- Email: [email protected]
Actin versus vimentin in myoepithelial cells of salivary gland tumors
Actin versusvimentin in myoepithelial cells of salivary gland tumors A comparative study Vera Cavalcanti de Araujo,a Yasmin Rodarte Carvalho,b and Ney Soares de Araujo,a SHoPaulo, Brazil DEPARTMENT
OF ORAL PATHOLOGY
SCHOOL OF DENTISTRY,
UNIVERSITY
OF S%O PAUL0
Vimentin versus actin expression was immunohistochemically studied in myoepithelial cells of 24 salivary gland tumors in which the participation of myoepithelial cells as a tumoral component has been postulated: two basal cell adenomas, seven pleomorphic adenomas, two myoepitheliomas, seven adenoid cystic carcinomas (two tubular, four cribriform, one solid), six polimorphous low-grade adenocarcinomas. lmmunostaining was carried out in formalin-fixed tissue serial sections (3 pm) by the avidin-biotin method, using the antibody vimentin (Dako Corp., Carpenteria, Calif.) and the antibody HHF35 anti-muscle actin (Enzo Biochemical, N.Y.). Our results have confirmed positive staining for vimentin in all salivary tumors studied, although in some tumors it was only in focal areas. The staining for the HHF35 antibody to muscle actin was only consistently found in the adenoid cystic carcinomas of the tubular and cribriform patterns. This study suggests that actin is at least somewhat replaced by vimentin in neoplastic tumoral cells. Therefore vimentin can be used to define the participation and distribution of myoepithelial cells in these tumors. (ORAL SURCORAL MED ORAL PATHOL 1994;77:387-91)
Immunohistochemical stainings for myoepithelial tumoral markers have included Sloe, myosin, glial fibrillary acidic protein, desmin, and keratin. l-t3 However, none of these has proved reliable for identifying myoepithelial cell-lines. It has been well demonstrated that muscle-specific actin is present in the normal myoepithelial cells of salivary glands.14-l7On the basis of this observation, the expression of muscle actin in myoepithelial tumoral cells has become a standard diagnostic procedure.‘* However, in our laboratory, we have found vimentin to be one of the earliest indicators of neoplastic myoepithelial differentiation.19 The purpose of the present study, therefore, was to compare the staining patterns of tumoral myoepithelial cells with the use of antibodies to vimentin and to actin.
Formalin-fixed tissue serial sections (3 pm) from the tumors were usedfor immunohistochemical staining. Immunostaining was carried out with the avidinbiotin method with the use of the antibody antivimentin (Biogenex Laboratories, San Ramon, Calif.) and the antibody HHF35 antimuscle actin (Enzo Biochemical, Farmingdale, N.Y.). The antibodies were applied at an appropriate dilution overnight at 4” C. Diaminobenzidine was used as a chromogen. Negative controls were incubated in a buffer without primary antibodies. Blood vessels and normal tissue included in the specimens served as internal controls for the immunohistochemical reactions. Normal major and minor salivary glands included in the specimens were also studied to verify the staining of actin in normal myoepithelial cells.
MATERIAL AND METHODS
RESULTS
Twenty-four gland tumors were retrieved from the files of the Oral Pathology Department of the University of Sgo Paulo, School of Dentistry. They consisted of two monomorphic adenomas,sevenpleomorphic adenomas, two myoepitheliomas, seven adenoid cystic carcinomas (two tubular, four cribriform, one solid) and six polimorphous low-grade adenocarcinomas. All histologic material from the tumors was reviewed to confirm the accuracy of the diagnoses.
Results that concern the tumors are presented in Table I. Normal tissues included in the specimens provided a useful internal control for both actin and vimentin. Vimentin was demonstrated in stromal fibroblasts and endothelial cells of the vasculature. Actin was demonstrated in smooth muscle cells of the vascular system and in normal myoepithelial cells around acinar cells and intercalated ducts when normal salivary gland were included in the slicesof either major or minor salivary glands (Figs. 1 and 2). As confirmed by these internal controls, all formalinfixed specimens were well preserved with respect to the antigens. Nevertheless, areas of some tumors excluding the stroma were totally negative.
aProfessor and Chairman. bAssistant Professor. Copyright @ 1994 by Mosby-Year Book, Inc. 0030-4220/94/%1 .OO+ .lO T/l4152598
387
388
de Araujo, Carvalho, and de Araujo
Table
1.Immunoreactivity of salivary gland tumors
ORAL SURGERY ORAL MEDICINE ORAL PATHOLOGY April 1994
HHF 35 ACTIh +-Tumor
Number
++
Basal cell adenoma Pleomorphic adenoma Myoepithelioma Adenoid cystic carcinoma Tubular Cribriform Solid Polimorphous low-grade adenocarcinoma
2 I 2
2
I 2
2
2 1
I
1
2 4 1 6
)
+
Vimentin 7----..+ J\‘rgutivr -~
-. . ..__---Xegative
++
I 3 2
I ? 2
I
I 2
I 2
I
5
1 I 4
1
Immunostaining was graded as ++ strong, + weak to moderate, and negative
Basal cell adenoma
Vimentin was present in a great number of nonluminal cells, even in those organized into duct-like structures. In some casesa cell that lined a luminal space was also vimentin positive. Tumor cells were negative for actin (Fig. 3). Pleomorphic
adenoma
Vimentin was detected in almost all the nonluminal cells (polygonal, spindle-shaped, and hyaline or plasmacytoid). Staining was weak in the clear polygonal cells and strong in the plasmacytoid cells. Actin stained only the outer cells of the ducts and faintly stained the polygonal cells (Figs. 4 and 5). Myoepithelioma
Vimentin was found in almost all tumor cells (hyaline or plasmacytoid and spindle-shaped cells). Tumor cells were negative for actin. Adenoid cystic carcinoma Tubular type. Vimentin was found in the narrow
band of the cytoplasm of the outer tubular cells. The staining for actin followed the samepattern (Fig. 6). Cribriform type. Cells that lined pseudocystsand peripheral cells of cylinders stained positively for vimentin and for actin (Fig. 7). Solid type. Neither vimentin or actin was detected in the tumor cells. Polymorphous
low-grade adenocarcinoma
Most of the cells in the lobular areas and the cells that lined the papillae were positive for vimentin. In areas of trabecular pattern, some cells were positive for vimentin. In two tumors the same pattern of staining was seenfor actin, although the staining was weaker than for vimentin. In four of the six tumors studied all the cells were negative (Figs. 8 and 9). DlSCUSSlON
The present results have confirmed positive staining for vimentin in salivary gland tumors in which myo-
epitelial cells have been postulated as a tumoral component.19Vimentin has already been found in pleomorphic adenomas,2,3,20-22 myoepitheliomas,23-25 monomorphic adenomas,’2 and adenoid cystic carcinomas.26,21Vimentin has also been detected in epithelial-myoepithelial carcinomas.19 In some specimens the whole tumor exhibited vimentin, but in others only focal areas of the tumor were positive. The strongest reactions occurred in the peripheral portion. The most obvious staining for vimentin was seen in modified myoepithelial cells-hyaline or plasmacytoid-of pleomorphic adenomasand myoepithelioma. We did not find vimentin in tumors thought to arise from excretory ducts28,29or in those with no evidence of the presenceof myoepithelial cells, such as the cystadenoma, papillary cystadenoma lymphomatosum, inverted ductal papilloma, mucoepidermoid carcinoma, salivary ductal carcinoma, and acinic cell carcinoma.30,31 These findings support the view that staining for vimentin is one of the earliest indicators of neoplastic myoepithelial cell differentiation. Our results have also shown that staining with the HHF35 antibody to muscle actin was only consistently found in the tumoral cells of adenoid cystic carcinomas of tubular and cribriform patterns. In adenoid cystic carcinomas that are composed of solid masseswith central necrosis, the staining was negative for both actin and vimentin. This fact was also detected by Hirano et al.18and merits further investigation. In other tumors studied, staining with actin was either extremely weak or absent. Our experience with actin is in keeping with that of Dardick et al.17 who have described the lack of immunoreactivity in most cells in pleomorphic adenomas and myoepitheliomas. This observation has been attributed to the effects of formalin fixation. On the other hand, Tsukada et a1.14y 32demonstrated that actin epitope recognized by the HHF35 antibody was stable to formalin, whereas someother authors17,33have claimed to achieve an improvement in staining with the use of material fixed in a modified methacarn fixative.
ORAL SURGERY ORAL MEDICINE ORAL PATHOLOGY Volume 77, Number 4
de Araujo,
Carvalho, and de Araujo
389
Fig. 1. Normal parotid gland. Positive to HHF35 in the myoepithelial cells of the acini. (Original magnification x400.) Fig. 2. Normal minor salivary gland. Positive to HHF35 in the myoepithelial cells of the acini, Positivestained blood vesselservesas an internal control. (Original magnification X300.) Fig. 3. Basal cell adenoma. A, Staining of the basal cells with vimentin. B, same cells are negative with actin. (Original magnification x550.) Fig. 4. Pleomorphic adenoma. A, Vimentin stains the majority of polygonal tumoral cells. B, polygonal cellsare weakly stained with actin. (Original magnification X200.) Fig. 5. Pleomorphic adenoma.A, Vimentin stains the majority of the plamocytoid cells. B, Plasmacytoidcells are negative. (Original magnification X550.) However, even when methanol-fixed samples were used, actin was detected in only a few of the pleomorphic adenomas and in none of the myoepitheliomas studied.t6* l7 In addition, our results have shown that the staining for actin of normal myoepithelial cells that are present in the major and minor normal salivary glands included in the specimens was not affected by formalin fixation. As already demonstrated by ultrastructural studies, myoepithelial tumoral cells usually lack many of the features that characterize normal myoepithelial
cells, that is, myofilaments, micropinocytotic vesicles, hemidesmosomes, and basement membranes.34 It is clear that a full differentiation into myoepithelial cells is seldom achieved in tumors.17 Probably the state of differentiation, including the appearance of actin, matches the stage of development at which the oncogenie/carcinogenic stimulus acts.35 Moreover, it is already known that the filament expression is not a rigidly invariable cell function and is not infrequently subject to change under a variety of circumstances including neoplastic transformation.36
3%)
de Araujo, Carvalho, and de Araujo
ORAL SURGERY ORAL MEDICINE ORAL PATHOLOGY April 1994
Fig. 6. Adenoid cystic carcinoma, tubular. A, Staining of the outer cells of the tubular structures with vimentin. B, Same cells are weakly stained with HHF35. (Original magnification x200.) Fig. 7. A and B, Ad-
enoid cystic carcinoma, cribriform types. Staining with vimentin and strong staining with actin is seenin the cells lining the pseudocystsand in the peripheral cells of the cylinders. (Original magnification x200.) Fig. 8. Polymorphous low-grade adenocarcinoma. A, Vimentin stains the majority of tumoral cells. B, Actin was not detected in this tumor. (Original magnification x550.) Fig. 9. Polymorphous low-grade adenocarcinoma.A, Staining with vimentin in a papillary area. B, Actin was not detected in this tumor. (Original magnification x550.)
It is well known that normal salivary myoepithelial cells do not express vimentin except transiently in coexpression with cytokeratin in a few fetal cells during the development- of the parotid gland.37 Therefore these results strongly suggest that actin is partially or totally replaced by vimentin in neoplastic myoepithelial tumoral cells. This finding is not unique and was
recently demonstrated in neoplastic renal cortical tubules by Ward et a1.38 Finally, we suggestthe useof vimentin to define and
to investigate the participation and distribution myoepithelial cells in salivary gland tumors.
of
REFERENCES
1. Caseiitz J, Loning J, Staquet MJ, Seifert G, Thivoleth J. Immunohistochemical demonstration of filaments structures in parotid gland: occurrence of keratin and actin in normal and tumoral gland with special respect to the myoepitheliak cells. J Cancer Res Clin 011~011981;100:59-68. 2. Caselitz J, O&cm M, Seifert G, Weber K. Intermediate-sized filament proteins (prekeratin, vimentin, desmin) in the normal
ORAL SURGERY ORAL MEDICINE ORAL PATHOLOGY Volume 77, Number 4 parotid gland and parotid gland tumors. Virchows Arch A Path01 Anat Histopathol 1981;393:273-86. 3. Caselitz J, Osborn M, Weber K. Intermediate-size filament protein (prekeratin, vimentin, desmin) in normal parotid gland and parotid gland tumors: search for specific markers of neoplastic epithelial duct and myopithelial cell lines established from human salivary gland and characterization of their growth in vitro. Cancer 1981;54:2959-67. 4. Nathrath WBJ, Wilson PD, Trejdosiewecz LK. Immunohistochemical localization of keratin and luminal epithelial antigen in myoepithelial and luminal epithelial cells of mammary and salivary gland tumours. Path01Res Pratt 1982;175:27988. 5. Sciubba JJ, Brannon RB. Myoepitheliom of salivary glands: report of 23 cases.Cancer 1982;49:562-72. 6. Hara K, Tlo M, Takeuchi J, Tijima S, Endo T, Hidaka H. Distribution of S-100 protein in normal salivary glands and salivary gland tumors. Virchows Arch A Path01Anat Histopath01 1983;401:237-49. I. Sato M, Hayashi Y, Yoshida H, Yanagawa T, Yura Y, Nitta T. Search for specific markers of neoplastic epithelial duct and myoepithelial cell lines established from human salivary gland and characterization of their growth in vitro. Cancer 1984; 54~2959-67. 8. Palmer RM, Lucas R. Immunocytochemical identification of cell types in pleomorphic adenoma, with particular reference to myoepithelial cells. J Path01 1985;146:213-20. 9. Nakasato Y, Yshida Y, Takahashi K, Susuki K. Immunohistochemical distribution of Siau protein and glial fibrilary acidic protein in normal and neoplasic salivary glands. Virchows Arch A Path01 Anat Histopathol 1985;405:299-310. 10. Palmer ER. The identification of myoepithelial cells in human salivary glands: a review and comparison of light microscopical methods. J Oral Path01 1986;15:221-9. 11. Mori M, Tsukitani K, Ninomiya T, Okada Y. Various expressions of modified myoepithelial cells in salivary pleomorphic adenoma: immunohistochemical studies. Path Res Pratt 1987; 182:632-46. 12. Morinaga S. Nakajima T, Shimosato Y. Normal and neoplastic myoepitelial cells in salivary glands: an immunohistochemical studv. Hum Path01 1987:18:1218-26. 13. GustafssbnH, Virtanen I, Thornell L-E. Glial fibrillary acidic protein and desmin in salivary neoplasms: expression of four different types of intermediated filament proteins within the same cell type. Virchows Arch B Cell Path01 1989;57:303-13. 14. Tsukada T, McNutt MA, Ross R, Gown AM. HHF35, a muscle actin-specific monoclonal antibody. Am J Path01 1987;127:389-402. 15. Gugliotta P, Sapino A, Macri L, Skalli 0, Gabbiani G. Specific demonstration of myoepithelial cells by anti-alpha smooth muscle actin antibody. J Histochem Cytochem 1988;36:65963. 16. Zarbo RJ, Bacchi CE, Gown AM, Muscle-specific protein expressionin normal salivary glands and pleomorphic adenomas; an immunocytochemical study with biochemical confirmation. Mod Path01 1991;4:621-6. 17. Dardick I, Ostrynski VL, Ekem JK, Leung R, Burford-Mason AP. Immunohistochemical and ultrastructural correlates of muscle-actin expression in pleomorphic adenomas and myoepitheliomas based on comparison of formalin and methanol fixation. Virchows Archiv A Path01 Anat Histouathol 1992; 421:95-104. 18. Hirano T, Gluckman JL, deVries EJ. The expressionof vascular smooth-muscle actin in salivary gland tumors. Arch Otolaryngol Head Neck Surg 1990;116:692-6. 19. Araujo VC, Araujo NS. Vimentin as a marker of myoepithelial cells in salivary gland tumors. Eur Arch Otorhinolaryngol 1990;247:252-5. 20. Erlandson RA, Cardon-Cardo C, Higgins PJ. Histogenesis of
de Araujo,
Carvalho, and de Araujo
391
benign pleomorphic adenoma (mixed tumor) of the major salivary glands: an ultrastructural and immunohistochemical studv. Am J Sure Path01 1984:8:803-20. 21. Burns BF, Dar&k I, Parks’ WR. Intermediate filament expression in normal parotid glands and pleomorphic adenomas. Virchows Arch A Path01 Anat Histopathol 1988; 413:103-12. 22. Yamada K, Shinohara H, Takai Y, Mori M. Monoclonal antibody-detected vimentin distribution in pleomorphic adenoma of salivary glands. J Oral Path01 1988;17:348-55. 23. Dardick I, Cave11S, Boivin M, et al. Salivary gland myoepithelial variants histological, ultrastructural, and immunocytochemical features. Virchows Arch A Path01 Anat Histopath01 1989;416:25-42. 24. Dardick I, Thomas MJ, van Nostrand AWP. Myoepithelioma, new concepts of histology and classification: a light and electron microscopy study. Ultrastruct Path01 1989;13:187-224. 25. Araujo VC, Souza SCOM, Souza LB, Araujo NS. Estudo histopatologico e imunohistoquimico do mioepitelioma de elbndula salivar. Rev Odont USP 1990:4:62-6. 26. Caselitz J, Schulze I, Seifert G. Adenoid cystic carcinoma of the salivary glands: an immunohistochemical study. J Oral Path01 1986;15:308-18. 21. Chomete G, Auriol M, Vaillant JM, Kasai T, Okada Y, Mori M. Heterogeneity and co-expression of intermediate filament proteins in adenoid cystic carcinoma of salivary glands. Path01Bio (Paris) 1991;39:110-6. 28. Eversole LR. Histogenetic classification of salivary tumors. Arch Path01 1971;92:433-43. 29. Regezi JA, Batsakis JG. Histogenesis of salivary gland neoplasm. Otolaryngol Clin North-Am 1977;10:297-307. 30. Arauio VC. Souza SOM. Nunes FD. Arauio NS. Immunohistochemical analysis of 89 benign salivary gland tumors. Rev Odont USP 1991;5:1-6. 31. Araujo VC, Souza SOM, Jaeger RG, Araujo NS. Immunohistochemical study of malignant salivary gland tumors: an analysis of 67 cases.Rev Odont USP 1991;5:37-42. 32. Tsukada T, Tippens DBA, Gordon D, Ross R, Gown AM. HHF35: a muscle-actin-specific monoclonal antibody. Am J Path01 1987;126:51-60. 33. Leoncini P, Cintorino M, Vindigni C, et al. Distribution of cytoskeletal and contractile proteins in normal and tumor-bearing salivary and lacrimal glands. Virchows Archiv A Path01 Anat Histopathol 1988;412:329-37. 34. Tandler B, Denning CR, Mandel ID, Kutscher AH. Ultrastructure of human labial salivary glands, III: myoepithelium and ducts. J Morph01 1970;130:227-46. 35. Batsakis JG. Salivary gland neoplasia: an outcome of modified morphogenesis and cyto differentiation. ORAL SURG ORAL MED ORAL PATHOL 1980;49:229-32.
36. Gould VE. The coexpressionof distinct classesof intermediate filaments in human neoplasms.Arch Path01 Lab Med 1985; 109:984-5. 37. GustafssonH, Kjorell U, Eriksson A, Virtanen I, Thornell LE. Distribution of intermediate filament proteins in developing and adult salivary glands in man. Anat Embryo1 1988;178:34351. 38. Ward JM, Stevens JL, Konishi H, et al. Vimentin metaplasia in renal cortical tubules of preneoplastic, neoplastic, aging, and regenerative lesions of rats and humans. Am J Path01 1992; 1411955-64. Reprint requests:
V. C. de Araujo Disciplina de Patologia Bucal Facuidade de Odontologia Universidade de Sgo Paulo C. P. 8216, Sao Paulo, S. P. Brasil
OF ORAL PATHOLOGY
SCHOOL OF DENTISTRY,
UNIVERSITY
OF S%O PAUL0
Vimentin versus actin expression was immunohistochemically studied in myoepithelial cells of 24 salivary gland tumors in which the participation of myoepithelial cells as a tumoral component has been postulated: two basal cell adenomas, seven pleomorphic adenomas, two myoepitheliomas, seven adenoid cystic carcinomas (two tubular, four cribriform, one solid), six polimorphous low-grade adenocarcinomas. lmmunostaining was carried out in formalin-fixed tissue serial sections (3 pm) by the avidin-biotin method, using the antibody vimentin (Dako Corp., Carpenteria, Calif.) and the antibody HHF35 anti-muscle actin (Enzo Biochemical, N.Y.). Our results have confirmed positive staining for vimentin in all salivary tumors studied, although in some tumors it was only in focal areas. The staining for the HHF35 antibody to muscle actin was only consistently found in the adenoid cystic carcinomas of the tubular and cribriform patterns. This study suggests that actin is at least somewhat replaced by vimentin in neoplastic tumoral cells. Therefore vimentin can be used to define the participation and distribution of myoepithelial cells in these tumors. (ORAL SURCORAL MED ORAL PATHOL 1994;77:387-91)
Immunohistochemical stainings for myoepithelial tumoral markers have included Sloe, myosin, glial fibrillary acidic protein, desmin, and keratin. l-t3 However, none of these has proved reliable for identifying myoepithelial cell-lines. It has been well demonstrated that muscle-specific actin is present in the normal myoepithelial cells of salivary glands.14-l7On the basis of this observation, the expression of muscle actin in myoepithelial tumoral cells has become a standard diagnostic procedure.‘* However, in our laboratory, we have found vimentin to be one of the earliest indicators of neoplastic myoepithelial differentiation.19 The purpose of the present study, therefore, was to compare the staining patterns of tumoral myoepithelial cells with the use of antibodies to vimentin and to actin.
Formalin-fixed tissue serial sections (3 pm) from the tumors were usedfor immunohistochemical staining. Immunostaining was carried out with the avidinbiotin method with the use of the antibody antivimentin (Biogenex Laboratories, San Ramon, Calif.) and the antibody HHF35 antimuscle actin (Enzo Biochemical, Farmingdale, N.Y.). The antibodies were applied at an appropriate dilution overnight at 4” C. Diaminobenzidine was used as a chromogen. Negative controls were incubated in a buffer without primary antibodies. Blood vessels and normal tissue included in the specimens served as internal controls for the immunohistochemical reactions. Normal major and minor salivary glands included in the specimens were also studied to verify the staining of actin in normal myoepithelial cells.
MATERIAL AND METHODS
RESULTS
Twenty-four gland tumors were retrieved from the files of the Oral Pathology Department of the University of Sgo Paulo, School of Dentistry. They consisted of two monomorphic adenomas,sevenpleomorphic adenomas, two myoepitheliomas, seven adenoid cystic carcinomas (two tubular, four cribriform, one solid) and six polimorphous low-grade adenocarcinomas. All histologic material from the tumors was reviewed to confirm the accuracy of the diagnoses.
Results that concern the tumors are presented in Table I. Normal tissues included in the specimens provided a useful internal control for both actin and vimentin. Vimentin was demonstrated in stromal fibroblasts and endothelial cells of the vasculature. Actin was demonstrated in smooth muscle cells of the vascular system and in normal myoepithelial cells around acinar cells and intercalated ducts when normal salivary gland were included in the slicesof either major or minor salivary glands (Figs. 1 and 2). As confirmed by these internal controls, all formalinfixed specimens were well preserved with respect to the antigens. Nevertheless, areas of some tumors excluding the stroma were totally negative.
aProfessor and Chairman. bAssistant Professor. Copyright @ 1994 by Mosby-Year Book, Inc. 0030-4220/94/%1 .OO+ .lO T/l4152598
387
388
de Araujo, Carvalho, and de Araujo
Table
1.Immunoreactivity of salivary gland tumors
ORAL SURGERY ORAL MEDICINE ORAL PATHOLOGY April 1994
HHF 35 ACTIh +-Tumor
Number
++
Basal cell adenoma Pleomorphic adenoma Myoepithelioma Adenoid cystic carcinoma Tubular Cribriform Solid Polimorphous low-grade adenocarcinoma
2 I 2
2
I 2
2
2 1
I
1
2 4 1 6
)
+
Vimentin 7----..+ J\‘rgutivr -~
-. . ..__---Xegative
++
I 3 2
I ? 2
I
I 2
I 2
I
5
1 I 4
1
Immunostaining was graded as ++ strong, + weak to moderate, and negative
Basal cell adenoma
Vimentin was present in a great number of nonluminal cells, even in those organized into duct-like structures. In some casesa cell that lined a luminal space was also vimentin positive. Tumor cells were negative for actin (Fig. 3). Pleomorphic
adenoma
Vimentin was detected in almost all the nonluminal cells (polygonal, spindle-shaped, and hyaline or plasmacytoid). Staining was weak in the clear polygonal cells and strong in the plasmacytoid cells. Actin stained only the outer cells of the ducts and faintly stained the polygonal cells (Figs. 4 and 5). Myoepithelioma
Vimentin was found in almost all tumor cells (hyaline or plasmacytoid and spindle-shaped cells). Tumor cells were negative for actin. Adenoid cystic carcinoma Tubular type. Vimentin was found in the narrow
band of the cytoplasm of the outer tubular cells. The staining for actin followed the samepattern (Fig. 6). Cribriform type. Cells that lined pseudocystsand peripheral cells of cylinders stained positively for vimentin and for actin (Fig. 7). Solid type. Neither vimentin or actin was detected in the tumor cells. Polymorphous
low-grade adenocarcinoma
Most of the cells in the lobular areas and the cells that lined the papillae were positive for vimentin. In areas of trabecular pattern, some cells were positive for vimentin. In two tumors the same pattern of staining was seenfor actin, although the staining was weaker than for vimentin. In four of the six tumors studied all the cells were negative (Figs. 8 and 9). DlSCUSSlON
The present results have confirmed positive staining for vimentin in salivary gland tumors in which myo-
epitelial cells have been postulated as a tumoral component.19Vimentin has already been found in pleomorphic adenomas,2,3,20-22 myoepitheliomas,23-25 monomorphic adenomas,’2 and adenoid cystic carcinomas.26,21Vimentin has also been detected in epithelial-myoepithelial carcinomas.19 In some specimens the whole tumor exhibited vimentin, but in others only focal areas of the tumor were positive. The strongest reactions occurred in the peripheral portion. The most obvious staining for vimentin was seen in modified myoepithelial cells-hyaline or plasmacytoid-of pleomorphic adenomasand myoepithelioma. We did not find vimentin in tumors thought to arise from excretory ducts28,29or in those with no evidence of the presenceof myoepithelial cells, such as the cystadenoma, papillary cystadenoma lymphomatosum, inverted ductal papilloma, mucoepidermoid carcinoma, salivary ductal carcinoma, and acinic cell carcinoma.30,31 These findings support the view that staining for vimentin is one of the earliest indicators of neoplastic myoepithelial cell differentiation. Our results have also shown that staining with the HHF35 antibody to muscle actin was only consistently found in the tumoral cells of adenoid cystic carcinomas of tubular and cribriform patterns. In adenoid cystic carcinomas that are composed of solid masseswith central necrosis, the staining was negative for both actin and vimentin. This fact was also detected by Hirano et al.18and merits further investigation. In other tumors studied, staining with actin was either extremely weak or absent. Our experience with actin is in keeping with that of Dardick et al.17 who have described the lack of immunoreactivity in most cells in pleomorphic adenomas and myoepitheliomas. This observation has been attributed to the effects of formalin fixation. On the other hand, Tsukada et a1.14y 32demonstrated that actin epitope recognized by the HHF35 antibody was stable to formalin, whereas someother authors17,33have claimed to achieve an improvement in staining with the use of material fixed in a modified methacarn fixative.
ORAL SURGERY ORAL MEDICINE ORAL PATHOLOGY Volume 77, Number 4
de Araujo,
Carvalho, and de Araujo
389
Fig. 1. Normal parotid gland. Positive to HHF35 in the myoepithelial cells of the acini. (Original magnification x400.) Fig. 2. Normal minor salivary gland. Positive to HHF35 in the myoepithelial cells of the acini, Positivestained blood vesselservesas an internal control. (Original magnification X300.) Fig. 3. Basal cell adenoma. A, Staining of the basal cells with vimentin. B, same cells are negative with actin. (Original magnification x550.) Fig. 4. Pleomorphic adenoma. A, Vimentin stains the majority of polygonal tumoral cells. B, polygonal cellsare weakly stained with actin. (Original magnification X200.) Fig. 5. Pleomorphic adenoma.A, Vimentin stains the majority of the plamocytoid cells. B, Plasmacytoidcells are negative. (Original magnification X550.) However, even when methanol-fixed samples were used, actin was detected in only a few of the pleomorphic adenomas and in none of the myoepitheliomas studied.t6* l7 In addition, our results have shown that the staining for actin of normal myoepithelial cells that are present in the major and minor normal salivary glands included in the specimens was not affected by formalin fixation. As already demonstrated by ultrastructural studies, myoepithelial tumoral cells usually lack many of the features that characterize normal myoepithelial
cells, that is, myofilaments, micropinocytotic vesicles, hemidesmosomes, and basement membranes.34 It is clear that a full differentiation into myoepithelial cells is seldom achieved in tumors.17 Probably the state of differentiation, including the appearance of actin, matches the stage of development at which the oncogenie/carcinogenic stimulus acts.35 Moreover, it is already known that the filament expression is not a rigidly invariable cell function and is not infrequently subject to change under a variety of circumstances including neoplastic transformation.36
3%)
de Araujo, Carvalho, and de Araujo
ORAL SURGERY ORAL MEDICINE ORAL PATHOLOGY April 1994
Fig. 6. Adenoid cystic carcinoma, tubular. A, Staining of the outer cells of the tubular structures with vimentin. B, Same cells are weakly stained with HHF35. (Original magnification x200.) Fig. 7. A and B, Ad-
enoid cystic carcinoma, cribriform types. Staining with vimentin and strong staining with actin is seenin the cells lining the pseudocystsand in the peripheral cells of the cylinders. (Original magnification x200.) Fig. 8. Polymorphous low-grade adenocarcinoma. A, Vimentin stains the majority of tumoral cells. B, Actin was not detected in this tumor. (Original magnification x550.) Fig. 9. Polymorphous low-grade adenocarcinoma.A, Staining with vimentin in a papillary area. B, Actin was not detected in this tumor. (Original magnification x550.)
It is well known that normal salivary myoepithelial cells do not express vimentin except transiently in coexpression with cytokeratin in a few fetal cells during the development- of the parotid gland.37 Therefore these results strongly suggest that actin is partially or totally replaced by vimentin in neoplastic myoepithelial tumoral cells. This finding is not unique and was
recently demonstrated in neoplastic renal cortical tubules by Ward et a1.38 Finally, we suggestthe useof vimentin to define and
to investigate the participation and distribution myoepithelial cells in salivary gland tumors.
of
REFERENCES
1. Caseiitz J, Loning J, Staquet MJ, Seifert G, Thivoleth J. Immunohistochemical demonstration of filaments structures in parotid gland: occurrence of keratin and actin in normal and tumoral gland with special respect to the myoepitheliak cells. J Cancer Res Clin 011~011981;100:59-68. 2. Caselitz J, O&cm M, Seifert G, Weber K. Intermediate-sized filament proteins (prekeratin, vimentin, desmin) in the normal
ORAL SURGERY ORAL MEDICINE ORAL PATHOLOGY Volume 77, Number 4 parotid gland and parotid gland tumors. Virchows Arch A Path01 Anat Histopathol 1981;393:273-86. 3. Caselitz J, Osborn M, Weber K. Intermediate-size filament protein (prekeratin, vimentin, desmin) in normal parotid gland and parotid gland tumors: search for specific markers of neoplastic epithelial duct and myopithelial cell lines established from human salivary gland and characterization of their growth in vitro. Cancer 1981;54:2959-67. 4. Nathrath WBJ, Wilson PD, Trejdosiewecz LK. Immunohistochemical localization of keratin and luminal epithelial antigen in myoepithelial and luminal epithelial cells of mammary and salivary gland tumours. Path01Res Pratt 1982;175:27988. 5. Sciubba JJ, Brannon RB. Myoepitheliom of salivary glands: report of 23 cases.Cancer 1982;49:562-72. 6. Hara K, Tlo M, Takeuchi J, Tijima S, Endo T, Hidaka H. Distribution of S-100 protein in normal salivary glands and salivary gland tumors. Virchows Arch A Path01Anat Histopath01 1983;401:237-49. I. Sato M, Hayashi Y, Yoshida H, Yanagawa T, Yura Y, Nitta T. Search for specific markers of neoplastic epithelial duct and myoepithelial cell lines established from human salivary gland and characterization of their growth in vitro. Cancer 1984; 54~2959-67. 8. Palmer RM, Lucas R. Immunocytochemical identification of cell types in pleomorphic adenoma, with particular reference to myoepithelial cells. J Path01 1985;146:213-20. 9. Nakasato Y, Yshida Y, Takahashi K, Susuki K. Immunohistochemical distribution of Siau protein and glial fibrilary acidic protein in normal and neoplasic salivary glands. Virchows Arch A Path01 Anat Histopathol 1985;405:299-310. 10. Palmer ER. The identification of myoepithelial cells in human salivary glands: a review and comparison of light microscopical methods. J Oral Path01 1986;15:221-9. 11. Mori M, Tsukitani K, Ninomiya T, Okada Y. Various expressions of modified myoepithelial cells in salivary pleomorphic adenoma: immunohistochemical studies. Path Res Pratt 1987; 182:632-46. 12. Morinaga S. Nakajima T, Shimosato Y. Normal and neoplastic myoepitelial cells in salivary glands: an immunohistochemical studv. Hum Path01 1987:18:1218-26. 13. GustafssbnH, Virtanen I, Thornell L-E. Glial fibrillary acidic protein and desmin in salivary neoplasms: expression of four different types of intermediated filament proteins within the same cell type. Virchows Arch B Cell Path01 1989;57:303-13. 14. Tsukada T, McNutt MA, Ross R, Gown AM. HHF35, a muscle actin-specific monoclonal antibody. Am J Path01 1987;127:389-402. 15. Gugliotta P, Sapino A, Macri L, Skalli 0, Gabbiani G. Specific demonstration of myoepithelial cells by anti-alpha smooth muscle actin antibody. J Histochem Cytochem 1988;36:65963. 16. Zarbo RJ, Bacchi CE, Gown AM, Muscle-specific protein expressionin normal salivary glands and pleomorphic adenomas; an immunocytochemical study with biochemical confirmation. Mod Path01 1991;4:621-6. 17. Dardick I, Ostrynski VL, Ekem JK, Leung R, Burford-Mason AP. Immunohistochemical and ultrastructural correlates of muscle-actin expression in pleomorphic adenomas and myoepitheliomas based on comparison of formalin and methanol fixation. Virchows Archiv A Path01 Anat Histouathol 1992; 421:95-104. 18. Hirano T, Gluckman JL, deVries EJ. The expressionof vascular smooth-muscle actin in salivary gland tumors. Arch Otolaryngol Head Neck Surg 1990;116:692-6. 19. Araujo VC, Araujo NS. Vimentin as a marker of myoepithelial cells in salivary gland tumors. Eur Arch Otorhinolaryngol 1990;247:252-5. 20. Erlandson RA, Cardon-Cardo C, Higgins PJ. Histogenesis of
de Araujo,
Carvalho, and de Araujo
391
benign pleomorphic adenoma (mixed tumor) of the major salivary glands: an ultrastructural and immunohistochemical studv. Am J Sure Path01 1984:8:803-20. 21. Burns BF, Dar&k I, Parks’ WR. Intermediate filament expression in normal parotid glands and pleomorphic adenomas. Virchows Arch A Path01 Anat Histopathol 1988; 413:103-12. 22. Yamada K, Shinohara H, Takai Y, Mori M. Monoclonal antibody-detected vimentin distribution in pleomorphic adenoma of salivary glands. J Oral Path01 1988;17:348-55. 23. Dardick I, Cave11S, Boivin M, et al. Salivary gland myoepithelial variants histological, ultrastructural, and immunocytochemical features. Virchows Arch A Path01 Anat Histopath01 1989;416:25-42. 24. Dardick I, Thomas MJ, van Nostrand AWP. Myoepithelioma, new concepts of histology and classification: a light and electron microscopy study. Ultrastruct Path01 1989;13:187-224. 25. Araujo VC, Souza SCOM, Souza LB, Araujo NS. Estudo histopatologico e imunohistoquimico do mioepitelioma de elbndula salivar. Rev Odont USP 1990:4:62-6. 26. Caselitz J, Schulze I, Seifert G. Adenoid cystic carcinoma of the salivary glands: an immunohistochemical study. J Oral Path01 1986;15:308-18. 21. Chomete G, Auriol M, Vaillant JM, Kasai T, Okada Y, Mori M. Heterogeneity and co-expression of intermediate filament proteins in adenoid cystic carcinoma of salivary glands. Path01Bio (Paris) 1991;39:110-6. 28. Eversole LR. Histogenetic classification of salivary tumors. Arch Path01 1971;92:433-43. 29. Regezi JA, Batsakis JG. Histogenesis of salivary gland neoplasm. Otolaryngol Clin North-Am 1977;10:297-307. 30. Arauio VC. Souza SOM. Nunes FD. Arauio NS. Immunohistochemical analysis of 89 benign salivary gland tumors. Rev Odont USP 1991;5:1-6. 31. Araujo VC, Souza SOM, Jaeger RG, Araujo NS. Immunohistochemical study of malignant salivary gland tumors: an analysis of 67 cases.Rev Odont USP 1991;5:37-42. 32. Tsukada T, Tippens DBA, Gordon D, Ross R, Gown AM. HHF35: a muscle-actin-specific monoclonal antibody. Am J Path01 1987;126:51-60. 33. Leoncini P, Cintorino M, Vindigni C, et al. Distribution of cytoskeletal and contractile proteins in normal and tumor-bearing salivary and lacrimal glands. Virchows Archiv A Path01 Anat Histopathol 1988;412:329-37. 34. Tandler B, Denning CR, Mandel ID, Kutscher AH. Ultrastructure of human labial salivary glands, III: myoepithelium and ducts. J Morph01 1970;130:227-46. 35. Batsakis JG. Salivary gland neoplasia: an outcome of modified morphogenesis and cyto differentiation. ORAL SURG ORAL MED ORAL PATHOL 1980;49:229-32.
36. Gould VE. The coexpressionof distinct classesof intermediate filaments in human neoplasms.Arch Path01 Lab Med 1985; 109:984-5. 37. GustafssonH, Kjorell U, Eriksson A, Virtanen I, Thornell LE. Distribution of intermediate filament proteins in developing and adult salivary glands in man. Anat Embryo1 1988;178:34351. 38. Ward JM, Stevens JL, Konishi H, et al. Vimentin metaplasia in renal cortical tubules of preneoplastic, neoplastic, aging, and regenerative lesions of rats and humans. Am J Path01 1992; 1411955-64. Reprint requests:
V. C. de Araujo Disciplina de Patologia Bucal Facuidade de Odontologia Universidade de Sgo Paulo C. P. 8216, Sao Paulo, S. P. Brasil