Miettinen M, Wahlstrom T, Virtanen I, Talerman A, Astengo-Osuna C
Am J Surg Pathol. 1985 Sep;9(9):640-51. doi: 10.1097/00000478-198509000-00003.
Seventy ovarian sex-cord-stromal and germ-cell tumors were immunohistochemically studied for the presence of intermediate-filament proteins of different types used as markers for cellular differentiation. Cells of ovarian granulosa-cell tumors constantly expressed vimentin and appeared to lack cytokeratin. Two tumors previously classified as granulosa-cell tumors were reclassified as poorly differentiated "common" epithelial tumors based on their cytokeratin positivity, vimentin negativity, and morphologic features. Dysgerminomas and Leydig-cell tumors showed only vimentin positivity. Tubular structures in androblastomas, which are considered to represent Sertoli-cell differentiation, were cytokeratin positive, and thus differed from the majority of normal Sertoli cells that are known to express vimentin and not cytokeratin. Embryonal carcinomas, choriocarcinomas, and endodermal sinus tumors showed cytokeratin positivity in the neoplastic cells whereas vimentin was observed in the stromal cells. In immature teratomas, epithelial differentiation was demonstrated with cytokeratin antibodies, and neural and glial differentiation was also frequently demonstrated by immunostaining with antibodies to neurofilaments and glial fibrillary acidic protein. The results show that antibodies to intermediate filaments can be used in the differential diagnosis between ovarian epithelial and nonepithelial tumors, and they provide a very accurate additional method to characterize the cellular differentiation of ovarian neoplasms.
对70例卵巢性索间质肿瘤和生殖细胞肿瘤进行免疫组织化学研究,以检测不同类型的中间丝蛋白,这些蛋白用作细胞分化的标志物。卵巢颗粒细胞瘤的细胞持续表达波形蛋白,似乎缺乏细胞角蛋白。根据其细胞角蛋白阳性、波形蛋白阴性和形态学特征,将2例先前分类为颗粒细胞瘤的肿瘤重新分类为低分化“普通”上皮肿瘤。无性细胞瘤和睾丸间质细胞瘤仅显示波形蛋白阳性。男性母细胞瘤中被认为代表支持细胞分化的管状结构呈细胞角蛋白阳性,因此与大多数已知表达波形蛋白而不表达细胞角蛋白的正常支持细胞不同。胚胎癌、绒毛膜癌和内胚窦瘤的肿瘤细胞呈细胞角蛋白阳性,而波形蛋白见于间质细胞。在未成熟畸胎瘤中,细胞角蛋白抗体显示上皮分化,神经丝蛋白和胶质纤维酸性蛋白抗体免疫染色也经常显示神经和胶质分化。结果表明,中间丝抗体可用于卵巢上皮性肿瘤和非上皮性肿瘤的鉴别诊断,为卵巢肿瘤细胞分化的特征化提供了一种非常准确的辅助方法。