Granulocyte Contamination of Ficoll-Hypaque Preparations of Mononuclear Cells following Thermal Injury May Lead to Substantial Overestimation of Lymphocyte Recovery

During ongoing flow cytometric studies of burned patient blood leukocytes, it was noted frequently that large numbers of granulocytes were present along with the mononuclear cells at the plasma/Ficoll-Hypaque (F-H) interface following centrifugation over F-H. Since differential WBC counts are not routinely performed on F-H interface cells, it is possible that many previous immunologic studies of burned patients have greatly overestimated numbers of lymphocytes recovered. The present study sought to quantify the extent to which granulocyte contamination of F-H separated cells occurs following burn injury. Blood from 15 thermally injured patients (7-55% total body surface area burn) was studied serially at 24 hr, 48 hr, and weekly thereafter through 6 weeks postburn (PB). Controls were age-matched normals (No. of control bloods=59). Three-part differential cell counts (lymphocytes, monocytes, and granulocytes) were performed on both F-H interface cells and RBC-lysed whole blood. Counts were performed by light scatter analysis on a flow cytometer. Except at 48 hr, at every time studied through 4 weeks PB, there was significant contamination of F-H interface cells with granulocytes. At 24 hr PB, 41 ± 9% of the interface cells were granulocytes while at 4 weeks, PB 24 ± 8% of the interface cells were granulocytes. The data did not support the interpretation that this increase in F-H interface granulocytes was simply reflective of the granulocytosis commonly observed after burn. Thus artificial generation of granulocytosis by addition of extra normal leukocytes to normal blood resulted in complete separation of granulocytes from mononuclear cells following centrifugation over F-H. Due to the high percentage of granulocytes at the F-H interface and a substantial decrease in the lymphocyte:monocyte ratio, it is demonstrated that numbers of lymphocytes could be overestimated by as much as 60% if quantitative differential cell counts are not performed. These results suggest that granulocytes from thermally injured individuals have lower densities than those of normals such that large numbers remain at the F-H interface. Further, these data are a caveat to further work using density preparative techniques without concomitant differential cell counting of the leukocytes obtained.