Search

Immunohistochemical staining for S-100 protein and the intermediate filaments keratin and vimentin, was made in 41 salivary adenomas. In pleomorphic adenomas, great heterogeneity in the staining, as well as multiple and co-expressions of these proteins were found in the outer tumor cells of tubulo-ductal structures and modified myoepithelial cells, but not in the luminal tumor cells. All the outer tumor cells stained for S-100 protein, 97% for K8.12 keratin and 85% for vimentin. Of these cells, 29% showed multiple expression of K8.12 keratin, vimentin, and S-100 protein, and 17% showed co-expression of K8.12 and S-100 protein. Modified and neoplastic myoepithelial cells showed similar expressions of these proteins to those of outer tumor cells; myoepithelioma cells displayed the most complicated pattern, being positive for KL1, PKK1, and K8.12 keratins, vimentin and S-100 protein. In luminal tumor cells there was a heterogeneous expression of KL1 and PKK1 in 82%, and of KL1, PKK1, and K8.12 in only 14.7%. Based on the immunohistochemical findings obtained with different monoclonal antibodies in pleomorphic salivary adenomas, outer tumor cells may be derived from ductal basal cells and luminal tumor cells from intercalated duct cells.

A peroxidase-antiperoxidase technique for S-100 protein has been applied to 68 salivary glands. These included 17 pleomorphic adenomas, seven adenoid cystic carcinomas, 23 adenolymphomas and a number of other neoplasms. In addition, five specimens of myoepithelial sialadenitis ('benign lymphoepithelial lesion') and five normal parotid glands were included. Consistent results were obtained, myoepithelial cells and cells in myxoid and chondroid areas in pleomorphic adenomas staining intensely. In adenoid cystic carcinoma, on the other hand, there was no staining. The adenolymphomas possessed intensely S-100 protein-positive cells in the interfollicular lymphoid areas; these were probably interdigitating reticulum cells. In addition, branching structures, probably corresponding to Langerhans' cells, were observed in the epithelium of adenolymphomas.

Immunohistochemical expression of 8 cases of mucoepidermoid carcinomas (G-I, 3 cases; G-II, 2 cases; and G-III, 3 cases) revealed marked heterogeneity of the proteins examined. Immunohistochemically detectable keratins (TK, KL1, and PKK1) were distributed in epidermoid cells, but were absent in mucous secreting cells. Strongly positive deposits of keratin proteins were detected in squamoid tumor cells in the G-I tumors. The tumor cells displayed positive staining for S-100 alpha, but did not stain with polyclonal S-100 antiserum or with monoclonal S-100 beta. The cells showing highest reactivity for S-100 protein were scattered in neoplastic foci and were probably Langerhans cells. Lactoferrin and lysozyme reactions were generally negative in tumor foci; but a positive reaction for lactoferrin was found in luminal tumor cells although rarely, and lysozyme staining was occasionally noted in histiocytes in the stroma. Amylase activity was usually absent in the tumor cells, with the exception of one case in which it was confined to the tumor cells. Mucoepidermoid carcinomas of various grades indicated marked heterogeneity in terms of various immunohistochemically detectable proteins.

Immunohistochemical studies were performed for the presence of S-100b protein in non-neoplastic and neoplastic salivary gland tissues by the peroxidase anti-peroxidase (PAP) method. Some cases of pleomorphic adenoma were investigated by immuno-electron microscopy. S-100b protein could not be detected in epithelial cells of intercalated ducts, acini, striated ducts and excretory ducts of non-neoplastic salivary gland. However, myoepithelial cells surrounding the acini and intercalated ducts were specifically stained by S-100b protein. In pleomorphic adenomas, S-100b protein-positive cells could be mostly observed in the myxoid and chondroid areas, and the basal layer cells of the double-layered ductal cells were also positive. In clear cell adenoma, the clear cells were also S-100b protein positive. In adenoid cystic carcinomas, S-100b protein-positive cells could be found in trabecular areas, but not in tumor cells showing cribriform-pattern. In other tumors (Warthin's tumor, oxyphilic adenoma, basal cell adenoma, mucoepidermoid tumor and acinar cell carcinoma), S-100b protein positive cells were seldom observed. Immuno-electron microscopically, S-100b protein was diffusely distributed in the cytoplasm of myoepithelial cells as well as of tumor cells of pleomorphic adenoma, being distributed especially on the membrane of endoplasmic reticulum and the outer nuclear membrane.

Pleomorphic adenoma of the salivary glands was studied by the peroxidase-antiperoxidase method and Ouchterlony's double diffusion for the presence of so-called nervous system-specific proteins, S-100 protein, glial fibrillary acidic protein (GFAP), and astroprotein. Positive immunostaining for S-100 protein was observed in tumor cells of epithelial, myxomatous, and chondroid areas. An immunodiffusion test confirmed its presence in the tumor. Normal salivary glands were also stained with anti-S-100 serum, but the reaction was less intense than that of tumor. Specific immunostaining for GFAP and astroprotein was found in a small number of cells of myxomatous and chondroid areas and occasionally in cells in epithelial areas. An immunodiffusion test suggests that the GFAP-related antigen of the pleomorphic adenoma showing a minor heterogeneity of antigenic determinants is almost identical with GFAP. Normal salivary glands did not stain with GFAP or astroprotein antisera, nor did they react with anti-GFAP serum by immunodiffusion. Thus, the S-100 protein and GFAP-related antigen may be actively synthesized in adenoma cells during the course of tumor development. In addition, the GFAP-related antigen is considered to be a tumor-associated antigen of pleomorphic adenoma of the salivary glands.

Immunohistochemical distribution of S-100 protein was evaluated in 129 tumors from major and minor salivary glands. Also, two sensitive immunoperoxidase avidin-biotin methods using either overnight incubation with primary antibody or pretreatment trypsin digestion and half-hour incubation were compared. Tumors with S-100 protein immunoreactivity were demonstrated in numerous benign and malignant histologic categories. Adenoid cystic carcinomas, carcinomas ex pleomorphic adenoma, clear cell carcinomas, and adenocarcinomas NOS showed inconsistent positive staining, whereas all monomorphic and pleomorphic adenomas and polymorphous low grade adenocarcinomas examined stained positively. No staining was observed in mucoepidermoid carcinomas or acinic cell carcinomas. Mesenchymal-like tumor cells with positive immunostaining were seen only in pleomorphic adenomas and trabecular-tubular adenomas. Equivalent results were found with both overnight and same-day digestion techniques. The consistent S-100 protein staining in some histologic tumor categories (pleomorphic and monomorphic adenoma and polymorphous low grade adenocarcinoma) compared to mucoepidermoid carcinoma that is devoid of S-100 protein immunoreactivity has application to some microscopic differential diagnostic situations. Inconsistent staining of adenoid cystic carcinomas and adenocarcinomas did not allow discrimination from other benign and malignant salivary gland tumors with similar histomorphology.

Immunohistochemical localization of S-100 protein, its alpha and beta subunits, and glial fibrillary acidic protein (GFAP) in normal and neoplastic salivary glands was studied by the peroxidase-antiperoxidase method and immunoblot analysis. Positive immunostaining for S-100 protein was observed in pleomorphic adenoma, adenolymphoma, tubular adenoma, adenoid cystic carcinoma, acinic cell tumour, adenocarcinoma and carcinoma in pleomorphic adenoma. S-100 protein was localized in myoepithelial cells, epithelial cells of intercalated ducts and serous acinar cells of normal salivary gland. Both alpha and beta subunits of S-100 protein showed almost identical distribution in normal and neoplastic salivary glands, but skeletal muscle cells were alpha-positive/beta-negative whereas Schwann cells and fat cells were alpha-negative/beta-positive in the stroma and neighbouring tissue. GFAP was only found in pleomorphic adenoma and its malignant counterpart. Immunoblot analysis showed that the GFAP-related antigen consisted of several polypeptide bands with a molecular weight ranging between 35,000 to 50,000 daltons.

The clinician is almost entirely dependent on the histopathologist to accurately diagnose minor salivary gland tumours, but in some cases the histological interpretation of the specimen is very difficult. Recently it has been demonstrated using immunohistochemical techniques that S-100 protein is present in some salivary gland tissues and its localization has been used as an aid in the differentiation of major salivary gland tumours. To assess its value in the diagnosis of minor salivary gland tumours it was localized in sections from 15 such tumours using both a standard peroxidase-antiperoxidase (PAP) and a newly developed immunogold-silver staining sequence (IGSS) technique. Strong staining for S-100 protein was seen in the nuclei and cytoplasm of the cellular areas and also in the cells in the chondroid and myxoid areas of pleomorphic adenomas. Generally the staining was more intense and widespread with the IGSS method. No staining was observed in any of the other tumour types. We conclude that S-100 protein localization is a valuable aid in the differentiation of minor salivary gland tumours. Furthermore, the IGSS method enables more sensitive 'reading' of the staining reaction.