Immunohistochemical detection of cytotoxic lymphocytes in malignant serous effusions.

The use of ancillary techniques to aid in the diagnosis of metastatic carcinoma in serous effusions has been the subject of numerous studies. In this article, we study 35 cases of malignant effusions (metastatic adenocarcinoma) and 20 benign effusions using a panel of immunohistochemical markers to determine whether changes in the subpopulations of accompanying lymphoid cells can be detected with this technique and whether such changes are associated with the presence of malignancy. We noted a significant increase in cytotoxic lymphocytes, defined as the percentage of all lymphoid cells staining with an antibody to TIA-1 (an antigen localized to the cytotoxic granule membranes of cytotoxic T cells and natural killer cells) in malignant compared with benign effusions (23% vs. 12%; P < 0.05). In addition, nearly all cases in which cytotoxic lymphocytes composed > 20% of the lymphoid cell population contained metastatic tumor. Thus, immunohistochemical staining for TIA-1 can reliably detect cytotoxic lymphocytes in cell blocks of serous effusions; in addition, a relative increase in their number is associated with the presence of malignancy.

Cercariform cells: are they specific for transitional cell carcinoma?

BACKGROUND: Cercariform cells (CCs) recently have been described as a significant clue to the cytologic diagnosis of transitional cell carcinoma (TCC). The CC is a single tumor cell with a nucleated globular body and a unipolar cytoplasmic process with a nontapering, flattened, or bulbous end. METHODS: To determine the specificity of CCs in FNAs, 45 cases of poorly differentiated malignant epithelial neoplasms (PDMEN) and 11 cases of metastatic TCC were reviewed, and the number of CCs present in each case was counted. All cases had 1-19 air-dried Diff-Quik stained slides, 1-6 ethanol fixed Papanicolaou stained slides, and one Millipore filter preparation. RESULTS: CCs were identified in 16 of 45 cases of PDMEN (36%). Nine cases had 1-4 CCs, 3 cases had 5-9 CCs, 2 cases had 10-19 CCs, and 2 cases had >20 CCs. The 2 cases with >20 CCs included PDMEN of the lung and metastatic PDMEN in the liver. CCs were identified in all 11 cases of metastatic TCC. Two cases had 1-4 CCs, 3 cases had 5-9 CCs, 1 case had 10-19 CCs, and 5 cases had >20 CCs. The metastatic sites included the lymph nodes (six cases), liver (three cases), lung (one case), and sacrum (one case). CONCLUSIONS: The finding of a few CCs is not specific for TCC. CCs were observed in 16 of 45 cases of PDMEN (36%). The presence of >20 CCs is significantly more likely to occur in metastatic TCC (5 of 11 cases; 45%) than PDMEN (2 of 45 cases; 4%) (P<0.001). If no CC is identified in an FNA, it is unlikely that the case represents a metastatic TCC because all 11 cases of metastatic TCC in the current study had at least 1 CC. It is interesting to note that all 11 cases of metastatic TCC were found to have CCs on the filter, suggesting that the morphology of the CC is not a smearing artifact.

Childhood melanoma: a clinicopathological study of 22 cases.

Childhood melanoma is a rare disease with an estimated incidence of one per million per year. Careful study of childhood melanoma patients is critical due to the limited data currently available pertaining to this disease. Twenty-two children 15 years of age or under with malignant melanoma were treated at the Pigmented Lesion Clinic of Massachusetts General Hospital over a 33-year period. The medical records of all patients were reviewed, as well as the histologic characteristics of the lesions. Patients who were initially diagnosed with malignant melanoma but on review found to have Spitz naevi were not included in our study. Ten patients were boys and 12 were girls. The median ages of the boys and girls in our study were 12.9 and 13.6 years, respectively. Among the classified primary melanomas, 10 were superficial spreading, three were borderline/minimal deviation, two were nodular and one was melanoma in situ. Four of 22 patients had a documented family history of melanoma, and two additional patients had a family history of dysplastic naevi. A majority of lesions (14/22) arose in association with a precursor lesion. Two children died of disease at 1 and at 7 years following initial diagnosis. Eight patients had documented metastases. Since the majority of melanomas arose in association with a precursor lesion, follow-up of children with congenital and/ or dysplastic naevi is recommended. An interesting finding was the sometimes paradoxical behaviour of relatively thin lesions with metastases. Thus, a high index of suspicion is needed by the clinician confronted with melanocytic lesions of childhood. We found children from age 12 to 15 to be more at risk for the development of melanoma than younger children. Melanoma presenting before the age of 10 is very unusual.