The effect of number of histogram events on reproducibility and variation of flow cytometric proliferation measurement.

Abstract

Sources of variation in synthesis phase fraction (SPF) calculation were studied, and the number of histogram events was found to be an important quality control consideration. Six cell cycle models (CCMs) for histograms composed of 1,000 to 20,000 events were compared. All CCMs were cytometer based, or available in Multicycle (MC) software. The experiment consisted of five consecutive acquisitions, on the same day, of the same propidium iodide (PI) stained sample of T24 human cell line, at each of nine "landmarks" between 1,000 and 20,000 events. The authors found (1) all CCMs evaluated required > or = 5,000 events for accurate, reproducible SPF; and (2) in the 5,000-20,000 event range the MC models provided the most accurate, reproducible SPF values. Therefore, histogram-dependent curve fitting models may enhance clinical applications of FCM proliferation measurements. The authors conclude that histogram rejection criteria for S-phase analysis should be established, and that two-color multiparametric DNA analysis "live" gating with tissue specific markers may assure acquisition of sufficient events for accurate SPF.