COORDINATE EXPRESSION OF CYTOKERATIN-7 AND CYTOKERATIN-20 DEFINES UNIQUE SUBSETS OF CARCINOMAS

We tested the hypothesis that the coordinate expression of cytokeratin 7 (CK 7) and cytokeratin 20 (CK 20) could distinguish among carcinomas arising from different primary sites. A total of 384 cases of carcinomas primary to various organs, as well as 16 cases of malignant mesothelioma, were evaluated using commercially available monoclonal antibodies and an avidin-biotin immunoperoxidase technique. The subset of tumors strongly expressing both CK 7 and CK 20 included virtually all bladder transitional cell carcinomas and the majority of pancreatic adenocarcinomas; the tumors negative for both CK 7 and CK 20 were largely restricted to hepatocellular, prostate, and renal cell carcinomas in addition to squamous cell and neuroendocrine carcinomas of lung. The CK 7-/CK 20+ immunophenotype, however, was highly characteristic of adenocarcinomas of colorectal origin, whereas CK 7+/CK 20- immunophenotype was typically seen in the vast majority of carcinomas arising from other sites, including ovary, endometrium, breast, and lung, as well as malignant mesothelioma. Gastric carcinomas were the most heterogeneous subgroup with respect to CK 7/CK 20 immunophenotype. In the subset of mucinous tumors, striking immunophenotypic differences were noted among those primary to the breast (CK 7+/CK 20-), gastrointestinal tract (CK 7-/CK 20+), and ovary (CK 7+/CK 20+). In all cases investigated, this CK immunophenotype was invariant in metastatic vs. primary tumors. It is concluded that, in the appropriate clinical setting, the CK 7/CK 20 immunophenotype of carcinomas is a valuable diagnostic marker in the determination of primary site of origin.

Adenocarcinoma in females detected in serous effusions: Cytomorphologic aspects and immunocytochemical reactivity to cytokeratins 7 and 20

OBJECTIVE: To investigate the usefulness of assessing the immunoreactivity of cytokeratins 7 (CK7) and 20 (CK20) as well as several cytomorphologic parameters in effusions with metastatic adenocarcinomas in the search for the primary site of the tumor. STUDY DESIGN: From the files of the Pathology Department, A. C. Camargo Hospital, we studied cytologic smears from 73 metastatic adenocarcinomas originally from the breast, 63 from the ovary, 40 from the lung and 32 from the stomach, looking for morphologic parameters that could have discriminant potential in suggesting the primary site in a routine situation, including intranuclear inclusions, prominent nucleoli, mitosis, signet-ring cells, psammoma bodies, nuclear crease, binucleation and multinucleation, papillary features, acinar profile (including ball cells) and single cells. Immunoreactions were performed with monoclonal antibodies to CK7 (OV-TL 12/30 and CK20 (Ks 20.8) and included morphologic analysis. Both analyses were studied in a blind fashion regarding the primary site of the tumors. RESULTS: Positivity ratios for breast, ovary, stomach and lung cases were 67.6%, 63.5%, 29.7% and 45.5%, respectively, for CK7 and 17.2%, 15.8%, 13.5% and 32.2%, respectively, for CK20. Discriminant analysis of morphologic and immunocytochemical parameters had an error rate of 42.9% in recognizing the primary site and a Wilk's lambda of .7290. CONCLUSION: The more efficient parameter with discriminant function was the papillary appearance showed by CK7, which should be used in further studies with a similar scope. The set of parameters used in this study were insufficient to discriminate the primary site of female adenocarcinomas in effusions with significant accuracy.

Differentiation of epithelial cells in the urinary tract

Uroplakins, cytokeratins and the apical plasma membrane were studied in the epithelia of mouse urinary tract. In the simple epithelium covering the inner medulla of the renal pelvis, no uroplakins or cytokeratin 20 were detected and cells had microvilli on their apical surface. The epithelium covering the inner band of the outer medulla became pseudostratified, with the upper layer consisting of large cells with stalks connecting them to the basal lamina. Uroplakins and cytokeratin 20 were not expressed in these cells. However, some superficial cells appeared without connections to the basal lamina; these cells expressed uroplakins Ia, Ib, II and III and cytokeratin 20, they contained sparse small uroplakin-positive cytoplasmic vesicles and their apical surface showed both microvilli and ridges. Cytokeratin 20 was seen as dots in the cytoplasm. This epithelium therefore showed partial urothelial differentiation. The epithelium covering the outer band of the outer medulla gradually changed from a two-layered to a three-layered urothelium with typical umbrella cells that contained all four uroplakins. Cytokeratin 20 was organized into a complex network. The epithelium possessed an asymmetric unit membrane at the apical cell surface and fusiform vesicles. Umbrella cells were also observed in the ureter and urinary bladder. In males and females, the urothelium ended in the bladder neck and was continued by a non-keratinized stratified epithelium in the urethra in which no urothelial cell differentiation markers were detected. We thus show here the expression, distribution and organization of specific proteins associated with the various cell types in the urinary tract epithelium.

Cytokeratin Profile in Exfoliative Cytology of Smokers

Fundação de Amparo à Pesquisa do Estado de São Paulo (FAPESP)