Search

The reliability of labelling routinely prepared air-dried peripheral-blood and bone-marrow smears with the alkaline-phosphatase/anti-alkaline-phosphatase (APAAP) method for phenotyping leukaemias was evaluated in 259 cases of haematological malignancy. All the cellular markers currently used in leukaemia diagnosis could be detected with this technique. Since the technique can be used on blood and marrow smears even after storage or postal transport, it is suggested that APAAP labelling should become the method of choice for immunophenotyping neoplastic samples in many haematology laboratories not equipped with specialised immunological diagnostic facilities. The method should also be valuable for epidemiological studies of leukaemia.

Epithelial membrane antigen (EMA) is generally assumed to be expressed only on epithelial cells. However, EMA was also found on reactive and neoplastic plasma cells, on some non-Hodgkin's lymphomas (particularly cases of T-cell lymphoma and "malignant histiocytosis"), and on Reed-Sternberg cells in cases of lymphocyte-predominant Hodgkin's disease. EMA was also induced on normal blood lymphocytes by exposure to human T-lymphotropic virus type II or phytohaemagglutinin. The specificity of these aberrant reactions was confirmed by the use of three different monoclonal anti-EMA antibodies (E29, HMFG2, and LICR.LON/M8), and by showing that EMA in lymphoid tissue is similar in molecular weight to EMA in human milk. These results indicate that additional anti-epithelial antibodies (eg, anti-cytokeratins) should be used in conjunction with anti-EMA for tumour diagnosis.

In order to assess the clinical value of immunohistological analysis of tumours of uncertain origin, 84 patients have been followed up over a period of up to 9 years. All biopsies had been referred because of diagnostic difficulties. In keeping with previous studies the majority of the cases (69%) were non-Hodgkin's lymphomas, followed by carcinoma (23%) with a mixture of melanomas and sarcomas making up the remaining 8%. The survival of lymphoma cases was clearly better (approximately twice as long) than that of patients with other tumour types. In spite of the difficulties in their histological diagnosis the survival curve for lymphomas, recognised by immunocytochemical means, was very similar to that for the series of high-grade lymphomas reported by the Kiel group. These findings show that lymphomas recognised immunocytochemically behave similarly to other lymphomas and confirm that immunohistological analysis of tumours of uncertain origin is of practical clinical relevance.

Conventional fixation and paraffin embedding of human tissue destroys most of the antigens detected with currently available antibodies, while cryostat sections give poor morphological detail and require freezer space for storage. However, when tissue was freeze-dried before being embedded in paraffin S then stained for a wide range of antigens recognised by monoclonal antibodies (including T-cell and B-cell antigens), morphological preservation was clearly better, and the labelling reactions were of equal (or greater) intensity, than with cryostat sections. As the number of diagnostically important monoclonal antibodies continues to increase the use of freeze-dried paraffin sections should facilitate the introduction of immunohistological techniques for both routine histopathological diagnostic work and immunological research.