PS995 IMPACT OF DNMT3A MUTATIONAL STATUS IN MOLECULAR MEASURABLE RESIDUAL DISEASE CLEARANCE IN ACUTE MYELOID LEUKEMIA WITH MUTATED NPM1

Abstract

Background:Acute myeloid leukemia (AML) with mutated NPM1 (NPM1mut) is a separated category in the latest WHO classification. Persistent detection of NPM1 mutations after therapy is a strong predictor of relapse and measurable residual disease (MRD) monitoring of NPM1 is recommended in the latest European LeukemiaNet (ELN) MRD assessment guidelines. DNMT3A is the third most frequent mutated gene in AML and has been associated with mutations in NPM1 and FLT3. Prognostic significance of mutated DNMT3A (DNMT3Amut) in AML has been controversial.Aims:To explore the differences in molecular MRD clearance in AML with mutated NPM1 according to DNMT3A mutational status.Methods:A total of 67 patients with de novo AML and NPM1mut were analysed from CETLAM database (median age 53 years (19–72); male:female 31:36 (0.86), median WBC 20.5x109/L (1.3–384), median bone marrow blasts 69%). All patients were included in the AML‐2003 (n = 9) and AML‐2012 (n = 54) CETLAM intensive treatment protocols, 63 patients achieved complete remission (CR) with 1 or 2 cycles of induction intensive chemotherapy (n = 55 vs 8) and followed 1 to 3 consolidation cycles. Notably, since 2012 patients with NPM1mut and FLT3‐ITD low ratio are not considered for transplant in first CR. Bone marrow (BM) samples from diagnosis were studied for DNMT3A mutations as previously described and NPM1 molecular MRD monitoring was performed at BM by quantitative RT‐PCR (sensitivity 10–4 to 10–6) at specific timepoints (diagnose, after induction and following each cycle of consolidation (C1, C2, C3)). The relationship between DNMT3A mutations and various patient characteristics was determined by the Fisher exact test for the categorical variables. Kaplan‐Meier method was used to estimate the distribution of OS.Results:DNMT3A mutations were detected in 31 patients, the majority were found in codon #R882 or corresponded to DNA insertion/deletions. FLT3‐ITD was observed in 28 patients (17 low wt/ITD ratio (<0.5) vs 11 high ratio). No association was found between DNMT3Amut and FLT3‐ITD mutational status (p = 0,133). In the whole series, bone marrow NPM1 quantitative MRD clearance was found significant when considering MRD log reduction (defined as the log10 reduction between BM NPM1 baseline level and NPM1 after each cycle of intensive chemotherapy), as well as NPM1 total transcripts (> or ≤ than median BM NPM1 transcripts after each cycle). Thus, after C1 patients with ≥3log reduction presented a significantly better outcome than those with <3log (5yr OS 82 ± 6,6% vs 54 ± 15% p = 0,013). In this setting, we contrasted DNMT3A mutational status to the log reduction after C1, and we observed a significant association between the presence of DNMT3Amut and a worse NPM1 clearance (p = 0,001) (Figure 1B). Similar findings of the association between DNMT3A and NPM1 MRD clearance were shown with log reduction after induction (p = 0,011) and NPM1 absolute transcript reduction after induction (p < 0,001) and C1 (p = 0,011).Summary/Conclusion:In this study, bone marrow NPM1 MRD log and absolute transcript reduction quantified at specific time points associates to a survival benefit. Mutations in DNMT3A are significantly related to a worse clearance of NPM1 MRD. Future studies with larger cohorts are needed to clarify prognoses of AML patients that harbour NPM1, FLT3‐ITD and mutated DNMT3A. image