Clinical and Biological Characterization of Children with FLT3ITD Mutated Acute Myeloid Leukemia (AML): A Report from the AIEOP AML-2002 Study Group

Purpose. While most of the recurrent molecular marker of AML have been already discovered and used for risk stratification in clinical protocols, post-treatment monitoring of these abnormalities can be useful in the clinical management of patients. In particular, monitoring of minimal residual disease (MRD), largely used in acute lymphoblastic leukemia, is gaining popularity also in AML, in the perspective of optimizing risk stratification of patients in terms of subsequent clinical relapse. While MRD monitoring in AML is mainly based on flow-cytometry approaches, molecular MRD measurements are not currently routinely used for taking clinical decision. Here we examine if the molecular MRD and Allelic Ratio (AR) levels monitoring may influence the survival of FLT3 internal tandem duplication (ITD)-mutated AML patients.

Patients and methods. We retrospectively analyzed 507 children with de novo AML for FLT3ITD mutation by RT-PCR. Mutation was sequenced and the AR was calculated by Genescan. Bone marrow samples after induction treatment were analyzed for MRD levels by Real-Time PCR. We correlated these parameters with both patient event-free survival (EFS) and gene expression profile (GEP) findings.

Results. 54/507 patients (10.6%) harboured FLT3ITD mutation. AR was calculated both at cDNA and DNA levels showing a reliable correlation (R=0.68), even if only AR measured on cDNA was found to be a significant poor prognostic feature. This latter observation supports the concept that expression of the mutation is more important than genetic bulk architecture at diagnosis. Patients with high AR showed a significant worse EFS as compared to those with low AR (19.2% for AR>0.51, vs 63.5% for AR<0.51), mainly because of a failure to achieve complete remission. Molecular monitoring of MRD showed that a low MRD reduction (<2 logs) after I induction course was an independent prognostic factor predicting both a high risk of incidence of relapse (cumulative incidence of relapse, CIR=68.7% vs 24.3, p <0.05), and a lower EFS (22.2% vs 59.4%, p<0.05). After the second induction course, a low MRD reduction also predicted a worse EFS. Through GEP analysis, we showed that high ITD-AR patients had a different transcriptional profile that those with low ITD-AR. Gene ontology revealed that patients with high AR had a negative regulation of genes involved in H3K27 tri-methylation tightly associated with inactivation of gene promoters as compared to those with low ITD-AR, suggesting that methylation might play a crucial role in this subgroup of FLT3ITD patients. Furthermore, we found that the cyclinA1 was significantly overexpressed. We hypothesised that cyclinA1-CDK2 complex is driving B-MYB phosphorylation, which, in turn, over-activates its own expression and sustains the B-MYB transcriptional oncogenic program in FLT3ITD patients with high AR. Support to this hypothesis is given by the observation that B-MYB signature was found to discriminate patients with low and high ITD-AR. GEP was also performed by comparing poor or good responders patients, showing that the former have a significant enrichment in stemness, in resistance to drugs and histone acetylation pathways.

Conclusion. Our results show that ITD-AR and MRD status are important independent prognostic factors for the management of of FLT3ITD patients, to be taken into consideration in planning the post-induction treatment. The high expression of the FLT3ITD mutation and the persistence of the mutated clone confer an aberrant hyperactivation of the downstream oncogenic FLT3 pathway influencing the outcome within these AML. Transcriptional profiles opens for further consideration of epigenetic targeting for FLT3ITD patients.