Validation of a mathematical procedure for computer analysis of flow cytometric DNA data in human tumors.

Abstract

The determination of tritiated thymidine labeling index and the percentage of cells with S phase DNA content was performed on cell suspensions obtained from 69 patients with non-Hodgkin's lymphoma. The distributions of cells in the cell cycle by computer analysis of flow cytometric data were obtained by two mathematical procedures: the widely adopted Fried model and a new one proposed by Bruni et al. A significant agreement was observed by checking the Spearman index (rs) between the percentages of cells in the different cell cycle phases (G0/1, rs = 0.76; S, rs = 0.60; and G2 + M, rs = 0.43; p less than 0.001) determined by the two procedures. Similarly, a good correlation was observed between the labeling index (LI) and the S phase values obtained by the Fried (rs = 0.45, p less than 0.001) and Bruni (rs = 0.69, p less than 0.001) models, but with a higher agreement for the latter one. The S phase by the Bruni model was also superior in predicting LI: in fact, by employing the S cutoff value of 12%, a better agreement between low LI and low S phase or high LI and high S phase was observed with the Bruni procedure (90%) than with the Fried model (72%). Finally, the analysis of the prognostic significance of the different kinetic variables confirmed the prognostic relevance of LI at any time; the S phase percentage as determined by Bruni et al. was discriminant of survival only at shorter times, and no prognostic significance could be ascribed to S phase according to the Fried procedure.