Introduction

FMS-like tyrosine kinase 3 (FLT3) is the most commonly mutated gene in acute myeloid leukemia (AML). About 30 % of AML patients harbor a FLT3 mutation [1••], either in the form of an in-frame internal tandem duplication, usually in the juxtamembrane (JM) domain of FLT3 (FLT3-ITD), or as a point mutation in the tyrosine kinase domain (FLT3 TKD). While the clinical significance of FLT3 TKD mutations in AML remains controversial [2, 3], patients with FLT3-ITD mutant AML have aggressive disease characterized by early relapse and decreased survival [2, 4–6] compared to those with FLT3 wild type (FLT3 WT) AML. Studies suggest that use of allogeneic stem cell transplant in first remission can improve survival [7, 8, 9•, 10•, 11•]; however, a large number of patients are either not suitable candidates for this procedure or relapse prior to transplant. There has been significant interest in targeting the FLT3 receptor tyrosine kinase with small molecule inhibitors for many years. Unfortunately, first-generation FLT3 inhibitors were hindered by lack of selectivity and potency, resulting in unimpressive clinical activity as monotherapy. Preclinically, quizartinib (AC220), a second-generation FLT3 tyrosine kinase inhibitor (TKI), has been shown to be a highly selective inhibitor of FLT3 signaling for both FLT3-ITD mutant and wild type receptors [12]. Clinically, as a single agent it has achieved high response rates in patients with relapsed or refractory AML and has rekindled enthusiasm for FLT3 inhibition as a therapeutic approach in patients with FLT3-ITD mutant AML [13••, 14••, 15••], though response duration has been limited by the development of secondary resistance-conferring kinase domain (KD) mutations [16••]. While the early clinical success of quizartinib in AML is promising, the optimal strategy for the use of FLT3 inhibitors in AML remains unclear. Here we will review a number of FLT3 inhibitors that have been evaluated or are undergoing evaluation in clinical trials and discuss challenges underlying use of these agents (Tables 1, 2 and 3).

Table 1 Comparison of properties of FLT3 inhibitors tested in AML clinical trials
Full size table
Table 2 Selected trials of type I FLT3 inhibitors
Full size table
Table 3 Selected trials of type II FLT3 inhibitors
Full size table

FLT3 Function and Signaling

FLT3 is a receptor tyrosine kinase that is normally expressed on immature hematopoietic cells and functions in the development of both stem cells and the immune system [17]. Murine models have shown that the targeted disruption of Flt3 results in adult mice with deficiencies in primitive B lymphoid progenitor cells. Additionally, stem cells lacking Flt3 have a decreased ability to reconstitute T-cells and cells of the myeloid lineage [18]. The wild type FLT3 receptor is a monomeric transmembrane protein, which contains five extracellular immunoglobulin-like domains, a transmembrane domain, a JM domain and a split intracellular kinase domain. FLT3 is activated when its ligand, FLT3 ligand (FL) binds to the extracellular domain promoting dimerization [19]. Subsequent phosphorylation of the kinase activates downstream signaling pathways, including Ras/MAPK, PI3K/Akt and STAT5 signaling.

FLT3 is the most frequently mutated gene in AML, with an estimated 30 % of AML patients harboring FLT3 mutations [1••, 20]. Approximately 20–25 % of AML patients harbor FLT3-ITD [21] mutations, which usually occur as in-frame duplications involving the FLT3 JM domain. Normally the JM domain exerts a negative regulatory function over the kinase domain [22]; however, with FLT3-ITD mutations this autoregulatory function is lost resulting in constitutive kinase activation. The added length conferred by the ITD mutation places tyrosines 589 and 591 in contact with aspartic acid 829 in the N-terminal of the activation loop, ultimately catalyzing phosphate transfer [23]. Recently, it was shown that about one-third of FLT3-ITD mutations are a result of an insertion outside of the JM domain, within the β1 sheet or the β2 sheet of the tyrosine kinase domain 1 (TKD1) [24, 25, 26•]. Another 5–7 % of AML patients have point mutations in the FLT3 kinase domain (FLT3 TKD) which result in single amino acid substitutions, usually within the activation loop of the FLT3 kinase domain at aspartate 835 [27, 28] and less commonly at tyrosine 842 or isoleucine 836 [28]. These FLT3 TKD mutations cause constitutive kinase activation by favoring the active kinase conformation. Recently, activating mutations at the TKD1 residue asparagine 676 have also been identified [29•].

Though both FLT3-ITD and FLT3 TKD mutations cause ligand-independent kinase activation, in vitro studies have identified differential autophosphorylation [30] and downstream signaling patterns for FLT3-ITD [31] compared to FLT3 TKD and native FLT3, particularly preferential activation of STAT5 [32] by FLT3-ITD, as well as increased proliferation and clonogenic growth potential [32]. These signaling differences may in part be the result of aberrant trafficking of FLT3-ITD mutant receptors resulting in prolonged retention in the endoplasmic reticulum and increased exposure to intracellular substrates such as STAT5 [33]. In a murine bone marrow transduction and transplantation model, FLT3 D835Y yields an oligoclonal lymphoid disorder with longer disease latency distinct from the myeloproliferative neoplasm (MPN) observed with FLT3-ITD [34]. Recently, Bailey et al. confirmed in a knock-in mouse model where Flt3 is expressed under control of the endogenous promoter that disease induced by FLT3 D835Y is phenotypically distinct from disease induced by FLT3-ITD. In contrast to FLT3-ITD mice, which exclusively develop MPN [35], FLT3 D835Y mice develop a MPN with longer latency and broader range of disease phenotypes, including extranodal B-cell masses and T-cell rich B-cell lymphoma, suggesting an increased permissiveness of FLT3 D835Y for lymphoid neoplasms.

FLT3 Mutations in AML: Clinical Features, Management and Outcomes with Standard Approaches

The presence of a FLT3-ITD mutation in young patients with normal cytogenetic AML confers a poor prognosis [5]. This effect is lost in older patients which may be because this group carries such a poor prognosis that it is difficult to detect any additional negative effect [36]. Despite clarity on the prognostic effect of FLT3-ITD mutations in young patients with normal cytogenetic AML, this is not the case in patients with FLT3 TKD mutations. Patients with FLT3-ITD mutant AML often present with marked leukocytosis and high percentages of peripheral blood and bone marrow blasts [37]. A number of studies have shown that remission rates with standard induction regimens are similar for patients with FLT3-ITD+ AML and patients with unmutated FLT3 [2, 4–6], but remission duration is short, with a median time to relapse of only 6–7 months [37]. Furthermore, after relapse, the prognosis for patients with FLT3-ITD+ AML is worse than those with FLT3 WT AML, with a median survival of less than 5 months [38, 39•]. As such, patients with FLT3-ITD mutant AML represent a particularly challenging population to treat. Given the aggressive nature of FLT3-ITD mutant AML, many studies have investigated the role of allogeneic stem cell transplant as consolidation therapy in first remission. While none of these studies are prospective in nature, most studies suggest that patients with FLT3-ITD mutant AML consolidated with allogeneic transplant in first remission have better outcomes than those patients consolidated with chemotherapy alone [7, 8, 9•, 10•, 11•, 40, 41•]. Most experts currently recommend consolidation with allogeneic transplant in first remission if possible based on donor availability and host suitability. Unfortunately, not all patients are transplant candidates and a large proportion of patients relapse prior to transplant.

FLT3 Inhibitors: Early Agents and Limitations

For patients with FLT3-ITD mutant AML, small molecule FLT3 inhibitors have been an attractive investigational treatment alternative for a number of years. A number of small molecule TKIs with activity against FLT3, including midostaurin (PKC412), lestaurtinib (CEP-701), tandutinib (MLN-518), KW2449, sunitinib (SU11248), and sorafenib (BAY 43–9006) were initially evaluated in patients with FLT3-ITD mutant AML, both as single agents and in combination with intensive chemotherapy. Despite initial optimism, response rates and response duration were limited in early phase I and phase II trials in patients with relapsed or refractory AML, at least in part due to lack of sustained FLT3 inhibition in vivo. The cytotoxic activity of these agents appears to be dependent on significant and sustained FLT3 inhibition. Unfortunately, because these agents are multikinase inhibitors, a lack of potency meant that many early agents were limited by the inability to achieve effective FLT3 inhibition without dose-limiting toxicity. Additionally, clinical efficacy of these agents was limited by high degrees of protein binding and/or short plasma half-lives in patients. Moreover, it was unclear if rare clinical responses to these early inhibitors were achieved through inhibition of FLT3 signaling or through off-target effects.

FLT3 Inhibitors: The Agents

Midostaurin (PKC-412)

Midostaurin (PKC412) is a small molecule TKI with broad activity against cKIT and PDGFR in addition to FLT3. Unfortunately, single-agent trials with midostaurin in patients with relapsed or refractory AML were unimpressive, resulting largely in clearance of peripheral myeloid blasts and rare bone marrow responses [42, 43]. Despite, the lack of single-agent efficacy, a phase Ib study provided some optimism for the benefit of midostaurin combined with chemotherapy in previously untreated patients. In this study, midostaurin was combined with standard induction and consolidation chemotherapy in patients with newly diagnosed AML. Of the 69 patients in the trial, 19 harbored a FLT3 mutation, 13 of which were evaluable for a response. In the FLT3 mutated patients, the complete remission (CR) rate was 92 % with a 1-year overall survival (OS) of 85 % and a 2-year OS of 62 % [44]. Based on these results, a multi-center, randomized, phase III trial of midostaurin or placebo combined with induction chemotherapy using daunorubicin/cytarabine and consolidation chemotherapy using high-dose cytarabine in patients with newly diagnosed FLT3 mutated AML was initiated. Results of this trial are expected to be reported in 2014.

Lestaurtinib (CEP-701)

Like midostaurin, lestaurtinib is a small molecule TKI with broad activity against a number of receptor tyrosine kinases including JAK2, VEGFR and TrkA in addition to FLT3. In a phase I/II trial, lestaurtinib was evaluated as a single agent in patients with relapsed or refractory FLT3 mutant AML. Five of the 17 patients were noted to have clinical evidence of biological activity and measurable clinical response, which, similar to midostaurin, manifested mostly as short-lived reductions in peripheral and bone marrow blasts [45]. A randomized phase III trial of lestaurtinib in combination with chemotherapy in FLT3 mutant AML patients in first relapse showed no benefit to the addition of lestaurtinib, though correlative studies suggested that only a small number of lestaurtinib-treated patients achieved target FLT3 inhibition [39•].

Tandutinib (MLN-518)

Tandutinib, is a small molecule TKI with activity against c-KIT and PDGFR in addition to FLT3. In a phase I trial in AML, single-agent tandutinib showed little clinical activity. Eight of 40 patients evaluated in this study had FLT3-ITD mutations. No CRs or PRs were observed at either of the dose levels tested [46]. A subsequent phase II trial showed similar results [47]. When combined with induction and consolidation chemotherapy in patients with previously untreated AML, the CR rate was 73 %, though only five of the 29 patients had a FLT3-ITD mutation [48].

KW2449

KW-2449 is a small molecule multikinase inhibitor with activity against FLT3, aurora kinase, FGFR-1, and ABL kinase, which was tested in a phase I trial in leukemia and myelodysplastic syndrome using a twice daily dosing schedule ranging from doses of 25 mg to 500 mg per day. In five of 11 patients with FLT3-ITD mutations, peripheral blast reductions of greater than 50 % were observed at day 14, without evidence of bone marrow responses. Careful correlative studies revealed that clinical response was limited by a high degree of plasma protein binding and short inhibitor half-life resulting in effective target inhibition in patients for only 4–6 h per dose [49].

Sunitinib (SU11248)

Sunitinib is a multikinase inhibitor which carries FDA indications for the treatment of a number of solid tumors including advanced clear cell renal cell carcinoma, gastrointestinal stromal tumors and advanced pancreatic neuroendocrine tumors. An early clinical trial assessing single-agent efficacy of sunitinib in patients with relapsed or refractory AML showed only modest activity with partial responses of short duration [50]. However, despite the lack of single-agent activity, a phase I/II study assessing the efficacy of sunitinib in combination with induction and consolidation chemotherapy in older patients with AML and FLT3 activating mutations showed some promise. CR rates were 53 % (8/15) and 71 % (5/7) for patients with FLT3-ITD and FLT3 TKD mutations, respectively. The 13 patients who achieved CR went on to be consolidated with high dose cytarabine and 7/13 received sunitinib maintenance. The median overall survival in this study was 18.8 months [51].

Sorafenib (BAY 43–9006)

Like sunitinib, sorafenib is a multikinase inhibitor that is FDA approved for the treatment of patients with advanced clear cell renal cell carcinoma, advanced hepatocellular carcinoma and radioactive iodine refractory thyroid cancer. The antileukemic effect of sorafenib was assessed in a phase I study of patients with relapsed or refractory AML. In this study, six of 16 patients had FLT3-ITD mutations; all FLT3-ITD+ patients achieved a reduction in peripheral and bone marrow myeloid blasts after sorafenib treatment. In patients who were FLT3 WT there was limited activity, with only three of seven patients responding. No activity was observed in patients with FLT3 TKD mutations [52]. Despite the relatively modest activity observed in this initial study, multiple studies have since reported the activity of sorafenib in FLT3-ITD+ AML. In a recent report of 65 patients with FLT3-ITD mutant AML, clinical efficacy of sorafenib monotherapy was seen in patients relapsed after both chemotherapy and allogeneic stem cell transplant [53••]. A number of additional studies have clinically assessed the antileukemic efficacy of sorafenib in a variety of settings [54–57], including a recent phase II study of sorafenib in combination with 5-azacitadine in relapsed/refractory FLT3-ITD mutant AML, which demonstrated a response rate of 46 %, mostly consisting of CR or CR with incomplete count recovery [58••]. In contrast, another recent randomized placebo-controlled study of sorafenib in combination with induction and consolidation chemotherapy in an elderly patient population showed no benefit and increased toxicity [59••]. In this trial, however, patients were not selected for FLT3 mutation status and only 14/95 placebo-treated and 15/102 sorafenib-treated patients in this study were FLT3-ITD+. Given the observed anecdotal clinical efficacy of sorafenib monotherapy in FLT3-ITD mutant AML patients, especially in the post-allogeneic transplant setting, the use of sorafenib in this patient population has become increasingly common despite the fact that this agent is not FDA approved for the indication. Underlying reasons for the particular efficacy of FLT3 inhibitors in the post-allogeneic transplant setting remain unclear, though may be attributable to selection of a FLT3-mutant clone at relapse post-transplant or due to unknown effects of FLT3 inhibitors on the graft immune system.

Quizartinib (AC-220)

Quizartinib (AC220) is a second generation FLT3 inhibitor, which unlike earlier FLT3 inhibitors, has resulted in a high rate of response in patients with relapsed or refractory AML. Studies have shown that quizartinib, in contrast to early FLT3 inhibitors, is both highly potent and selective for WT FLT3 and FLT3-ITD [12]. The safety and tolerability of quizartinib was evaluated in a phase I dose escalation trial of unselected patients with relapsed or refractory AML. The most commonly observed treatment-related adverse events were prolongation in QTc, gastrointestinal symptoms, and myelosuppression. The trial included 76 patients: 17 were FLT3-ITD+, 37 were FLT3-ITD negative and 22 were indeterminate for FLT3 mutation status. Nine of 17 patients (53 %) who were FLT3-ITD mutant+ responded versus five of 37 (14 %) FLT3-ITD negative patients. The median response duration was 13.3 weeks, with responses noted at doses as low as 18 mg/day [60•]. The efficacy of single-agent quizartinib was subsequently assessed in a phase II study in patients with relapsed or refractory AML. This study consisted of two treatment cohorts, one consisting of older patients (>60 years of age) after one line of chemotherapy and one consisting of patients >18 years of age relapsed or refractory to second-line therapy or transplant. The primary end point was the composite CR rate (CRc), which was defined as combination of CR, CR with incomplete platelet recovery (CRp) and CR with incomplete hematologic recovery (CRi). Overall, the CRc rate (mostly CRi or CRp) in patient with FLT3-ITD mutant AML in both cohorts was an impressive 44–54 % [14••, 15••] with continuous daily dosing of 90 mg/day (females) or 135 mg/day (males); a notable ~30 % CRc rate was also observed in FLT3-ITD negative patients in both cohorts. It is also significant to note that patients who had relapsed after allogeneic stem cell HSCT had the highest remission rates and that a large proportion (35 %) of the FLT3-ITD+ patients in the younger cohort were bridged to transplant, suggesting a benefit of quizartinib in the peri-transplant setting. As was seen in the previous phase I study, the most common adverse events were gastrointestinal effects, reversible QTc prolongation and myelosuppression

Because of high rates of QTc prolongation and myelosuppression in the initial phase II study, a second randomized phase II study explored lower doses of quizartinib (30 or 60 mg) in a similar patient population. Again, an ~50 % rate of CRc was observed at both dose levels and was associated with a decreased rate of QTc prolongation; however, most remissions still occurred in the setting of incomplete neutrophil or platelet recovery [13••]. It is also notable that instead of the hypocellular response associated with chemotherapy, in some patients response to quizartinib appears to be associated with a syndrome of terminal myeloid differentiation resulting in marrow hypercellularity associated with a neutrophil surge and inflammatory tissue infiltrates [61••], further suggesting that remissions on FLT3 kinase inhibitor treatment may appear different from those achieved with standard chemotherapy. The lack of traditional CR with full neutrophil and/or platelet count recovery observed in these studies has sparked controversy as to whether the non-conventional endpoint of CRc is associated with true clinical benefit and/or prolongation of overall survival compared to standard chemotherapy. To answer this question, a randomized phase III clinical trial of quizartinib monotherapy versus salvage chemotherapy in FLT3-ITD+ AML patients in first relapse is expected to begin in 2014.

The single-agent clinical efficacy of quizartinib in patients with relapsed or refractory FLT3-ITD mutant AML has led to a number of studies assessing its activity in different clinical settings. Two recent phase I clinical trials assessed the safety and tolerability of quizartinib in combination with induction and consolidation chemotherapy in younger [62•] and older [63•] adult patients with newly diagnosed AML. Another trial assessed the safety of quizartinib in combination with etoposide and cytarabine in a pediatric population with relapsed AML [64•]. In all three trials, quizartinib was safely administered in combination with chemotherapy and further combination studies are being planned. Currently, a phase I trial assessing the efficacy of quizartinib as maintenance therapy after allogeneic stem cell transplant is also recruiting patients.

Even with the increased response rates observed with quizartinib, the lack of clinical activity of early FLT3 inhibitors called into question whether the activity of quizartinib could be definitively attributed to inhibition of FLT3 signaling. The development of drug-resistant KD mutations on the FLT3-ITD allele at the time of relapse in eight of eight patients who achieved remission on quizartinib monotherapy confirmed that these initial responses were achieved via FLT3 inhibition and that relapse was mediated by reactivation of FLT3 signaling [16••]. The most common resistance mutations occurred at the activation loop residue D835. Mutations at the FLT3 gatekeeper residue F691 were less commonly found. These mutations have also been associated with acquired clinical resistance to sorafenib [65••, 66•]. Taken together, these observations establish that inhibition of FLT3 signaling holds therapeutic promise, provided that sufficiently potent and prolonged inhibition can be achieved.

Novel FLT3 Inhibitors

The emergence of TKI-resistant FLT3 KD mutations at the time of relapse on clinically active FLT3 inhibitors such as quizartinib has emphasized the need for additional potent and selective FLT3 inhibitors that are also invulnerable to resistance-causing FLT3 KD mutations. To this end, a number of FLT3 inhibitors that have demonstrated pre-clinical activity against quizartinib-resistant mutations have undergone or are currently undergoing evaluation in early phase clinical trials in FLT3 mutant AML.

The less clinically common FLT3-ITD F691L “gatekeeper” mutation appears to cause quizartinib resistance by disrupting a π-π ring stacking interaction between the aromatic side chain of the phenylalanine residue and the benzo-imidazol-thiazol moiety of quizartinib [16••]. Mutations at this residue may be overcome by structurally diverse inhibitors less dependent on this interaction. The ABL/FLT3 inhibitor ponatinib, which is FDA approved for the treatment of TKI-resistant CML, achieved bone marrow remission in 2/7 TKI naïve FLT3-ITD+ AML patients treated on aphase I study [67•] and has demonstrated pre-clinical activity against the quizartinib-resistant F691L mutation [68•]. However, ponatinib’s clinical activity in AML and its ability to suppress the F691L mutation needs to be assessed in larger clinical trial experience. It is unclear if the dose-limiting cardiovascular toxicity recently described in CML patients will impact the ability to administer ponatinib at clinically efficacious does in a FLT3 mutant AML population. PLX3397 is an inhibitor of KIT, FMS and FLT3 that has also demonstrated equipotent preclinical activity against the FLT3-ITD F691L mutation and is currently being evaluated in a phase I/II trial in FLT3-ITD+ AML. Results from this trial have not yet been reported.

To date, mutations at the FLT3 activation loop residue D835 have been the most frequently implicated in resistance to both quizartinib and sorafenib [16••, 65••]. D835 mutations appear to cause resistance to quizartinib by favoring an active kinase conformation, in which the phenylalanine at the base of the kinase activation loop is flipped “in” (“DFG-in”) relative to its conformation in the inactive state (“DFG-out”). Quizartinib is a “type II” inhibitor, which binds to the kinase only in the inactive “DFG-out” conformation; therefore, D835 mutations induce a kinase conformation disadvantageous to quizartinib binding. By the same mechanism, D835 mutations and mutations at other activation loop residues such as Y842 and D839 cause similar resistance to other type II FLT3 inhibitors including sorafenib [16••], ponatinib [68•] and PLX3397 [69•], and pre-clinical mutagenesis studies have identified activation loop mutations as a common cause of resistance to these type II FLT3 inhibitors [16••, 68•, 69•]. The ideal FLT3 inhibitor which would be invulnerable to activation loop D835 mutations would likely be a “type I” inhibitor capable of binding the active, “DFG-in” kinase conformation. While earlier FLT3 inhibitors such as midostaurin and lestaurtinib are indeed type I inhibitors, lack of potency and selectivity limit their clinical efficacy. Crenolanib is a next-generation type I inhibitor that is highly selective for FLT3 and has demonstrated pre-clinical activity against FLT3 D835 mutations both in the presence and absence of an ITD [70•, 71•, 72•]. However, while phase II clinical trials in FLT3 mutant AML are ongoing, data on the clinical efficacy of crenolanib are not yet available.

FLT3 Inhibitors: How Should We Use Them?

While quizartinib has demonstrated high single-agent remission rates in FLT3-ITD+ patients, questions about the optimal use of FLT3 inhibitors in AML remain. Evidence suggests that AML is a polyclonal disease [73•]. While all patients with FLT3-ITD mutations are considered to have aggressive disease with a propensity for early relapse, significant evidence suggests that prognosis is related to the allelic burden of disease, with those patients with high mutant to wild type allele ratio having the worst prognosis [2, 74]. This phenomenon may be a reflection of the polyclonality of AML at the time of diagnosis, including disease cells with and without FLT3 mutations especially as FLT3 mutations are generally thought to occur later in leukemogenesis and not as “founder mutations” likely to exist in all leukemic clones. As the disease progresses, a dominant FLT3 addicted clone can emerge, which may account for the increased allelic burden of disease in many patients at the time of relapse [37]. This suggests that early in the disease course single-agent therapy with a FLT3 inhibitor would only affect a small subset of AML cells and may, therefore, have little clinical utility. While a dominant FLT3-addicted clone does seem to emerge in many patients at the time of relapse, the development of polyclonal drug-resistant FLT3 KD mutations may limit response duration to FLT3 inhibitors used as monotherapy [16••]. Together, the polyclonality of AML and the aggressive nature of relapsed FLT3-ITD mutant AML suggest that the optimal use of FLT3 inhibitors would be in combination with other agents, such as standard chemotherapy, hypomethylating agents or stem cell transplant, with a particular role for maintenance therapy for the suppression of a FLT3-addicted clone.

Based on clinical response rates, FLT3 inhibitors are most appropriately used in the patients with FLT3 mutations, particularly FLT3-ITD mutations. However, the ~30 % response rate observed with quizartinib in FLT3-ITD negative patients [14••, 15••] is intriguing and it is unclear if a subset of these response can be attributed to other FLT3 mutations not detected by clinical testing. The identification of other AML patient populations that might be ultimately responsive to potent FLT3 inhibitors will require careful molecular analysis of non-FLT3 mutant responders.

Conclusions

The clinical activity of early FLT3 inhibitors was limited by a lack of potency, selectivity and favorable pharmacokinetic properties. While quizartinib monotherapy has achieved a high rate of bone marrow remissions with incomplete recovery of blood counts in patients with relapsed and refractory FLT3-ITD mutant AML, the clinical meaningfulness of such responses is still in question. Furthermore, response duration on quizartinib and other FLT3 inhibitors has been limited by the rapid development of secondary drug-resistant FLT3 KD mutations that appear to be an Achilles heel for the majority of FLT3 inhibitors currently in clinical development. Future efforts will need to concentrate on elucidating the ideal strategy for the clinical use of FLT3 inhibitors, particularly in drug combinations or as remission maintenance, as well as identifying potent and selective FLT3 inhibitors with decreased vulnerability to resistance mutations.