The assessment of measurable residual disease (MRD) in acute myeloid leukemia (AML) by flow cytometry (FC) is challenging and requires full standardization to provide reliable results for patient management. Here is shown the analytical validation of a 10-color panel for AML MRD detection.

6 normal, 50 regenerative and 68MRD positive bone marrow (BM) samples were prepared using bulk lysis for acquisition of at least 1 million cells per tube. The panel composition was addressed to detect leukemia-associated immunophenotypes (LAIPs) and different-than-normal (DfN) immunophenotypes (clonal evolution), as well as leukemia stem cell (LSC) identification, using 5 tubes with 6 backbone markers (CD33APC/CD34PECy5.5/CD38APC-H7, CD45V500/CD117PE-Cy7,HLA-DRV450), combined with CD15/CD13/CD11b/CD7, CD2/CD56/CD123, CD97/CD99/CD54/CD244/CD18, CLL1+TIM3 /CD123/CD96/CD19, CD36/CD300e/CD45RA/CD14/CD64 in FITC/PE/AF700/BV421/BV605, respectively. Fluorescent markers were selected according to minimum spillovers between fluorescence channels and their stability after treatment. This panel was validated by comparing with our established 8-color panel. Samples were acquired in BD FACSCanto II and BD FACSLyric 10, adjusted according to EuroFlow's standard operating procedures. An unstained blank tube was run prior to sample acquisition to assess compensation and carry-over contamination was avoided by cleaning with distilled water between tubes. Infinicyt (Cytognos™) software was used for data analysis. Statistical analysis: R software (v 4.1.2) and Excel were used for statistical analysis, considering a p value ≤ 0.05 as significant. Bland-Altman plots and Pearson's correlation coefficient were used to assess agreement values between 10- and 8-color panels. Levey-Jennings plots, Ranksum tests, and chi-square tests were used to assess variances on the median fluorescence intensity (MFI) of the markers in CD34+ progenitor cells in normal and regenerated BM and in blast cells. The chi-square test was also used to compare the expressions of the monoclonal markers at diagnosis and during treatment. The linearity of the tests to determine the lower limit of quantification (LLOQ) was assessed using Pearson's correlation coefficient.

There was high accuracy and reproducibility comparing the performance of both panels and both flow cytometers (R2 > 0.96). The analytical sensitivity was determined by the limit of blanc (LOB) calculated using normal BM samples not stained with CD34 and CD117, limit of detection (LOD) with LOB+1.645SD; and lower limit of quantification LLOQ, using sequential dilutions of AML samples. Linearity was assessed using the results of blast cell percentages obtained from three samples diluted in duplicate, with high correlation between the measurements (R2 = 0.99; p-value < 0.001). The expression patterns of markers were evaluated according to their MFI in CD34+ progenitor cells from 6 normal and 50 regenerative BM samples compared with the MFI in blast cells from 68 MRD+ samples and there was significance difference of most values (p < 0.05). The hemodilution control was done by the percentages of mast cells (CD117+ bright/HLA-DR-) < 0.002% and of mature granulocytes (CD13 and CD11b bright) > 90%.

The results showed a successful process of laboratory validation for AML MRD assessment, as recommended by the international consensuses (ICCS and ELN) and to authorize its use for clinical decision making.