Accuracy of routine flow-cytometric bitmap selection for three leukocyte populations

A double-blind study was performed with peripheral blood of 41 human subjects to check the accuracy of determination of lymphocyte, monocyte, and granulocyte windows with which every flow cytometric analysis of leukocyte markers starts. White blood cell suspensions were prepared according to the whole blood method and analyzed on an EPICS-C flow cytometer using the two-parameter 90 degrees light scatter vs. forward angle light scatter (granularity vs. cell size) data distribution. Windows (bitmaps) for lymphocytes, monocytes, and granulocytes were drawn and numbers of cells determined in each. The proportions of lymphocytes, monocytes, and granulocytes were calculated in relation to total cell number, counted and in relation to the sum of cells in three bitmaps, and then compared with proportions determined by microscopic whole blood cell (WBC) differential and a WBC differential determined in an automated hematology analyzer. Average proportions of lymphocytes obtained by the flow cytometer were significantly lower than those obtained by either microscopic or automated differential, suggesting that some of the relevant cells were not included in the bitmaps. Granulocyte proportion related to total cell number was lower and that related to bitmap cell number higher than that obtained by microscopic and automatic differentials, suggesting that nongranulocytic cells were included in the granulocyte bitmaps. Proportions of lymphocytes and granulocytes obtained by the flow cytometer correlated well with those obtained by both microscopic and automatic differential. In contrast, the proportions of monocytes showed a poor correlation, which is probably due to their low number and delicate position in the distribution, and which makes them difficult to delineate.