Measurable Residual Disease (MRD) by Flow Cytometry in Adult B-Acute Lymphoblastic Leukaemia (B-ALL) and Acute Myeloid Leukaemia (AML): Correlation with Molecular MRD Testing and Clinical Outcome at One Year

SIMPLE SUMMARY: Measurable residual disease monitoring is an important prognostic tool in haematological malignancies commonly performed by two modalities: flow cytometry and molecular methods. In this paper, we studied consecutive adult participants requiring flow cytometric measurable residual disease monitoring. This is one of the largest prospective Australian studies to date, providing a unique insight into the Australian context. We described five distinctive patterns associated with disease relapse and survival and also investigated correlation with molecular methods. Our results provide additional evidence that the correlation between molecular and flow cytometric methods is moderate in B-lymphoblastic leukaemia and poor in acute myeloid leukaemia. There was a strong association between flow cytometry results and relapse in acute myeloid leukaemia but less so for B-lymphoblastic leukaemia. Our novel data indicate that the pattern of change in measurable residual disease over time was associated with the risk of relapse, particularly in acute myeloid leukaemia and highlight the divergent ways measurable residual disease testing can be employed across different leukaemias. ABSTRACT: Measurable residual disease (MRD) detected by flow cytometry (FC) is well established in paediatric B- lymphoblastic leukaemia (B-ALL) and adult chronic lymphocytic leukaemia (CLL), but its utility in adult B-ALL and adult acute myeloid leukaemia (AML) is less clear. In this prospective MRD study, one of the largest in Australia to date, we examined consecutive bone marrow aspirates from adult participants with B-ALL (n = 47) and AML (n = 87) sent for FC-MRD testing at a quaternary referral hospital in Sydney. FC-MRD results were correlated to corresponding Mol-MRD testing where available and clinical outcomes at three-month intervals over 1 year. B-ALL showed a moderate positive correlation (r(s) = 0.401, p < 0.001), while there was no correlation between FC-MRD and Mol-MRD for AML (r(s) = 0.13, p = 0.237). Five FC-MRD patterns were identified which had significant associations with relapse (X(2)(4) = 31.17(4), p > 0.001) and survival (X(2)(4) = 13.67, p = 0.008) in AML, but not in B-ALL. The three-month MRD results were also strongly associated with survival in AML, while the association in B-ALL was less evident. There was a moderate correlation between FC-MRD and Mol-MRD in B-ALL but not AML. The association of FC-MRD with relapse and survival was stronger in AML than in B-ALL. Overall, these findings suggest divergent utilities of FC-MRD in AML and B-ALL.