A noncanonical FLT3 gatekeeper mutation disrupts gilteritinib binding and confers resistance

The recent FDA approval of the FLT3 inhibitor, gilteritinib, for AML represents a major breakthrough for treatment of FLT3 mutated AML. However, patients only respond to gilteritinib for 6-7 months due to the emergence of drug resistance. Clinical resistance to gilteritinib is often associated with expansion of NRAS mutations, and less commonly via gatekeeper mutations in FLT3, with F691L being the most common. We developed an in vitro model that charts the temporal evolution of resistance to gilteritinib from early microenvironmental-mediated resistance to late intrinsic resistance mutations. Our model system accurately recapitulates the expansion of NRAS mutations and the F691L gatekeeper mutations found in AML patients. As part of this study, we also identified a novel FLT3N701K mutation that also appeared to promote resistance to gilteritinib. Using the Ba/F3 system, we demonstrate that N701K mutations effectively act like a gatekeeper mutation and block gilteritinib from binding to FLT3, thereby promoting resistance. Structural modeling of FLT3 reveals how N701K, and other reported gilteritinib resistance mutations, obstruct the gilteritinib binding pocket on FLT3. Interestingly, FLT3N701K does not block quizartinib binding, suggesting that FLT3N701K mutations are more specific for type 1 FLT3 inhibitors (gilteritinib, midostaurin, and crenolanib). Thus, our data suggests that for the FLT3N701K mutation, switching classes of FLT3 inhibitors may restore clinical response. As the use of gilteritinib expands in the clinic, this information will become critical to define clinical strategies to manage gilteritinib resistance.