ANALYSIS OF INTRACELLULAR TH1 CYTOKINE SECRETION DATA USING PARAMETRIC METHODOLOGY

Flow cytometric methods are widely used for analyzing cytokine release from immune cell populations. However, difficulties are frequently en countered in the analysis of flow cytometric data from intracellular c ytokine staining because the cytokine positive and cytokine negative h istograms overlap considerably. This study compares models containing Gaussian, Giddings, Haarhoff-van der Linde (HVL) and Weibull distribut ions for fitting flow cytometric intracellular cytokine histogram peak s, The results show that now cytometric data for the Th1 cytokines, in terferon-gamma and interleukin-2, in peripheral blood are well describ ed by a model consisting of the sum of two log-normal distributions bu t the other distributions tested also showed satisfactory fits. The mo del-based approach may potentially eliminate the need to use markers d erived from isotype control staining, The results obtained using peak fitting were also compared to the widely practised 99% division line o r marker method, The percent positive calculated using the 99% divisio n line or marker methods correlates poorly with the percent area under the cytokine-positive peak. However, when the model is used to calcul ate a cytokine-positive percentage analogous to that determined by the 99% division line or marker method, the two methods correlate well. ( C) 1998 Wiley-Liss, Inc.