Analysis and sorting of apoptotic cells from fine-needle aspirates of excised human primary breast carcinomas.

Numerous recent studies have indicated the central role of apoptosis as a determinant of the growth abnormalities occurring with malignancy and of the effectiveness of a wide range of therapeutic manoeuvres in cancer treatment. However, there has been a relative paucity of studies measuring apoptosis in human solid tumours, because of the low incidence of apoptotic cells, the difficulty of identifying cells undergoing apoptosis, and the ethical and practical restrictions on obtaining repeat biopsies from patients during therapy. Fine-needle aspirates (FNAs) may be obtained from breast carcinomas as a minimally invasive technique allowing repeat sampling. We describe an approach in which the in situ end labelling (TUNEL) assay is applied to cells in FNAs prior to their analysis by flow cytometry, which allows many thousands of cells to be analysed automatically by objective criteria. Cells that were discriminated as apoptotic on flow cytometric analysis were sorted onto microscope slides and found to show nuclear morphology typical of apoptotic cells. A statistically significant relationship was found between the flow cytometric analysis and the conventional application of TUNEL on histological sections (P = 0.03). Repeat analyses of FNAs from 12 carcinomas showed a median 2.05% apoptotic cells and an overall coefficient of variation of 34.9%. Of the total variability in 12 tumours, 80% was attributed to variation between tumours, 12% between batches, and 8% was random. Thus, this technique should be able to detect the major differences in the percentage of apoptotic cells that occur between different tumours (range 0.3-11.3% by flow cytometry) and between different phases of treatment, and should provide a useful tool for further research on this process in solid tumours.