Absence of Epstein-Barr virus in medullary carcinoma of the breast as demonstrated by immunophenotyping, in situ hybridization and polymerase chain reaction.

Medullary carcinoma of the breast is an epithelial malignant proliferation that shares many characteristics (macroscopic, microscopic, epidemiologic, and prognostic) with lymphoepithelioma-like carcinomas of various sites. The authors hypothesized that they could also share the same etiologic agent, the Epstein-Barr virus (EBV). Epstein-Barr virus, a virus of the herpesvirus family, is to be associated with lymphoepithelioma-like carcinomas of the nasopharynx, stomach, lung, thymus, and salivary gland. Therefore, the authors looked for the virus in a series of 10 medullary carcinomas of the breast. Using immunohistochemistry, in situ hybridization and polymerase chain reaction, this investigation failed to show evidence of EBV. Similar negative results have been reported in lymphoepithelioma-like carcinomas arising in the skin and in the uterine cervix, which like the breast do not originate in the foregut. These results suggest that the pathogenesis of these tumors is not unique, implicating probably different etiopathogenic entities.

Medullary carcinoma of the breast: a tumour lacking keratin 19.

The presence of keratin 19 (K19) was searched for by immunostaining in 16 medullary carcinomas, comprising 12 typical and four atypical cases, in 29 undifferentiated high-grade carcinomas (NOS-HG) with conspicuous lymphoid response and in 12 well differentiated low-grade carcinomas (NOS-LG). The medullary carcinomas were all negative whereas 23 of the high-grade and all 12 low-grade carcinomas expressed K19. Staining for K19 could be of value in the differential diagnosis of these tumours. Furthermore, these findings, with other observations, raise the possibility that medullary carcinoma cells could be linked to precursor cells of the terminal duct lobular units because both populations share several characteristics.

Immunohistochemical identification of myoepithelial, epithelial, and connective tissue cells in canine mammary tumors.

Fifty-eight formalin-fixed paraffin-embedded canine mammary tumors, 19 malignant and 39 benign, were used in this study. Tumors were obtained from dogs submitted for surgical resection of lesions at private veterinary practices in Brussels or from the surgery unit of the Faculty of Veterinary Medicine, University of Liège. Immunohistochemical evaluation was performed, using monoclonal antibodies directed against keratins 8-18 and 19, vimentin, desmin, and alpha-actin and polyclonal antibodies directed against high-molecular-weight keratins and S-100 protein. The main cell types, epithelial, myoepithelial, and connective, were identified, and myoepithelial cells represented the major component of most tumors, both benign and malignant. Myoepithelial cells had five patterns: resting and proliferative suprabasal cells, spindle and star-shaped interstitial cells, and cartilage. Reactivity to keratin 19, vimentin, alpha-actin, and S-100 protein suggested a progressive transformation from resting cells to cartilage. Epithelial cell reactivities were limited to keratins; only keratinized cells were positive for polyclonal keratins. Myofibroblasts were positive for both vimentin and alpha-actin, and connective tissue cells were positive for vimentin. Myoepithelial cells appeared to be the major component of carcinomas, justifying reevaluation and simplification of histomorphologic classifications, with a "pleomorphic carcinoma" group including all carcinomas except squamous, mucinous, and comedo carcinomas. Immunohistochemical evaluation, in addition to routine hematoxylin and eosin histopathologic evaluation is recommended for precise classification of canine mammary tumors.

[Immunocytochemical assay of estrogen receptors in puncture products of breast carcinomas]. / Dosage immunocytochimique du récepteur d'oestrogène dans les produits de ponctions de carcinomes mammaires.

In a pilot study, estrogen receptors (ER) were assayed on 42 surgically removed breast tumors by the following 3 methods: biochemical assay with dextran coated charcoal (DCC), Abbott immunoenzymatic (ER-EIA) and immunocytochemical (ER-ICA) technics. DCC and ER-EIA were performed on biopsy specimens while ER-ICA was run on cytocentrifugated cells obtained by fine needle aspiration (FNA). ER contents were expressed according to an index taking into account the proportion of colored neoplastic cells and the intensity of staining. Statistical correlation coefficient (Spearman and Kendall) concordance, sensitivity and specificity between the results were calculated (ER - ICA/ER - EIA: P less than 0.001, r = 0.38, concordance = 83%, sensitivity = 86%, specificity = 77%; ER - ICA/DCC: P less than 0.05, r = 0.22, concordance = 77%, sensitivity = 85%, specificity = 63%; ER - EIA/DCC: P less than 0.001, (r = 0.60). As previously reported, both immunoassays showed good agreement. The weaker but nevertheless significant correlation found with reference DCC may be due to the heterogeneity of tumoral ER content. This hypothesis is supported by the variability of ER - ICA assays on multiple FNA performed in 16 cases from our series. Use of multidirectional FNA slightly improved the results. Nevertheless, ER - ICA appear to be a good semi-quantitative method and might be helpful in the follow-up of metastasis treated with anti-estrogen, especially in small lesions not assayable by DCC.

Human medullary thyroid carcinoma: cell cultures and xenotransplants in nude mice. Immunocytochemistry and calcitonin secretion.

Occult primary and recurrent medullary thyroid carcinomas (MTC) detected only by elevated calcitonin levels in the peripheral blood, generally after pentagastrin-test stimulation, are difficult to localize. Some new imaging procedures with radionuclide tracers or radiolabelled monoclonal antibodies against carcinoembryonic antigen seem to bring some potentially therapeutic benefits. We report our results with cell cultures and xenotransplants of human MTC with the intention of establishing reproducible models in vitro and in vivo. Cell cultures secrete calcitonin at up to 1200 pg/ml for periods ranging from 3 to 13 weeks. Immunocytochemistry detects cytoplasmic granules positive for calcitonin in polygonal epithelioid cells with dendritic processes. Xenotransplants in nude mice fare better in the subcutaneous axilla than in the subrenal capsule assay. In the former location the tumor-take is good and calcitonin is detected in the blood of the tumor-bearing animals, at levels ranging from 286 to more than 20,000 pg/ml. These models would be potentially usable as targets for radionuclide tracers and/or radiolabelled monoclonal antibodies.