Systematic misestimation of cell subpopulations by flow cytometry: a mathematical analysis

Various sources of variability in flow cytometric determination of cell concentration have previously been investigated with respect to andrologic applications. Although common aspects related to the variation between samples, variation between operators, and accuracy have been extensively studied, specific sources of false-count estimation have found less attention. In particular, a major and well-recognized source of misestimation of cell counts (i.e., contamination of the sample by non-sperm particles) has not to date been characterized in detail. We show here by means of original mathematical research that not only the cell counts but also the percentages of cells expressing different fluorescence patterns are affected by the presence of alien particles often neglected in studies involving flow cytometric characterization. We demonstrate that there is a systematic overestimation in the proportion of unstained (viable) cells detected by flow cytometry in cases where the non-sperm particles are not excluded from analysis by additional identification other than light-scatter characteristics. Moreover, we provide an exact mathematical estimate for the magnitude of this overestimation, and we discuss the consequences for diagnostic applications and studies on sperm physiology, specifically for studies on sperm capacitation and evaluation of cryopreserved semen. Finally, equations are derived for the correction of the flow cytometric values for use in practical applications.