Pretransplant NPM1-MRD Levels Predict Outcome after Allogeneic Stem Cell Transplantation in Adult Patients with Acute Myeloid Leukemia

Background: Allogeneic hematopoietic stem cell transplantation (SCT) is considered to be the treatment strategy with the highest anti-leukemic efficacy for acute myeloid leukemia (AML) patients in morphological complete remission (CR). The prognostic impact of pretransplant minimal residual disease (MRD) levels measured by real-time quantitative polymerase chain reaction (RT-qPCR) on outcome is currently unclear.

Aims: To evaluate the prognostic impact of pretransplant NPM1 MRD levels in correlation to clinical characteristics and genetic abnormalities assessed at initial diagnosis in a cohort of adult AML patients receiving SCT.

Methods: We retrospectively studied 238 AML patients (median age at time of SCT, 53.5 years; range, 17-73 years) who received a SCT between 2005 and 2013 at the University of Heidelberg. Patients with acute promyelocytic leukemia were excluded from the analysis. Based on material availability, the mutational status of NPM1 and FLT3-ITD were analyzed in 208 and 215 of the patients, respectively. Sixty-seven of the 208 patients (32%) were NPM1-mutated; of those, 28 patients had a concurrent FLT3-ITD (42%). Source of donor was matched-related in 20, matched-unrelated in 45 and haplo-identical in 2 of the 67 patients, respectively. The majority of patients (n=61) received reduced intensity conditioning. SCT was performed in first CR in 31, in second CR in 20 and in 16 refractory patients, respectively. Pretransplant MRD could be measured in 53 of the 67 patients with NPM1-mutated AML (79%), and in 39 of the 51 NPM1-mutated patients who were in morphological CR at SCT (76%). MRD negativity prior to SCT was defined as less than 100 copies of mutated NPM1/ 10⁴ABL copies; the sensitivity level was 10^-5-6.

Results: Overall survival (OS) for NPM1-mutated patients transplanted in morphological CR was significantly longer as compared to patients with active disease (5-year estimates of OS: 63% vs. 38%; p=0.004) irrespective of first or second CR (p=0.74). There was no difference in outcome in CR patients with or without MRD measurement (OS: p=0.84; relapse-free survival (RFS): p=0.29). However, when focusing on patients with MRD measurement, MRD-positive (MRDpos) patients (n=20) had a higher incidence of concurrent FLT3-ITD mutations (p=0.008) as compared to MRD-negative (MRDneg) patients (n=19). There was a highly significant difference in RFS and OS between MRDpos and MRDneg patients: estimated 5-year RFS and OS were each 40% vs. 89% (p=0.001, each). In a multivariate Cox model for RFS, the hazard ratio (HR) for pretransplant MRD was 13.20 (95% confidence interval: 3.22-77.70; p<0.001), whereas FLT3-ITD measured at diagnosis was not significant (p=0.94). OS of patients receiving SCT in active disease was 38% at five years and was comparable to that observed in patients with MRDpos disease before transplant (40%; p=0.42).

Conclusions: Pretransplant NPM1 MRD positivity was a significant predictor of poor outcome in patients with NPM1-mutated AML, heralding a prognosis not better than that of patients transplanted with refractory disease.