Clinical use of flow cytometric (FCM) DNA analysis requires effective quality controls. Thirty‐two laboratories with various degrees of FCM experience participated in the first phase of a quality control program organized by the Association Française de Cytométrie. All received diskettes containing ten list‐mode files and ten histogram files that were derived from FCM analysis of various unfixed tumor specimens. A total of 610 responses on DNA ploidy and cell cycle were obtained with three different DNA analysis softwares: CellFit used by (44% of responses), MultiCycle (44%), and ModFit (12%). After statistical analysis, 31% of the responses were excluded from the final analysis for precise reasons. The groups were too small to carry out a valid analysis of the slight differences in the percentage of cells in the DNA synthesis phase (S%) between CellFit and MultiCycle. To estimate the influence of gating on the final cell‐cycle results, five of the histogram files were derived from corresponding list‐mode files, but the participating laboratories were unaware of this. A good correlation (r = 0.98) was obtained for S% values in the five paired files.
The fact that 31% of the responses had to be excluded clearly reflects inadequate training in the use of these analysis softwares and, in some cases, a failure to grasp the biological meaning of the results. In contrast, the laboratories fulfilling consensus recommendations obtained remarkably homogeneous results, showing that standardization is feasible. © 1996 Wiley‐Liss, Inc.