Affordable CD4(+) T cell counts by flow cytometry II. The use of fixed whole blood in resource-poor settings

We tested the feasibility and precision of affordable CD4(+) T cell countin g in resource-poor settings using a recently standardised fixative, TransFi x (TM) in whole blood (WB) by flow cytometry (FCM). The precision of the as says was established under optimal conditions for single-platform FCM such as the volumetric CytoronAbsolute and the bead-based FACSCan. Fresh WB samp les from HIV-seropositive and seronegative patients were tested in Tanzania and South Africa, fixed and sent to the UK for reanalysis 7 days later. Co rrelation, bias and limits of agreements were analysed by linear regression and the Bland-Altman test. Absolute CD4(+) T cell counts remained stable f or at least 10 days when TransFix was added to WB in 1:10 dilution at 20-25 degreesC, and for 7 days when added in 1:10 or 1:5 dilution to samples sto red to mimic 'tropical' conditions at 37 degreesC. Higher temperatures such as 42 degreesC were tolerated for only short periods since the recovery ha d decreased to 63% by day 3. The reproducibility of lymphocyte subset analy sis remained unchanged by TransFix with coefficient of variations < 6% for all T cell subsets. Absolute CD4(+) T cell counts and CD4(+) T cell % value s on fixed samples in the UK showed a high correlation with the results usi ng fresh samples in Tanzania (r = 0.993 and 0.969, respectively) and with t he samples handled in Johannesburg (r = 0.991 and 0.981) with minimal bias. Primary CD4 gating using only a single CD4 antibody also remained accurate in TransFixed samples (r = 0.999). Thus, TransFix permits optimal fixation and transport of WB samples in the developing world for FCM to local regio nal laboratories and for quality assurance in international centres. When u sed together with inexpensive primary CD4 gating, TransFix will allow relia ble and affordable CD4(+) T cell counting by FCM in resource-poor settings. (C) 2001 Elsevier Science B.V. All rights reserved.