Abstract Activating mutations in FMS-like tyrosine kinase 3 (FLT3) are among the most frequently identified mutations in acute myeloid leukemia (AML). Most of these mutations occur as internal tandem duplications (FLT3-ITD), which are associated with high rates of relapse. FLT3 inhibitors (FLT3i) are now incorporated as standard of care therapy FLT3-ITD AML. Responses to FLT3i monotherapy in relapsed AML are often characterized by differentiation of leukemic cells with persistent FLT3-ITD clonal hematopoeisis (McMahon et al., Blood Advances, 2019). We recently demonstrated downregulation of the histone methyltransferase, EZH2, in response to FLT3i as a mechanism of drug-induced myeloid differentiation in FLT3-ITD AML (Sung et al., Leukemia, 2024). EZH1, a close homolog of EZH2, was stable or increased with FLT3i, suggesting a potential compensatory effect that could contribute to resistance. Combined FLT3 and EZH1/2 inhibition using gilteritinib and valemetostat increased the anti-leukemic effect in vitro in primary FLT3-ITD AML and in vivo in genetically engineered mouse models (GEMMs). Mice treated with the combination exhibited increased myeloid differentiation and reduced leukemic burden. For further pre-clinical development, we aimed to demonstrate efficacy of gilteritinib and valemetostat in patient-derived xenograft (PDX) models of FLT3-ITD AML. Two primary FLT3-ITD AML specimens were transplanted into NOD.Cg-Prkdcscid Il2rgtm1Wjl/SzJ (NSG) mice and monitored for human cell engraftment by serial bone marrow assessment. After confirmation of engraftment of 10%, animals were treated for 4 weeks with vehicle, gilteritinib (60 mg/kg oral thrice weekly), valemetostat (100 mg/kg oral daily), or the combination. Spleen and bone marrow were harvested after 4 weeks on treatment for analysis of leukemic burden and differentiation states using a 27-color spectral flow cytometry panel for simultaneous assessment of stem/progenitors and mature myeloid species. Both models had initial mutational profiles of FLT3-ITD, NPM1 W288Cfs, and DNMT3A (R484Afs*8 or R882H) and demonstrated significant reductions in spleen and marrow leukemic involvement with the combination versus vehicle control at 4 weeks. In the bone marrow for PDX #1, gilteritinib enriched for phenotypic leukemia stem cells (LSCs), presumably through clearance of more mature cell types. In contrast, valemetostat alone reduced LSCs though overall demonstrated a modest increase in percent of leukemic blasts. This is consistent with previous reports of valemetostat increasing LSC cycling (Fujita et al., Leukemia,2018). The combination of gilteritinib plus valemetostat reduced total blasts and LSCs while increasing neutrophils and monocytes. This is suggestive of enhanced differentiation effects that were previously demonstrated in our GEMM experiments. Similar effects were seen in the spleen for PDX #1, but surprisingly engraftment in the spleen was nearly completely from a FLT3 wild-type and TP53 R156Afs*14 NF1 I679Dfs*21 mutated AML. Despite the absence of a FLT3 mutation, gilteritinib and valemetostat therapy reduced total leukemic burden and LSCs in the spleen. In PDX #2, gilteritinib alone was highly effective, but the addition of valemetostat again reduced the number of phenotypic LSCs. Single cell RNA-Seq analysis of PDX #2 demonstrated qualitative differences in differentiation state with the combination treatment. Gilteritinib alone led to an expansion of the monocytic compartment whereas the combination exhibited more neutrophil-primed precursors. Longer term treatment is likely needed to evaluate differences in terminal maturation for this model. Additional analyses are ongoing and will be updated at the Annual Meeting. Conclusions: Our findings validate the FLT3i and EZH1/2 combination as a novel approach to therapy in FLT3-ITD AML and suggest a potential effect of this combination through inhibition of wild-type FLT3. Early assessment suggests that valmetostat decreases LSC frequency and that combination therapy with FLT3i may improve responses. Additional studies in FLT3 wild-type AML are planned.
Eunice S. Wang (Mon,) studied this question.