Quantitative Impact of Mutated NPM1 MRD in Patients with AML Undergoing Allogeneic Stem Cell Transplantation

Introduction: In NPM1-mutated (NPM1_mut) acute myeloid leukemia (AML) the detection of measurable residual disease (MRD) using quantitative PCR methods for NPM1_mut allows for risk stratification and therapy adjustments during disease course. Even though it is well understood that a persistent NPM1_mut burden at allogeneic stem cell transplantation (HSCT) is associated with relapse and shorter survival, the quantitative impact of NPM1_mut at HSCT has not been investigated in detail.

Methods: We analyzed 81 NPM1_mut AML patients (pts) who received their first allogeneic peripheral blood HSCT at a median age of 61 (range 33-76) years in morphologic complete remission (CR) or CR with incomplete peripheral recovery (CRi). Conditioning regimens were non-myeloablative (70%), reduced-intensity (11%), or myeloablative (19%) according to EBMT guidelines. At diagnosis, cytogenetics and co-mutational status of FLT3, IDH1, IDH2, DNMT3A, and CEBPA were assessed. NPM1 MRD up to 28 days prior to HSCT in CR/CRi was measured by digital droplet PCR using a competitive probe approach, and results were normalized to ABL1 (limit of detection 0.001%NPM1_mut/ABL1). For patients with a positive NPM1 MRD test result, receiver operator characteristics (ROC) curves were used to analyze distinct NPM1/ABL1 MRD threshold. Presence of chronic graft-versus-host disease (GvHD) was assessed according to Glucksberg criteria. Median follow up after HSCT was 4.4 years.

Results: ROC curves identified 0.01% NPM1_mut/ABL1 as the most informative cut-off to define MRD^pos and MRD^negpts in both, bone marrow and blood, and irrespective of conditioning intensity. Using this cut-off 31 (38%) pts were NPM1_mut MRD^pos at HSCT and had a significantly higher cumulative incidence of relapse (CIR, P<.001, after 2 years 4% vs 66%) and shorter overall survival (OS, P<.001, after 2 years 77% vs 38%). Patients transplanted NPM1_mut MRD^pos had a 20.8-times higher risk of relapse than those transplanted MRD^neg. While there were no differences regarding the gene mutation profile at diagnosis in pts with or without NPM1_mut MRD at HSCT (Figure 1A), pts with NPM1_mut MRD at HSCT had a significantly higher diagnostic WBC (49.6 vs 17.9; P=.03), LDH (12.9 vs 6.7; P=.02), and more CD34+ cells in bone marrow at diagnosis (1.4% vs 0.5%; P=.02).

We observed a negative correlation of the NPM1_mut /ABL1 MRD levels before HSCT and the time to relapse; the higher the NPM1_mut MRD burden at HSCT the earlier the relapse occurred (R= -0.5, P=.03). This correlation was strongest in pts with a FLT3-ITD present at diagnosis (R= -0.82, P=.002). When utilizing two cut-offs for low (≥ 0.01% but < 1%) and high (≥ 1%) NPM1_mut /ABL1 MRD levels there was a significantly shorter time to relapse in relapsing AML pts with high NPM1_mut MRD levels (median time to relapse: 73 days) before HSCT compared to pts with low MRD levels (median time to relapse: 149 days). In addition, compared to patients with low NPM1_mut MRD levels, pts with high NPM1_mut MRD levels showed a trend for shorter OS (P=.10; OS after 2 years: 77% for NPM1_mut MRD^negpatients vs 48% for low NPM1_mut MRD^pos patients vs 21% for high NPM1_mut MRD^pos patients; Figure 1B), indicating that pts transplanted with higher NPM1_mut MRD may be harder to salvage. Presence of a FLT3-ITD and age at HSCT did not alter the negative impact on outcomes of low or high NPM1_mut MRD at HSCT.

We also evaluated the prognostic impact of the presence of chronic GvHD as a surrogate marker for graft-versus-leukemia effects with respect to the NPM1_mut MRD status at HSCT. In a landmark analysis for pts surviving longer than 100 days, NPM1 MRD^neg (< 0.01% NPM1_mut/ABL1) pts had the lowest CIR (4% after 2 years), followed by NPM1_mut MRD^pos pts developing a chronic GvHD (25% after 2 years) and by NPM1 MRD^pos pts who did not develop a chronic GvHD and had a very high CIR (92% after 2 years, P<.001).

Conclusion: AML patients with > 0.01% NPM1_mut/ABL1 MRD before allogeneic HSCT had a significant higher risk of relapse and shorter OS. High WBC, LDH, and CD34+ cells in bone marrow at diagnosis associated with persisting NPM1_mut MRD at HSCT. The higher the NPM1_mut MRD burden at HSCT, the earlier the relapse, and the lower the chances of successful salvage therapy after relapse. Development of a chronic GvHD after HSCT was associated with a reduced risk of relapse in NPM1_mut MRD^pos pts, indicating that bolstering immunologic effects post-HSCT for NPM1_mut MRD^pos AML pts may be a promising option to improve outcomes.

Schwind:Novartis: Honoraria. Herling:Abbvie: Honoraria, Research Funding; EDO-Mundipharma: Honoraria, Research Funding; Janpix: Honoraria, Research Funding; Jazz: Honoraria, Research Funding; Novartis: Honoraria, Research Funding; Roche: Honoraria, Research Funding; Takeda: Honoraria, Research Funding. Merz:BMS Celgene: Honoraria; Janssen: Honoraria. Metzeler:Daiichi Sankyo: Honoraria; Pfizer: Consultancy; Jazz Pharmaceuticals: Consultancy; Novartis: Consultancy; Celgene/BMS: Consultancy, Honoraria, Research Funding; Curis: Research Funding; Astellas: Honoraria; AbbVie: Honoraria. Franke:Novartis: Honoraria; Jazz Pharmaceuticals: Honoraria; Pfizer: Honoraria; Gilead: Other: Travel support; Incyte: Honoraria; BMS: Honoraria; Takeda: Other: Travel support. Vucinic:Novartis, Gilead Kite, Takeda, MSD, BMS Celgene, Abbvie, Amgen: Honoraria; MSD, BMS Celgene, Novartis, Gilead Kite, Takeda: Consultancy, Membership on an entity's Board of Directors or advisory committees; Sobi, BMS Celgene: Other: travel, accommodations, expenses. Jentzsch:Novartis: Honoraria.