P538: fast-track measurable residual disease detection by multiparametric flow cytometry in patients with acute myeloid leukemia undergoing allogeneic hematopoietic cell transplantation

Background: In acute myeloid leukemia (AML), measurable residual disease (MRD) during therapy and follow-up provides prognostic information. MRD positivity (MRDpos) before allogeneic hematopoietic cell transplantation (alloHCT) is associated with increased relapse rates. Hence, peri-transplant treatment strategies might be stratified by MRD status. However, therapeutic interventions (e.g. conditioning intensity) usually require prompt availability. MRD monitoring is mainly based on molecular genetics and multiparametric flow cytometry (MFC). For 60% of patients (pts) no established molecular marker (mNPM1, CBFB::MYH11, RUNX1::RUNX1T1) is available. Besides MRD, mechanisms of immune escape (e.g. adaptive loss of HLA class II expression) determine the relapse risk. Aims: We evaluated the prognostic value of a standardized, reproducible, and fast turnover MFC MRD approach validated at the end of intensive induction therapy (Röhnert, Leukemia 2022) before alloHCT for pts with and without established molecular MRD marker. Furthermore, we assessed the a priori HLA-DR (an HLA class II molecule) expression on residual leukemic cells as a surrogate for immunogenicity. Methods: Bone marrow (BM) aspirates of pts with AML were analyzed for the presence of MRD immediately before start of conditioning therapy. Results of this MFC MRD approach were available within 24 h, but not used to guide treatment decisions. This approach allows for the simultaneous detection of MRD by the leukemia-associated immunophenotype (LAIP) and the different from normal (DfN) method. Briefly, the LAIP-based DfN analysis is based on a hierarchical gating strategy with fixed gates and encompasses the core antigens recommended by ELN including HLA-DR. Within less than 5 min, the analysis checks for the presence of 32 populations with aberrant expression profile (both HLA-DRpos, and HLA-DRneg). Finally, these aberrant populations are grouped into four aberrant categories (deficiency of CD13 or CD33, cross-lineage expression of CD7 or CD56). Results: Between 2016 and 2022, we identified 179 consecutive pts with available samples and MFC-based MRD assessment within a median of 8 days (range 4 – 27 days) before alloHCT. Of these pts, 139 (78%) had no established molecular MRD marker. The median follow-up was 30.8 months. Overall survival (OS) is reported. Before start of conditioning, 72% pts were MRDpos by MFC. In a univariable Cox regression analysis, the OS of the MRDpos cohort was significantly shorter compared to MRDneg pts (HR 2.7, p=0.011). When restricted to pts without established molecular MRD marker, this assay retained its prognostic significance (HR 2.4, p=0.045). When MRDpos pts were subdivided by the presence of HLA-DR (Figure 1), the absence of any HLA-DR expressing aberrant population (MRDposHLA-DRneg) was associated with a shorter OS compared to MRDneg (HR 2.9, p=0.009). The OS of MRDposHLA-DRpos pts was indistinguishable to MRDneg pts (HR 1.1, p=0.940). These differences were independent of the ELN risk category. Conclusion: The MRD status as determined by our fast-track LAIP-based DfN approach provides relevant prognostic information prior to conditioning for alloHCT for pts without established molecular MRD marker. Interestingly, the presence of aberrant populations without expression of HLA-DR was associated with an unfavorable outcome in the context of alloHCT. Therefore, the LAIP-based DfN approach might be a suitable prognostication tool before starting the preparative regimen for alloHCT and might help to guide potential treatment modifications. Longer follow-up, more pts and an independent cohort are necessary to confirm these findings.Keywords: Minimal residual disease (MRD), MRD, AML, Acute myeloid leukemia