Impact of hemodilution on flow cytometry based measurable residual disease assessment in acute myeloid leukemia

Jesse M. Tettero,Maaike E. Heidinga,Tim R. Mocking,Glenn Fransen,Angèle Kelder,Willemijn J. Scholten,Alexander N. Snel,Lok Lam Ngai,Costa Bachas,Arjan A. van de Loosdrecht,Gert J. Ossenkoppele,David C. de Leeuw,Jacqueline Cloos,Jeroen J. W. M. Janssen
DOI: https://doi.org/10.1038/s41375-024-02158-1
2024-01-26
Leukemia
Abstract:Measurable residual disease (MRD) measured in the bone marrow (BM) of acute myeloid leukemia (AML) patients after induction chemotherapy is an established prognostic factor. Hemodilution, stemming from peripheral blood (PB) mixing within BM during aspiration, can yield false-negative MRD results. We prospectively examined hemodilution by measuring MRD in BM aspirates obtained from three consecutive 2 mL pulls, along with PB samples. Our results demonstrated a significant decrease in MRD percentages between the first and second pulls ( P = 0.025) and between the second and third pulls ( P = 0.025), highlighting the impact of hemodilution. Initially, 39% of MRD levels (18/46 leukemia-associated immunophenotypes) exceeded the 0.1% cut-off, decreasing to 30% (14/46) in the third pull. Additionally, we assessed the performance of six published methods and parameters for distinguishing BM from PB samples, addressing or compensating for hemodilution. The most promising results relied on the percentages of CD16dim granulocytic population (scarce in BM) and CD117 high mast cells (exclusive to BM). Our findings highlight the importance of estimating hemodilution in MRD assessment to qualify MRD results, particularly near the common 0.1% cut-off. To avoid false-negative results by hemodilution, it is essential to collect high-quality BM aspirations and preferably utilizing the initial pull for MRD testing.
oncology,hematology
What problem does this paper attempt to address?