Technical Aspects of Flow Cytometry-based Measurable Residual Disease Quantification in Acute Myeloid Leukemia: Experience of the European LeukemiaNet MRD Working Party

Abstract

Measurable residual disease (MRD) quantified by multiparameter flow cytometry (MFC) is a strong and independent prognostic factor in acute myeloid leukemia (AML). However, several technical factors may affect the final read-out of the assay. Experts from the MRD Working Party of the European LeukemiaNet evaluated which aspects are crucial for accurate MFC-MRD measurement. Here, we report on the agreement, obtained via a combination of a cross-sectional questionnaire, live discussions, and a Delphi poll. The recommendations consist of several key issues from bone marrow sampling to final laboratory reporting to ensure quality and reproducibility of results. Furthermore, the experiences were tested by comparing two 8-color MRD panels in multiple laboratories. The results presented here underscore the feasibility and the utility of a harmonized theoretical and practical MFC-MRD assessment and are a next step toward further harmonization.

Figure 1.

Influence of anticoagulant. (A, B, C) CD11b expression on a sample with heparin anticoagulant. (D, E, F) The same sample but now collected in an EDTA tube. One noteworthy influence of EDTA anticoagulants on a sample is the change of expression patterns of specific antigen, such as CD11b. CD11b expression diminishes with EDTA (D and E) compared to heparin coated tubes (A and B).

Figure 2.

Nonviable sample. (A and B) A sample is preferably measured within 72 hours of collection. Prolonged time between collection and measuring will decrease the viability. (C and D) The sample is measured after approximately 240 hours after collection. More cells die causing FSC and SSC to diminish. FSC = forward scatter; SSC = side scatter.

Figure 3.

Effect of lysis solution. When choosing a lysis solution, FSC and SSC properties should be closely monitored to detect the introduction of artifacts. Lysis solutions can decrease the cell size and therefore decrease the FSC resulting in a poor distinction between red blood cells, lymphocytes, and WBC. (A and B) The distinction between WBC (blue), lymphocytes (green), and red/dead cells (red) are clearly seen with ammonium chloride (NH₄Cl) as lysis buffer. (C and D) The same sample is lysed with FACS Lyse sample, but other settings remained the same, resulting in less difference between cell populations. FACS = fluorescence-activated cell sorting; FSC = forward scatter; SSC = side scatter; WBC = white blood cells.

Figure 4.

Increased event rate. The duration of measuring can be reduced by increasing the event rate. (A) The events acquired should clearly be separated from the y-axis (SSC-A). (B) However, increasing the event rate can affect the result by reducing the SSC, resulting in a distorted picture and hamper interpretation of expression patterns (B and C). SSC = side scatter.

Figure 5.

Disturbances in time. (A) Example of similar amount of events acquired over time. (B) Turbulences or disturbances may substantially alter the detection of rare events. Therefore, plotting time against parameters sensitive to fluidic alterations could allow to perform postacquisition corrections during analysis.

Figure 6.

Details of ELN tube and HOVON P1 tube composition. (A) The ELN tube is designed based on the most common LAIPs and expertise of the ELN consortium as reported in Schuurhuis et al. There are 2 clone options for the CD45 antigen. (B) The HOVON P1 tube composition consists of the same fluorochromes but has different clones and antigens in different channels. ELN = The European LeukemiaNet; LAIPs = leukemia-associated immunophenotypes.

Figure 7.

HOVON-P1 and ELN tube comparison. (A) Comparison of log-transformed blast percentage of 62 samples (both diagnose and follow-up) between HOVON-P1 tube (the first tube used from the Heamato Oncology Foundation for Adults in the Netherlands) (x-axis) and the ELN tube (y-axis) with a correlation coefficient of r = 0.99 (P < 0.001). (B) LAIP percentage compared on the same 62 samples as panel A (r = 0.98, P < 0.001). (C) Measurable residual disease (MRD) percentage (y-axis; split into 2 segments on MRD 0.25%) of all 12 follow-up samples per center. A total of 6 follow-up samples are measured with 2 different tubes (HOVON P1 and ELN tube). The centers and patient data are anonymized. The associated samples are next to each other on the x-axis with corresponding numbers. Samples used are at different time points during therapy and with different cytogenetic alterations. Three (HOVON sample 1 and 6; ELN sample 4) of the 12 samples did not have an anonymous MRD result between the 4 centers if the 0.1% cutoff would be used, although not all samples were collected after induction therapy. These differences could be explained when LAIP gates were individually examined. Other 9 samples had identical MRD result (4 MRD+ and 5 MRD–). ELN = The European LeukemiaNet; LAIPs = leukemia-associated immunophenotypes; MRD = measurable residual disease.