Multicentric Standardization of Minimal/Measurable Residual Disease in Acute Myeloid Leukemia Using Multiparametric Flow Cytometry - a Brazilian Society of Bone Marrow Transplantation (SBTMO) Project

Introduction Measurable Residual Disease (MRD) is a critical tool for assessing treatment response in Acute Myeloid Leukemia (AML). AML MRD negativity after induction and consolidation correlates with improved overall survival (OS) and relapse-free survival (RFS). Conversely, AML MRD positivity both pre- and post-hematopoietic stem cell transplantation (HSCT) is linked to a higher incidence of relapse. One of the available methods for MRD detection is multiparametric flow cytometry (MFC), which is applicable to over 90% of patients. MFC sensitivity is influenced by post-treatment phenotype instability, emergence of new clones, and variability in sample processing, detection strategies and expertise of flow cytometry personal. Material and Methods The project was divided into three phases. The first included an educational program designed based on laboratory experience in AML MRD. This program involved online training on pre-analytical processes using Euroflow SOPs, data analysis of normal/regenerative bone marrow (BM), AML MRD positive samples, and leukemia stem cell (LSC) detection. The second phase, or dry phase, evaluated laboratory performance in data analysis. Participating laboratories analyzed eight de-identified flow cytometry standard (FCS) files using trained gate strategies and reported their results, receiving feedback and reference values for comparison. The final phase, or wet phase, involved laboratories preparing and analyzing four samples using trained SOPs and analysis strategies, submitting de-identified AML MRD FCS files for centralized analysis. Sixteen laboratories from Brazil, with varying levels of AML MRD experience, participated. Eleven used BD-FACSCanto II (8 colors), and five used FACSLyric (10 colors). Samples were prepared using bulk lysis and acquired aiming for one million events per tube. Flow cytometers were adjusted following Euroflow SOPs, with compensation matrices automatically calculated. Laboratory reports included data for comparability, such as MRD and non-leukemic CD34+ myeloid cells percentages, hematopoietic stem cells (HSC), LSC, immunophenotype descriptions, number of analyzed cells, limit of detection (LOD), lower limit of quantification (LLOQ), and hemodilution assessment. Pre-analytical parameters like forward (FSC) and side scatter (SSC) and median fluorescence intensity (MFI) were compared with reference values. Results In the dry phase, 125 results from the eight shared FCS files showed 81% agreement between laboratories and coordinating laboratory, with 19% discordant MRD+/MRD- (6% discordant >1 log). In the wet phase, 80% of 63 sample reports agreed with central analysts' results; 20% MRD+/MRD- disagreement (6% discordant >1 log). Besides the global similarity between phases, 10 (62%) laboratories improved their MRD result agreement between the two phases; nine achieved 100% agreement. LSC detection and quantification varied significantly, with 59% and 73% agreement between laboratory and central analysis in the dry and wet phases, respectively. The intraclass correlation coefficient for MRD results was 95%, and for LSC it was 45% between both central analysts. Normal CD34+ progenitors and HSC frequencies showed 58% in both parameters in the dry phase and 80% and 78% respectively in the wet phase. The number of cells analyzed, LOD and LLOQ and hemodilution control had agreement between laboratory results and central analyzes at more than 80% and 90% in the dry and wet phases respectively. Some pre-analytical parameters showed greater variability than others, such as CD56PE (62%) and CD54AF700 (80%). Discussion This multicenter standardization project aimed to reduce AML MRD result heterogeneity among Brazilian laboratories. The educational program and implementation of SOPs improved the performance of most laboratories. However, four laboratories showed reduced agreement and the main reason was the non-suitable preanalytical processes that hampered MRD interpretation, emphasizing the need for continuous training. This study demonstrated that standardized flow cytometry protocols are reproducible with rigorous SOP implementation. Continuous education and external quality controls are essential for maintaining reliable AML MRD testing. The project's results underscore the need for ongoing training to build expertise and ensure applicable results for clinical decision-making.