Quality control in the flow cytometric measurement of T-lymphocyte subsets: the multicenter AIDS cohort study experience. The Multicenter AIDS Cohort Study Group.

Since 1984, the Multicenter AIDS Cohort Study (MACS) has utilized four flow cytometry laboratories to measure T-lymphocyte subset levels semiannually in a large cohort of homosexual men. This report summarizes the steps taken in the MACS laboratories to attain comparability of lymphocyte subset determinations across the centers and over time. Identical flow cytometers, monoclonal antibodies, and analytic procedures have been used, and over a period of time, the procedure for sample preparation was also standardized. Interlaboratory proficiency testing utilizing identical specimens analyzed in the four laboratories was performed to evaluate the comparability of the data among the laboratories. Our results verify that such testing can identify technical bias in flow cytometric evaluations performed at different laboratories. Temporal laboratory consistency in flow cytometric measurements was evaluated using data from each site's HIV-seronegative homosexual reference group. Both sequential 95% confidence intervals (mean +/- 2 x SEM) and the within-person standard deviations of the immune measurements were considered. Significant variation in CD3, CD4, and CD8 lymphocyte subset percentages over time in the seronegative reference population was observed. Our observations indicate that the lymphocyte subset values of this seronegative group should be used to adjust those obtained on the seropositive study participants during a particular time period, thereby allowing improved discrimination of the effects of HIV on T cells in infected individuals. The data presented are of use for designing epidemiologic and intervention studies in HIV-1-infected individuals, especially for calculating sample sizes. The methods we have used to assess the quality of data in the MACS have general application to quality control programs in flow cytometry laboratories. In particular, comparison of sequential confidence intervals and within-person standard deviations for lymphocyte subset determinations on control populations are essential to a comprehensive proficiency testing program because they permit assessment of consistency within a laboratory over time.