Accurate quantification of cells recovered by bronchoalveolar lavage

Quantification of the differential cell count and total number of cells recovered from the lower respiratory tract by bronchoalveolar lavage is a valuable technique for evaluating the alveolitis of patients with inflammatory disorders of the lower respiratory tract. The most commonly used technique for the evaluation of cells recovered by lavage has been to concentrate cells by centrifugation and then to determine total cell number using a hemocytometer and differential cell count from a Wright-Giemsa-stained cytocentrifuge preparation. However, we have noted that the percentage of small cells present in the original cell suspension recovered by lavage is greater than the percentage of lymphocytes identified on cytocentrifuge preparations. Therefore, we developed procedures for determining differential cell counts on lavage cells collected on Millipore filters and stained with hematoxylin-eosin (filter preparations) and compared the results of differential cell counts performed on filter preparations with those obtained using cytocentrifuge preparations. When cells recovered by lavage were collected on filter preparations, accurate differential cell counts were obtained, as confirmed by performing differential cell counts on cell mixtures of known composition, and by comparing differential cell counts obtained using filter preparations stained with hematoxylin-eosin with those obtained using filter preparations stained with a peroxidase cytochemical stain. The morphology of cells displayed on filter preparations was excellent, and interobserver variability in quantitating cell types recovered by lavage was less than 3%. In contrast, the percentage of lymphocytes identified on cytocentrifuge preparations was 73 ± 29% of that identified on filter preparations, suggesting that lymphocytes were selectively lost during preparation of cytocentrifuge preparations. The magnitude of the loss was variable, and altering the procedures used in making cytocentrifuge preparations did not reduce this selective loss of lymphocytes. In addition, we noted that concentration and 'washing' of lavage cells results in cell loss (22 ± 20% cells lost after 3 centrifugations). This cell loss was variable in magnitude and was not restricted to a single cell type. Therefore, quantification of total cells recovered is best performed by counting an aliquot of the original cell suspension. These observations suggest that the results of differential counts and total counts of cells recovered by bronchoalveolar lavage must be interpreted in the context of the methods used to obtain these data. Determination of differential cell counts on cells displayed on filter preparations and determination of total cell counts on unconcentrated cell suspensions appear to accurately reflect the number and types of cells originally recovered by lavage.