Patients diagnosed with AML with NPM1mutation (NPM1mut AML) included in the European LeukemiaNet favorable genetic risk category (ELNfav, i.e., without FLT3-ITD or with a low allelic burden FLT3-ITD comutation [FLT3-ITD/FLT3wt <0.5; FLT3-ITDLOW]) do not benefit from an allogeneic stem cell transplantation (alloSCT) in first complete remission (CR1). However, a significant proportion of these patients fail to frontline chemotherapy and require salvage therapy. Persistence or detection of MRD after post-CR treatment is associated with a high relapse risk and worse prognosis. With this background, the cooperative group CETLAM proposed an early therapeutic intervention in CETLAM-2012 protocol for patients with ELNfavNPM1mut AML patients not achieving a sustained MRD clearance after consolidation therapy, defined as molecular failure (MF). Herein we analyzed the outcome and predictive risk factors of MF.


All patients diagnosed with ELNfav NPM1mut AML treated according to CETLAM-12 protocol who achieved CR after 1 or 2 courses of induction chemotherapy were included in the study. Intended post-CR therapy consisted of 3 courses of high-dose cytarabine chemotherapy (HDAC). MRD was assessed in bone marrow samples after each chemotherapy course and thereafter, at regular 3-month interval during 3 years, in reference labs from CETLAM group following standard NPM1-mutation specific RQ-PCR. MF was defined as failure to achieve a molecular response (molCR) after consolidation therapy (i.e., NPM1mut/ABL·100 ratio > 0.05) or MRD reappearance after molCR. All MFs were confirmed in a second sample collected at least 4 weeks apart from previous sample. For patients who present a MF, alloSCT was recommended, without a predefined indication of debulking previous salvage chemotherapy.


Out of 145 patients with ELNfav NPM1mut AML (75 M/70 F; median age, 56, 18-74), 132 (91%) achieved CR after 1 (n=124) or 2 courses (n=8). After a median follow-up of 25 months, 2-year overall survival (OS), event-free survival (EFS), leukemia-free survival (LFS) of the entire cohort were 79.9% (±3.6), 71.8% (±4) and 79.4% (±3.8), respectively. Among patients with available complete MRD information (n=89), 33 patients developed an hematological and/or molecular failure (Mol/Hem failure), resulting into a Mol/Hem failure-free survival at 2 years of 62% (±5.6%); median time from CR to Mol/Hem failures was 8.2 months (2-43). Fourteen patients presented with an overt hematological relapse (hREL), preceded by a detectable MF in 8. Overall, 14 received salvage therapy in morphological CR, MRD(+) status (MF), and 14 For hREL. Among MF, 5 patients received HDAC-type salvage treatment (followed by alloSCT in 3) and 9 proceeded directly to alloSCT. In 5 additional MF patients with MRD persistence after consolidation, subsequent MRD monitoring showed decreasing MRD levels until complete clearance (n=4) or intermitent detection (n=1) and have not received further therapy. After salvage therapy, 12/14 (86%) patients with MF achieved molCR, and 10/14 (71%) treated for hREL achieved CR2 (MRD negative in 7, 50%), followed by alloSCT in 9. OS after Mol/Hem failure was 64.8% (± 8.4) at 2 years, 88.8 (±7.5%) in patients treated for MRD(+)-status (MF) and 32.1% (±13.6%) in patients treated with overt hREL (p<0.001; figure 1).

Potential predictive factors of Mol/Hem failure were investigated. Interestingly, a ratio of NPM1mut/ABL*100 ≥1 after induction allowed distinction of two patient subpopulation with strikingly different Mol/Hem failure risk: Mol/Hem failure-free survival at 2 years of 84±6% vs. 33±10% in patients with a lower (<1) and higher (³1) tumor burden (p<0.001; figure 2). Remarkably, concurrence of FLT3-ITDLOWdid not correlate with outcome or Mol/Hem failure risk.


Despite a significant proportion of Mol/Hem failures, NPM1mut AML patients allocated in the ELN favorable risk group presented a favorable overall outcome. NPM1mut-based MRD surveillance is able to anticipate most hematological relapses, and a MRD-driven early intervention policy, at time of MF, allowed a successful rescue of a significant proportion of patients. Moreover, an early measure of residual tumor burden, after induction therapy, might identify those patients with a high risk of molecular or hematological subsequent failure, allowing the potential implementation of preemptive intensification strategies.


No relevant conflicts of interest to declare.

Author notes


Asterisk with author names denotes non-ASH members.

Sign in via your Institution