HLA‑DR Expression on Leukemic Progenitor Cells before Allogeneic Hematopoietic Cell Transplantation Adds Prognostic Value in the Presence of MRD

Introduction

The prognosis of patients with acute myeloid leukemia (AML) is influenced by multiple factors, including patient-related characteristics, disease specifics, and treatment response. Measurable residual disease (MRD) monitoring during therapy and follow-up offers prognostic information essential for individualized treatment decisions. MRD positivity (MRD^pos) prior to allogeneic hematopoietic cell transplantation (alloHCT) is associated with higher relapse rates. Alongside these factors, immunological interactions also influence the outcome of alloHCT. Some therapeutic interventions such as modifying conditioning intensity, require the prompt availability of the MRD status prior to transplant.

Aims

This study evaluated the prognostic value of a standardized, rapid (less than 5 minutes per sample) multiparametric flow cytometry (MFC) MRD approach (Röhnert, Leukemia 2022) before alloHCT. Additionally, we assessed the impact of HLA‑DR expression before alloHCT on post-transplant outcomes as a potential prognostic biomarker of immune escape.

Methods

Bone marrow (BM) aspirates from AML patients undergoing alloHCT at the University Hospital TU Dresden were analyzed for MRD presence immediately before conditioning therapy. This MFC MRD approach allows the simultaneous MRD detection by the leukemia-associated immunophenotype (LAIP) and the different from normal (DfN) method. Our LAIP-based DfN analysis employs a hierarchical gating strategy with fixed gates, encompassing the core antigens recommended by ELN, including HLA‑DR. The analysis checks simultaneously for 32 aberrant populations (both HLA‑DR^pos and HLA‑DR^neg). The MFC MRD results were not used to guide treatment decisions.

Results

From 2016 to 2022, we consecutively analyzed 185 patients by MFC‑based MRD assessment within a median of 9 days (IQR 7-14 days) before alloHCT. The median follow‑up time for the entire cohort was 32.7 months after alloHCT. Patients with an increased blast percentage (n=69) were included due to the increasing uncertainty about the importance of cytological remission before alloHCT. Before the start of the conditioning, a total of 76% (140/185) of patients were MRD^pos. MRD^pos before alloHCT was significantly correlated with inferior overall survival (OS) and relapse-free survival (RFS). The 2‑year OS rate was significantly lower in the MRD^pos cohort compared to MRD^neg patients (65% vs. 89%, HR 4.3, p=0.002). Of all 185 patients, 116 were in hematologic remission before HCT. In this subgroup, the proportion of MRD^pos patients was significantly lower compared to patients with persistent disease (63% versus 97%, p<0.001); remarkably 3% (2/69) of patients not in hematologic remission, were still MRD^neg. Remarkably, the two patients who were not in remission by cytomorphology but were MRD^neg by MFC did not relapse after alloHCT and showed an OS comparable with the MRD^neg responder group. Among patients in hematologic remission before alloHCT (n=116), MRD^pos patients exhibited higher relapse rates compared to MRD^neg patients (33% vs. 14%, p=0.025), leading to significantly shorter 2-year OS (74% vs. 88%, HR 2.8, p=0.037) and 2‑year RFS (59% vs. 81%, HR 2.1, p=0.048). Multivariable Cox regression models confirmed the prognostic significance of the MRD status. Notably, MRD^pos patients with preserved HLA‑DR expression (n=14) before alloHCT demonstrated a more favorable 2‑year OS (89% vs. 62%, HR 0.3, p=0.08) compared to those with HLA‑DR loss (n=126), with outcomes comparable to MRD^neg patients. Molecular MRD assessment complemented MFC methods, showing concordant results in 75% of patients.

Conclusions

Our rapid semi‑automated method of MRD evaluation by MFC allows a personalized risk prediction in AML patients undergoing alloHCT as a potential base for individual surveillance strategies. The association between pre‑transplant HLA‑DR expression on residual leukemic cells and post‑transplant outcome indicates a possible interaction between phenotype and susceptibility to Graft-versus-Leukemia effect. Further research is warranted to elucidate the mechanistic basis of immune escape and its implications for post‑transplant relapse, guiding the development of novel therapeutic strategies.