Measurement of LSC17 score under various sample collection and processing conditions. (A) Plot showing LSC17 scores measured in 4 PB samples processed either fresh (blue) or after being viably frozen and thawed (red). The dotted line represents the median LSC17 score (0.51). (B) Plot showing correlation between LSC17 score measured in Ficoll-separated and unseparated (non-Ficoll) BM samples from 41 patients with AML. (C) Plot showing LSC17 scores measured by the LDT in PB (blue) and BM (red) samples collected from 10 patients with AML at diagnosis. The dotted line represents the median LSC17 score. Patient samples are shown in order left to right by increasing PB blast percentage. Triangles, Ficoll-separated samples; circles, unseparated (non-Ficoll) samples. (D) Plot showing the difference (Δ) between paired PB and BM LSC17 scores compared with PB blast percentage. The horizontal dotted line indicates the extent of technical variation (Δ0.1 = 2 standard deviations); the vertical red line indicates the threshold (20% PB blasts) below which BM should be used as the sample source for the assay. (E) Plot showing LSC17 scores measured in 13 BM samples processed 8, 24, and/or 48 hours after collection. (F) Plot showing correlation between LSC17 scores measured in BM samples collected in EDTA tubes processed 8 hours after collection vs samples collected in PAXgene tubes and processed up to 5 days after collection. R, Spearman correlation coefficient.