Indeterminate measurable residual disease by multiparameter flow cytometry is associated with an intermediate risk of clinical relapse in adult patients with acute leukaemia

Measurable residual disease (MRD) is useful for prognostication and for monitoring response to treatment in patients with acute leukaemia. MRD by multiparametric flow cytometry (MFC-MRD) utilises the leukaemia-associated immunophenotype (LAIP) and difference from normal (DfN) strategies to identify the leukaemic clone. Difficulties arise when the LAIP overlaps with normal regeneration, there is clonal evolution, or when the abnormal clone population is exceptionally small e.g., <0.01% of CD45⁺ cells. Such cases are reported as ‘indeterminate’; however, there is little international consensus on this reporting. The relationship between clinical outcomes and indeterminate MFC-MRD is unknown. Here we determine the rate of indeterminate MFC-MRD reporting, its relationship to concurrent molecular MRD results when available, and to clinical outcomes to 12 months. We performed an internal audit of all adult testing for MFC-MRD between January and December 2021. A total of 153 consecutive patients with a diagnosis of acute leukaemia were included. Successive MFC-MRD results and clinical outcomes were recorded over a 12-month period from time of inclusion into the study. In total, 460 MFC-MRD tests from 153 patients were reviewed and 73 (16%) MFC-MRD tests from 54 (35%) patients were reported as indeterminate. The majority (70%) were at low levels between 0.01–0.1% of CD45⁺ cells. Compared to patients with a negative result, acute myeloid leukaemia (AML) was more frequent in patients who had an indeterminate MFC-MRD (70% vs 36%), and B-cell acute lymphoblastic leukaemia was less common (20% vs 55%). In patients with indeterminate MFC-MRD results, one-third had received either chemotherapy or allogeneic haemopoietic stem cell transplant (aHSCT) within the preceding 3 months. Agreement between MFC and molecular MRD testing was low. Patients with indeterminate MFC-MRD had leukaemia relapse rates below patients with a positive MFC-MRD, but greater than those with negative MFC-MRD (positive 33% vs indeterminate 21% vs negative 8%, p = 0.038). Overall, these findings indicate that indeterminate MFC-MRD results are more common in adults with AML and also in those who have received chemotherapy or aHSCT within the previous 3 months. We report for the first time that indeterminate MFC-MRD is a finding of potential clinical significance, which associates with a numerically higher median relapse rate within 12 months when compared to a negative MFC-MRD result.