Multiparametric Flow Cytometry-MRD Assay: Lesson from Phase II Trail REL AML 001

Background: Core-binding factor (CBF) AML with RUNX1/RUNX1T1 (AML1/ETO) and CBFB/MYH11 fusion genes is known as a heterogenous group, accounting for about 10% to 15% of AML. In the Phase II Trial REL AML 001 (EudraCT Number 2017-002094-18; ClinicalTrials ID: NCT 03686345) adult CBF leukemia patients treated with a continuation therapy with Midostaurin were included, and the measurable residual disease (MRD) was performed by molecular MRD assessment (qPCR) and multiparametric flow cytometric-MRD (MCF-MRD) assay. The findings of MCF-MRD assay were compared to qPCR. Methods: 31 patients with AML1/ETO (n=15) and CBFB/MYH11 (n=16) were evaluated between December 2018 to July 2022 with both techniques. A total of 145 determinations (n. AML1/ETO = 56 and n. CBFB/MYH11= 89) were performed. MCF panel included 2 10-color tubes: CD15 FITC/ CD56 PE/ CD34 PerCP-cy5.5/ CD117 PE-cy7/ CD7 APC/ CD13 APC-R700/ CD19 APC H7/ HLADR V450/ CD45 V500C/ CD33 BV630, specific for RUNX1/RUNX1T1 cases and CD64 FITC/ CD2 PE/ CD34 PerCP-cy5.5/ CD117 PE-cy7/ CD200 APC/ CD13 APC-R700/ CD14 APC H7/ HLADR V450/ CD45 V500C/ CD33 BV630, specific for CBFB/MYH11 cases. A mean of 1,733,240 CD45 positive events were acquired, with a mean lower limit of detection (LLOD) of 0.001% (ranging 0.01-0.0003%). In 10 cases (6 RUNX1 and 4 CBFB/MYH11) the molecular and phenotypical features at onset were not available. MCF analyses were performed by InfinicytTM software and reference image tool was used wherever possible. Results: We found a concordance of 76% between MCF cases and RUNX1/RUNX1T1 qPCR, and a concordance of 83% between MCF cases and CBFB/MYH11 qPCR (Tables 1a and 1b). In relapsed cases we observed a dynamic and continuous increase of abnormal MCF events, that was not found in cases without clinical relapse although with positive qPCR (Table 2). The HLADR expression evaluated as the geo mean intensity at presentation and/or at relapse resulted higher in RUNX1/RUNX1T1 patients than in CBFB/MYH11 cases (29845 vs 8979, respectively, p=0.0039 Figure 1). Interestingly, we found in a single case with atypical CBFB/MYH11 fusion gene a high value of HLADR geo mean (44522). The down-regulated expression of HLADR in CBFB/MYH11 blasts may be a useful clue to distinguish AML blasts from the scarcely represented normal CD2 positive myeloid precursors, and from the strongly HLADR positive elements that are found in a regenerating bone marrow background. Conclusion: Although MCF-MRD assay showed lower sensitivity than CBF qPCR, MCF offered interesting informations about the dynamics of the abnormal clone size with time, evaluated as the increasing number of MRD clustering events, which may predict an impending relapse.