EVALUATION OF SENSITIVITY IN DNA ANEUPLOIDY DETECTION USING A MATHEMATICAL-MODEL

Detection of DNA aneuploid populations in samples analyzed by flow cytometry generally requires recognition of more than a single G0/G1 peak on the resulting DNA histogram. This assessment is critical in clinical samples if diagnosis or subsequent patient treatment are influenced. In this paper we evaluate three major distribution parameters that contribute to recognition or obscuration of closely spaced peaks on a DNA histogram. These parameters are (1) coefficient of variation (CV), (2) proportion of cells within each peak, and (3) separation of peaks (DNA index, DI). Through the analysis of computer-synthesized DNA histograms, we produced a single chart illustrating the minimum CV required for two peaks (representing DNA diploid and DNA aneuploid cells) in mixtures of various proportions (1:99 to 99:1) and DI values (1.01 to 1.20) to be resolved. This information can be utilized to provide insight into the ability of flow cytometry to resolve a DNA aneuploid population in a histogram. (C) 1994 Wiley-Liss, Inc.