FLT3-ITD MUTATION PROMOTES LEUKEMIC DIFFERENTIATION TOWARDS AN IMMATURE CYTOTOXIC STAGE

FLT3 mutations drive leukemogenesis and poor prognosis in acute myeloid leukemia (AML). Despite targeted therapies, high relapse rates persist, likely due to FLT3-ITD–mediated stem cell survival and immune evasion. This study aimed to investigate how FLT3-ITD burden and clonal dominance influence immune status and treatment response. NK and T cells were assessed by flow cytometry in a Flt3-ITD knock-in mouse model. Bone marrow from 148 de novo AML patients (normal karyotype, known FLT3/NPM1 status) was evaluated for leukemia arrest stage and NK subsets. Drug sensitivity was analyzed using the Beat AML 2.0 dataset. In the murine model, the Flt3 mutation was associated with reduction in NK cell (CD45hiCD19-CD3-NK1.1+) content characterized by predominance of immature phenotypes (CD27+CD11b-) and a reduction in cytotoxic subpopulations (CD27-CD11b+). These results were corroborated by an extended phenotypic analysis using CD122, NK1.1, CD49b, and NKp46 expression to identify four NK cell maturation stages. Among these, a reduction in the frequency of more mature subsets (NKII to NKIV) was observed in Flt3-mutated homozygous mice, whose NK cells exhibited an increased expression of DNAM-1 and TIM-3. In the T cell compartment, an increase in short-lived effector cells (CD127-KLRG1+) was observed among CD4⁺ T cells specifically in homozygous animals. In the CD8⁺ T cell population, homozygous animals showed an increased frequency of central memory (CD62L+CD44+) and memory precursor effector cells (CD127+KLRG1-), accompanied by elevated expression of activation and inhibitory receptors, including NKG2D, TIM-3, and CTLA-4. Furthermore, γδ T cells were expanded in homozygous mutant animals compared to wild-type mice. The influence of Flt3 mutation content on the immune profile prompted us to investigate NK cells in AML samples across different stages of leukemia arrest, which revealed that FLT3mut/NPM1wt patients (n = 36) were predominantly associated with CMP-L (CD34+/-CD117+CD13+CD33+/-HLA-DR+MPO+) and GMP-L (CD34+/-CD117+/-CD13+/- CD33+HLA-DR+MPO+) leukemias, whereas FLT3wt/NPM1mut patients (n = 80) predominantly presented GP-L (CD34-CD117+/-CD13+CD33+HLA-DR-MPO+) leukemias. Double-mutant patients (FLT3mut/NPM1mut, n = 32) exhibited a similar profile to FLT3wt/NPM1mut patients, with a predominance of MP-L and GP-L stages. As compared to other maturation stages of leukemia arrest, GMP-L leukemias were associated with disease persistence after the first induction therapy and a trend toward earlier relapse, although impact on overall or relapse-free survival was not observed. In agreement, in silico analysis using the Beat AML 2.0 dataset indicated higher resistance to cytarabine, midostaurin, and azacitidine in patients with GMP-L FLT3mut/NPM1wt cells. No differences were observed in the frequencies of CD56+ NK cells, CD56bright, or CD56dim subsets among the FLT3/NPM1 profile. However, in FLT3-mutated patients, high AR showed a trend toward fewer CD56dim and more CD56bright NK cells compared to low AR. Regarding leukemic maturation, a trend toward increased CD56bright and decreased CD56dim populations was observed in differentiated leukemias (GMP-L, MP-L, and GP-L). Collectively, our findings indicate that FLT3-ITD promotes immune evasion by disrupting the maturation and activation of cytotoxic immune cells. Immature leukemic subsets may contribute to this disruption, potentially leading to resistance and poor outcomes.