The leukocyte depletion capacity and performance of 5 filters designed for filtration of red cell concentrates (RCC) were compared by counting leukocytes, measuring red cell volumes and by histological examination of the filters after use. To eliminate interdonor differences, 5 buffy-coat-poor RCC were pooled (in each of 10 experiments) and subsequently split up into the original bags. The RCC were passed over the Cellselect filter, a column filled with cellulose acetate, and over flat-bed polyester filters: the Cellselect Optima, the Pall RC 50, the Leukostop and the Sepacell R-500. The filtration was shortest with the Pall RC 50 (p less than 0.001 compared to the other 4 filters). Leukocyte removal was most effective with the cellulose acetate filter (p less than 0.01 compared to the other 4 filters) followed by the Cellselect Optima polyester filter (p less than 0.02 compared to the remaining 3 filters). Residual leukocytes did not exceed 50 x 10(6) for any brand of filter. Red cell recovery was similar for all 5 filters with mean values from 86.1 to 89.2%. The leukocyte numbers, counted in Türk's solution or in propidium iodide, gave comparable values in hemocytometers applying light microscopy or fluorescent microscopy, respectively. Histological examination showed that lymphocytes were mainly removed by trapping, whereas granulocytes showed a variable pattern: adhesion in presence of platelets or trapping.

Leukocyte filtration was performed with HTLV-I-infected blood and with blood supplemented with cultured HTLV-I-transformed cells. Reduction of infectivity upon leukocyte filtration was determined by the polymerase chain reaction (PCR) using primers indicative for the HTLV-I-pol and tax genes. Two different commercially available filters were used: a column-shaped cellulose acetate and a flat-bed polyester filter. Both filters yielded reduction of at least 3 10logs for cultured HTLV-I-infected cells. When blood from HTLV-I-infected individuals was used for filtration, the number of infected cells was reduced by 1-3 10logs. Although filtered blood as yet cannot be regarded as safe, it is concluded that leukocyte filtration of HTLV-I-infected blood potentially contributes to reducing the spread of HTLV-I by blood transfusion.

A flow cytometric method for the detection of low amounts of lymphocytes, monocytes, and granulocytes in filtered red cells (RBCs) was evaluated. In this procedure, the RBCs in the samples were lysed by ammonium chloride treatment and the white cells (WBCs) were detected by flow cytometry according to their specific light-scattering properties. The identity of the WBC subpopulations was confirmed by immunofluorescence with monoclonal antibodies specific for each cell type. Flow cytometric determination of WBCs in filtered RBCs correlated with numbers obtained by both a hemocytometer (r = 0.76) and a radioimmunoassay (r = 0.79). Total numbers of WBCs in RBCs measured by flow cytometry were 59 +/- 13 percent (n = 7) of those measured by electronic particle counting, 32 +/- 6 percent (n = 25) by hemocytometer, and 48 +/- 11 percent (n = 29) by radioimmunoassay. Lymphocytes added to filtered RBCs in a concentration of 1.37 cells per microL were detected at an average of 0.56 +/- 0.22 cells per microL (n = 3). Results with monoclonal antibodies indicated an altered expression of membrane markers on granulocytes after RBC filtration, as seen with cell activation. The inefficiency of the flow cytometric method to detect the total number of WBCs calculated by other methods may reflect filtration-induced changes in light-scattering properties of the WBCs. Although the method described does not accurately quantitate the total numbers of WBCs present in filtered RBCs, it may provide useful information on qualitative aspects of WBC subpopulations.

The prevalence of human T-cell leukaemia virus-I and -II infection was studied in a cohort of 346 intravenous and nonintravenous drug users in Amsterdam. Three participants (0.86%) had antibodies to HTLV-I by two commercially available HTLV-I enzyme immunoassays (EIA). Infection in these three subjects was confirmed by radioimmunoprecipitation assay. In the immunoblot study, only two of the three subjects were considered positive, since the serum of the third subject had antibodies to p24 only. By means of the polymerase chain reaction two participants (male intravenous drug users infected with human immunodeficiency virus; HIV) appeared to be infected with HTLV-I and one subject (a male nonintravenous drug user from Surinam) with HTLV-II. It is concluded that HTLV-I and HTLV-II circulate sporadically among drug users in Amsterdam and that risky injecting behaviour, which led to an HIV epidemic among intravenous drug users, has not led so far to an appreciable transmission of the other retroviruses among this group.

No abstract