Background: FMS-like tyrosine kinase 3 (FLT3) is a homodimeric transmembrane protein that is expressed on normal hematopoietic stem (HSC) and progenitor cells. FLT3 is also uniformly present on malignant blasts in acute myeloid leukemia (AML) providing a target for immunotherapy with chimeric antigen receptor (CAR)-modified T cells. FLT3 is a driver of leukemia-genesis in AML cases with internal tandem duplication (FLT3-ITD) and other gain-of-function mutations in the intracellular tyrosine kinase domain. Accordingly, the FLT3 inhibitor midostaurin has recently been approved in combination with chemotherapy to treat FLT3-mutated AML. Here, we explored the antileukemia efficacy of FLT3 CAR-T cells against FLT3 wild-type (wt) and FLT3-ITD+ AML, alone and in combination with midostaurin to determine whether the two therapeutic modalities could be used synergistically, particularly in high-risk FLT3-ITD+ AML.
Methods: A FLT3-CAR (BV10 scFv_IgG4 hinge_CD28_CD3ζ_EGFRt) was encoded in the lentiviral vector epHIV7 for gene-transfer into T cells of healthy donors and AML patients (n>6). CAR-T cell cytolytic activity was evaluated in FACS-/luminescence-based assays, cytokine production analyzed by ELISA and proliferation assessed by CFSE dye dilution. Immunodeficient NSG mice were engrafted with MOLM-13/ffLuc AML cells and treated with 5x106 CAR-modified or control T cells (CD4:CD8 ratio = 1:1). MOLM-13 cells were cultured in the presence of 10nM midostaurin and doses titrated up to 50nM to induce resistance (MOLM-13R).
Results: We confirmed specific recognition and high-level cytolytic activity of CD8+ FLT3 CAR-T cells against a panel of AML cell lines that included THP-1 (FLT3 wt), MOLM-13 (FLT3-ITD+/-) and MV4;11 (FLT3-ITD+/+), as well as FLT3-ITD+ primary AML blasts obtained from multiple patients. Both CD8+ and CD4+ FLT3 CAR-T cells produced IFN-γ and IL-2, and underwent proliferation after stimulation with native AML cell lines and primary AML blasts.
We then treated AML cell lines with midostaurin and detected a significant increase in FLT3 expression in MOLM-13 and MV4;11 by flow cytometry (delta MFI = 8986 and 1442, respectively, p<0.05), but not THP-1 cells, indicating that upregulation specifically occurred in FLT3-ITD+ AML. We sought to determine whether higher FLT3 expression on midostaurin-resistant AML target cells translated into superior antileukemia efficacy of FLT3 CAR-T cells. Indeed, we observed superior cytolytic activity, cytokine production and proliferation of FLT3 CAR-T cells after stimulation with MOLM-13R compared to native MOLM-13 cells (n=3 experiments; p<0.05). In contrast to FLT3, expression of CD33 and CD123 as alternative CAR target antigens was decreased on MOLM-13R.
We confirmed upregulation of FLT3 occurred on native MOLM-13 AML cells in vivo during midostaurin treatment in NSG mice, and observed synergistic antileukemia efficacy of FLT3 CAR-T cells and midostaurin in combination therapy. Intriguingly, we detected higher frequencies of CAR-T cells in all mice that received the combination treatment, compared to mice that received FLT3 CAR-T cells alone. At the end of the observation period, all mice in the combination treatment group had achieved complete remission of AML from peripheral blood, bone marrow and spleen.
Importantly, we also show that FLT3 CAR-T cells recognize normal HSCs and interfere with hematopoiesis in colony formation assays in vitro . However, midostaurin treatment did not lead to an increase of FLT3 on normal HSC consistent with their expression of wt FLT3.
Summary: Collectively, our data demonstrate that FLT3 CAR-T cells mediate potent reactivity against FLT3 wt and FLT3-ITD+ AML cells in vitro and in vivo . We further demonstrate that FLT3 CAR-T cells and midostaurin act synergistically, i.e. midostaurin-induced upregulation of FLT3 in FLT3-ITD+ AML cells enhances their recognition and elimination by FLT3 CAR-T cells. Due to recognition of normal HSC, the clinical use of FLT3 CAR-T cells is likely restricted to a defined therapeutic window prior to allogeneic HSC transplantation, followed by CAR-T cell depletion and hematopoietic reconstitution. Our data provide an operational model to augment the antileukemia efficacy of FLT3 CAR-T cells in high-risk FLT3-ITD+ AML patients, and to mitigate the risk for relapse with FLT3-negative AML variants that may otherwise develop under therapeutic pressure.
No relevant conflicts of interest to declare.
Asterisk with author names denotes non-ASH members.