No abstract

Aim-To discriminate between adenocarcinomas that are primary to the ovary and metastatic to the ovary, especially of colonic and breast origin, by immunohistochemistry, using stepwise discriminant analysis or a decision tree. Methods-312 routinely processed, formalin fixed tissue specimens were used. The tumours were divided into a learning set (n = 159), composed of primary tumours of ovary, breast, and colon, and a test set, comprising 134 metastases from these sites and an additional 19 primary ovarian carcinomas. The immunohistochemical panel was composed of antibodies against cytokeratin 7 (CK7) and 20 (CK20), CA125, vimentin, carcinoembryonic antigen (CEA), gross cystic disease fluid protein-15 (GCDFP-15), and the oestrogen receptor (ER). The staining results of the tumours were expressed as the product of the staining intensity and the percentage of positive tumour cells. Analyses were first performed on the learning set and then evaluated on the test set. Results-Although the immunostaining patterns showed a considerable overlap between the three types of adenocarcinoma, the breast carcinomas were typically positive for GCDFP-15 and often for ER, and negative for vimentin. Ovarian carcinomas were always positive for CK7 and to a lesser extent for CA125. Colonic carcinomas showed prominent positivity for CEA and CK20, while no staining was seen for ER and vimentin. In discriminant analysis, six antibodies ( CK7, CK20, CA125, CEA, ER, and GCDFP-15) appeared to be necessary for optimal classification: 89% of the learning set and 82% of the test set were classified correctly. In the decision tree, only four antibodies ( CK7, CEA, ER, and GCDFP-15) were used to obtain a correct classification score of 89% for the learning set and 84% for the test set. Conclusions-Using a semiquantitative assessment of the immunostaining results by a restricted panel of six antibodies with stepwise discriminant analysis, 80-90% of the adenocarcinomas of colon, breast, and ovary can be correctly classified. Discriminant analysis is computer aided and therefore an easy method and for each case a probability value of the classification result is obtained. The intuitive decision tree method provides a slightly better result, requires only four antibodies, and oVers a more practical method for the surgical pathologist. (J Clin Pathol 1999;52:283-290) Keywords: immunohistochemistry; ovarian carcinoma; colorectal carcinoma; breast carcinoma; diVerential diagnosis; discriminant analysis; decision tree Ovarian cancers constitute about 2.3% of all cancers. In the USA it is the fourth most common source of cancer death in women, exceeded only by breast, colorectal, and lung malignancies.1 Cancer of the ovary constitutes about 25% of the gynaecological cancers. Up to 7% of the ovarian cancers, however, comprise metastases from various other locations.2-4 About 40% of these originate from the colon.5-7 These metastases often form large cystic masses or show abundant necrosis, mimicking primary endometrioid or mucinous ovarian ca...

Large mononuclear cells with long, actively moving cytoplasmic veils were observed in lymph coming from the skin. The enzyme histochemistry and ultrastructure of these cells suggested that they are related to epidermal Langerhans cells and interdigitating cells in the lymph node. It has been reported that Langerhans cells and interdigitating cells play a role in contact hypersensitivity by taking up antigen and presenting it to thymus-dependent lymphocytes, and it is likely that the veiled cells in the lymph are also involved. After skin-painting with 1-fluoro-2,4-dinitrobenzene (DNFB), the veiled cells in lymph coming from the site of painting became more active and were observed contacting other cells present in the lymph; many large cellular aggregates were found. Since neutrophilic leucocytes and mononuclear phagocytes were the predominating cell types in this lymph, there was no evidence for a massive recruitment of immunocompetent lymphocytes at the site of painting. Neonatally thymectomized pigs do not develop allergic reactivity to DNFB. It is of interest that the number of veiled cells and their ability to form large cellular aggregates was not affected in these animals. Therefore, it is unlikely that the defect in responsiveness can be attributed to a failure in the function of veiled cells.