeratin reactivity in nonmyoepithelial cells in terminal ductal and lobular units (17), the expression of these cytokeratins need not necessarily indicate true myoepithelial cell differentiation in tumor cells. In a minority of breast carcinomas (both in situ and invasive), however, reactivity for muscle-specific actins could be seen. This reactivity was located in the neoplastic cells, indicating that some malignant cells may express true myoepithelial cell differentiation. This possibility is also suggested by other immunohistochemical investigations which reveal CALLA (9), GFAP (7) or vimentin (4, 7) immunoreactivity in a minority of breast carcinomas. In some ultrastructural studies myoepithelial cell differentiation has also been described (6). Finally, myoepithelial cell differentiation can be seen in SV40-transformed breast epithelial cell lines (1 8). A substantial number of tumors were positive for S100. SlOO positivity has been described in myoepithelial cells and may be considered, in a restricted sense, to indicate myoepithelial cell differentiation (15). However, in benign nonmyoepithelial ductal cells in breast tissues, reactivity for SlOO can also be observed ( 5 ) . SlOO protein reactivity has, in addition, been shown in various types of mesenchymal and epithelial tumors, including breast carcinomas (15 , 20). In benign ductal hyperplasias and in situ carcinomas we evaluated the continuity of the myoepithelial cell layer around the ducts. In benign lesions this ranged from 70 to 100% depending on the antibody used (data not shown). In in situ carcinomas it ranged from 0 to 100% with mean values of 44%, 62% and 11% for 34BE12, HHF35 and S100, respectively (data not shown). The partial loss of the myoepithelial cell layer in intraductal hyperplasias and cystically dilated glands is possibly due to the expanded forces generated by the dilated ducts. Although in in situ carcinomas this may also play a part, most of the effect is probably due to different collagenases and proteases produced by the neoplastic cells which disrupt the basement membrane on which the myoepithelial cells rest. Production of such collagenases by the tumor cells has been shown immunohistochemically, and immunoreactivity for collagenase IV in the pro906
liferating cells has even been suggested as a marker for malignant breast disease (1 3). Another point to be noted is the difference between the immunoreactivities with the different antibodies. Since myoepithelial cells are most certainly responsible for the staining around the ductal structures, the ability of the antibodies to recognize myoepithelial cells can be tested in this way. According to the results, HHF35 is the most sensitive of the antibodies, while SlOO seems to be the least sensitive. This also suggests that not all 34BE12 and SlOO positivity found in carcinomas necessarily indicates myoepithelial cell differentiation. In conclusion, the results show that none of the antibodies employed in this investigation can be used to distinguish between benign and malignant breast epithelial proliferations because immunoreactivities for all antibodies can be shown in in situ and invasive carcinomas in addition to benign lesions. We are grateful to Mr A1 Kovatich, the Immunopathology Laboratory, Thomas Jefferson University Hospital, for assistance in the performance of the immunostainings. This study was supported by a grant from the European Cancer Society (UICC).
REFERENCES 1. Altsmannsberger, M., Dirk, T , Droese, M., Weber, K. & Osborn, M.: Keratin polypeptide distribution in benign and malignant breast tumors: subdivision of ductal carcinomas using monoclonal antibodies. Virchows Arch. B. 51: 265-275, 1986. 2. Bartek, J., Taylor-Papadimitriou, J,, Miller, N. & Millis, R. :Patterns of expression of keratin 19 as detected with monoclonal antibodies in human breast tissues and tumors. Int. J. Cancer 36: 299-306, 1985. 3. Carter, D.:Interpretation of breast biopsies. Biopsy interpretation series. Raven Press, New York 1984, pp. 25-75. 4. Domagalu, W , Lasota, J., Dukowicz, A , , Markiewski, M., Striker, G., Weber, K, & Osborn, M.: Vimentin expression appears to be associated with poor prognosis in node-negative ductal NOS breast carcinomas. Am. J. Pathol. 137: 1299-1304, 1990. 5 . Gillett, C. E., Bobrow, L. G. & Millis, R. R.: SlOO protein in human mammary tissue-immunoreactivity in breast carcinoma, including Paget’s disease of the nipple, and value as a marker of myoepithelial cells. J. Pathol. 160: 19-24, 1990.
CYTOARCHITECTURE OF BREAST LESIONS
6. Gould, K E., Juo, W & Buttiforu, H.: Ultrastructural analysis in the differential diagnosis of breast tumors. The significance of myoepithelial cells, basal lamina, intracytoplasmic lumina and secretory granules. Pathol. Res. Pract. 167; 45-70, 1980. 7. Gould, !l E., Koukoulis, G. K., Junsson, D. S., Nugle, R. B., Frunke, W W & Moll, R.: Coexpression patterns of vimentin and glial filament protein with cytokeratins in the normal, hyperplastic, and neoplastic breast. Am. J. Pathol. 137: 1143-1 155, 1990. 8. Gown, A . M. & Vogel, A. M.: Monoclonal antibodies to human intermediate filament proteins. 111. Analysis of tumors. Am. J. Clin. Pathol. 84: 413424, 1985. 9. Gusterson, B. A , , Monughun, I! M., Muhendrun, R., Ellis, J. & O’Hure, M. J.: Identification of myoepithelial cells in human and rat breasts by anti-common acute lymphoblastic leukemia antigen antibody A12. J. Natl. Cancer Inst. 77: 343-349, 1986. 10. Hsu, S. M., Ruine. L. & Funger, H.: Use of avidinbiotin-peroxidase complex (ABC) in immunoperoxidase techniques: a comparison between ABC and unlabelled antibody (PAP) procedures. J. Histochem. Cytochem. 29: 577-580, 1981. I 1. Jurusch, E.-D., Nugle, R. B., Kuufmann, M., Maurer, C. & Bocker, W J.: Differential diagnosis of benign epithelial proliferations and carcinomas of the breast using antibodies to cytokeratins. Hum. Pathol. 19: 276-289, 1988. 12. Moll, R., Frunke, W W , Schiller, D. L., Geiger, B. & Krepler, R.: The catalog of human cytokeratins: patterns of expression in normal epithelia, tumors and cultured cells. Cell 31: 11-24, 1982. 13. Monteugudo, C., Merino, M . J., Sun-Juan, J., Liottu, L. A . & Stetler-Stevenson, W G.:Immuno-
histochemical distribution of type IV collagenase in normal, benign and malignant breast tissue. Am. J. Pathol. 136: 585-592, 1990. 14. Nugle, R. B., Bocker, W, Davis, J. R., Heid, H. W , Kuufmunn, M., Lucus, D. & Jarusch, E.-D.: Characterization of breast carcinomas by two monoclonal antibodies distinguishing myoepithelial from luminal epithelial cells. J. Histochem. Cytochem. 34: 869-881, 1986. 15. Nukujimu, T , Watunube, S., Suto, II, Kumeyu, T , Hirotu, T & Shimosuto, II; An immunoperoxidase study of S-100 protein distribution in normal and neoplastic tissues. Am. J. Surg. Pathol. 6: 715-727, 1982. 16. Ruju, U., Crissmun, J. D., Zurbo, R. J. & Gottlieb, C.: Epitheliosis of the breast. An immunohistochemical characterization and comparison to malignant intraductal proliferations of the breast. Am. J. Surg. Pathol. 14: 939-947, 1990. 17. Rudlund, El S. & Hughes, C. H.: Immunocytochemical identification of cell types in human mammary gland: variations in cellular markers are dependent on glandular topography and differentiation. J. Histochem. Cytochem. 37: 1087-1 100, 1989. 18. Rudlund, l? S., Ollerheud, G.& Burruclough. R.: Isolation of simian virus 40-transformed human mammary epithelial stem cell lines that can differentiate to myoepithelial-like cells in culture and in vivo. Develop. Biol. 136: 167-180, 1989. 19. Sloune, J. El; Biopsy pathology of the breast. First Edition. Chapman & Hall, London 1985, pp. 93-123. 20. Stroup, R. M. & Pinkus, G. S.: S-100 immunoreactivity in primary and metastatic carcinoma of the breast: a potential source of error in immunodiagnosis. Hum. Pathol. 19: 949-953, 1988.
907
