Christine M. McMahon & Alexander E. Perl

ABSTRACT
Introduction: The receptor tyrosine kinase FLT3 is the most commonly mutated gene in acute myeloid leukemia (AML). FLT3-internal tandem duplication mutations are associated with an increased risk of relapse, and a number of small molecule inhibitors of FLT3 have been developed. The highly potent and selective FLT3 kinase inhibitor gilteritinib is the first tyrosine kinase inhibitor approved as monotherapy for the treatment of relapsed and/or refractory FLT3-mutated AML.Areas covered: We review the biology and prognostic significance of FLT3 mutations in AML and discuss the pharmacology, clinical efficacy, and toxicity profile of gilteritinib. We also summarize important differences among the various FLT3 inhibitors that are currently approved or under devel- opment and highlight areas of ongoing research.Expert opinion: Gilteritinib has been shown to improve survival compared to salvage chemotherapy in relapsed and/or refractory FLT3-mutated AML. Gilteritinib is orally available with a favorable toxicity profile and as such is quickly becoming the standard of care for this patient population. Ongoing clinical trials are evaluating gilteritinib in combination with frontline chemotherapy, in combination with other agents such as venetoclax and azacitidine for patients who are ineligible for standard induction therapy, and as a maintenance agent.

KEYWORDS:Gilteritinib; FLT3; tyrosine kinase inhibitor; acute myeloid leukemia

1.Introduction
Acute myeloid leukemia (AML) is an aggressive hematologic malignancy that is diagnosed in approximately 20,000 people per year in the United States at a median age of 68 years [1]. The most common recurrently mutated gene in AML is fms- related tyrosine kinase 3, which encodes the type III receptor tyrosine kinase FLT3 [2]. FLT3 is expressed by normal hemato- poietic stem and early progenitor cells where it has important roles in myeloid and lymphoid differentiation and cell prolif- eration; in AML, FLT3 is highly expressed by a majority of leukemic blasts [3–9]. Internal tandem duplication (ITD) muta- tions of FLT3, which occur in the juxtamembrane domain, have been identified in 20%-23% of AML cases, while point muta- tions in the tyrosine kinase domain (TKD), most commonly at the D835 residue, occur in approximately 7% of AML cases [10–15]. Both FLT3-ITD and TKD mutations lead to the consti- tutive activation of FLT3 and the downstream signaling path- ways STAT5, AKT/mTOR, and Ras/MAPK [10–13]. A simplified schematic of FLT3 structure and signaling is shown in Figure 1 [16,17]. In vitro, FLT3 mutations promote cellular transforma- tion via generation of growth factor independence [18]; in vivo, FLT3-ITD generates a lethal myeloproliferative disorder when retrovirally introduced into murine hematopoietic stem cells [19], with loss of wild type (WT) FLT3 contributing to even greater aggressiveness in a similar FLT3-ITD murine model [20].

Clinically, FLT3-ITD mutations are associated with a poor prognosis, which is largely due to an increased risk of relapse in comparison to FLT3-WT and FLT3-TKD mutant AML [21–25]. FLT3-ITD mutations are predominantly seen and are of parti- cular prognostic importance in patients with intermediate cytogenetic risk, often normal karyotype AML[21–25]. A higher FLT3-ITD mutant to WT allelic ratio (≥ 0.51), as mea- sured by polymerase chain reaction, has been associated with an increased risk of relapse and a worse overall survival (OS), although even low level FLT3-ITD mutations appear to confer a higher risk of relapse in comparison to FLT3-WT AML [23,26,27]. One exception may be those patients with a low FLT3-ITD allelic burden who have a concomitant NPM1 muta- tion, as these patients have been reported to have somewhat more favorable outcomes [28]. Regardless, for patients who experience relapse, the treatment of relapsed and/or refrac- tory FLT3-mutated (FLT3mut+) AML is challenging, as second remissions are RIPA Radioimmunoprecipitation assay uncommonly achieved with salvage che- motherapy and are generally of very short durability [29].FLT3 has been studied as a promising drug target for a number of years, and various small molecule kinase inhibi- tors targeting FLT3 have been evaluated. FLT3 inhibitors that have been developed include the multi-kinase inhibitors les- taurtinib, midostaurin,and sorafenib, and the more potent and selective newer generation FLT3 inhibitors gilteritinib, quizartinib, and crenolanib.

A summary of selected FLT3 inhi- bitors that are currently approved or under development is provided in Table 1. Early trials of the multi-kinase inhibitors were somewhat disappointing as monotherapy with few if any complete remissions observed [30–32]. A phase I/II study of sorafenib combined with azacitidine in relapsed/refractory FLT3-ITD+ AML showed more encouraging results with a rate of complete remission (CR) and CR with incomplete count recovery (CRi) of 43% (16/37). However, the median duration of response was only 2.3 months (range 1– 14.3 months) [33], suggesting much room for improvement.In the frontline setting, however, the addition of midostaurin to standard chemotherapy has been shown to improve out- comes in newly diagnosed patients with FLT3mut+AML [34]. The RATIFY study randomized participants to midostaurin versus placebo in combination with cytarabine-based induction (‘7 + 3’) and consolidation chemotherapy, which was followed by up to 1 year of midostaurin maintenance therapy [34]. The addition of midostaurin improved OS (hazard ratio (HR) 0.78; 1-sided p = 0.009) and event-free survival (HR 0.78; 1-sided p = 0.002) [34]. Based on the results of this study, the U.S. Food and Drug Administration (FDA) approved midostaurin in combi- nation with induction and consolidation for newly diagnosed FLT3mut+ AML in April 2017, and the addition of midostaurin to intensive chemotherapy has now become the standard of care for these patients [35], though notably the US drug approval does not include single agent midostaurin use as maintenance.

Thebroadly selective‘multi-kinase inhibitors’like midos- taurin have been shown to be relatively weak inhibitors of FLT3 in correlative assays from clinical trials[36], which is thought to underlie their limited single-agent activity in relapsed/refractory FLT3mut+ patients. Therefore, second and third generation FLT3 inhibitors were developed specifically for their greater potency and selectivity against FLT3 [37–40]. These newer generation FLT3 inhibitors were designed to maximize anti-leukemic efficacy and hopefully pair this activity with fewer off-target side effects. Gilteritinib and crenolanib are type I kinase inhibitors that are able to inhibit both FLT3- ITD and FLT3-D835 mutations, while quizartinib (like sorafenib) is a type II kinase inhibitor with activity against FLT3-ITD but not FLT3-D835 mutations [38–42]. Recently, the results of ran- domized phase III trials comparing gilteritinib(ADMIRAL, NCT02421939) and quizartinib (QuANTUM-R,NCT02039726) to salvage chemotherapy in patients with relapsed and/or refractory FLT3mut+AML have been reported, with both agents demonstrating a statistically significant improvement in OS compared to salvage chemotherapy [43,44].Preliminary results of early phase trials evaluating crenolanib both as monotherapy and in combination with chemotherapy have also been promising [45–47], and phase 3 testing of crenola- nib in combination regimens for newly diagnosed and relapsed/refractory FLT3mut+ patients has begun.Based on interim results of the ADMIRAL study, gilteritinib was approved by the U.S. FDA in November 2018.

Figure 1. Overview of FLT3 structure and signaling.
FLT3 is a receptor tyrosine kinase. In normal function, binding of FLT3 ligand leads to the dimerization of FLT3 and the activation of signaling. ITD insertions in the juxtamembrane domain and D835 point mutations in the activation loop of the tyrosine kinase domain (TKD) lead to the constitutive activation of FLT3 and the downstream signaling pathways STAT5, AKT/mTOR, and Ras/MAPK [16,17]. The activation of FLT3 signal transduction promotes proliferation, impairs differentiation, and antagonizes apoptosis. FLT3-D835 TKD and FLT3-F691L ‘gatekeeper’ mutations have been identified as mediators of resistance to certain FLT3 kinase inhibitors. treatment of adults with relapsed and/or refractory AML with a FLT3 mutation as detected by an FDA-approved companion diagnostic, the LeukoStrat CDx FLT3 mutation assay. In this article, we review the pharmacology, safety, and clinical devel- opment of gilteritinib. We also discuss gilteritinib in compar- ison to other FLT3 inhibitors that have been developed, highlighting differences among these agents and summariz- ing important areas of ongoing research.

2.Methods
A literature search (English language only) was performed using PubMed through July 2019. References were also obtained by a review of bibliographies and from the conference proceedings of major international hematology meetings.

3.Gilteritinib
3.1.Mechanism of action
Gilteritinib is an oral tyrosine kinase inhibitor that is highly selective for FLT3. It is a type I kinase inhibitor with potent activity against both FLT3-ITD and FLT3-D835 TKD mutations, as well as weak activity against FLT3-F691 gatekeeper mutations [40,41]. In addition to FLT3, gilteritinib shows in vitro inhibition of Axl [41], a receptor tyrosine kinase that is a member of the TYRO, AXL, and MER (TAM) receptor tyrosine kinase family with important roles in cell survival, apoptosis, and chemoresistance [48]. Axl is over- expressed in a number of cancers, including AML, and has been shown to have a potential role in resistance to chemotherapy and to the FLT3 inhibitors quizartinib and midostaurin [49–53].

3.2.Chemistry
The chemical name of gilteritinib is 2-pyrazinecarboxamide, 6-ethyl-3-[[3-methoxy-4-[4-(4- methyl-1-piperazinyl)-1-piperidi- nyl]phenyl] amino]-5-[(tetrahydro-2H-pyran-4-yl)amino]-, (2E)-2-butenedioate (2:1), and the molecular formula is (C29 H44N8O3)2·C4 H4O4 [54].

3.3.Dose
Gilteritinib tablets are supplied in a single strength of 40 mg [54]. Gilteritinib is approved at a dose of 120 mg daily, which is the starting dose that was evaluated in the phase III ADMIRAL trial [43,54]. Of note, escalation of gilteritinib dose to 200 mg daily was allowed for patients enrolled to ADMIRAL who were not in complete remission after the first 28 days of therapy [43], but the U.S. label does not include a recommendation for dose increase in cases of suboptimal response.

3.4.Pharmacokinetics
In phase I testing,plasma concentrations of gilteritinib were
proportional to dose at doses ranging from 20 mg to 450 mg [55].Maximum plasma concentration of gilteritinib was achieved within 4 to 6 hours of an oral dose from a fasted state, with a delay of approximately 2 hours to peak concentra- tion when a high-fat meal is consumed [55]. The area under the curve (AUC) was reduced by approximately 10% with co- administration of a high fat meal. At a dose of 120 mg daily, the mean (± standard deviation) steady-state maximum plasma concentration of gilteritinib was 374 ng/mL (± 190 ng/mL) [54]. Substantial accumulation was observed until steady-state levels were reached by day 15 [55]. Metabolism was primarily via CYP3A4, and the elimination half-life was 113 hours [54].

3.5.Pharmacodynamics
In the phase I/II dose-escalation and expansion study of gilter- itinib (CHRYSALIS, NCT02014558), target inhibition was evalu- ated by plasma inhibitory assay (PIA) at each dose level of gilteritinib [55]. As previously described by Levis et al. [36], PIA assays were performed by collecting plasma from participants at baseline and after treatment with gilteritinib at multiple timepoints. The plasma was incubated in vitro with a FLT3- ITD-expressing AML line, which was then assessed for inhibition of FLT3 phosphorylation [36].FLT3 inhibition was observed at all dose levels of gilteritinib, and ≥ 90% FLT3 inhibition was noted in samples from participants who received a gilteritinib dose of at least 80 mg daily [55].

4.Clinical efficacy
4.1. Phase I/II trials
The safety and efficacy of gilteritinib was assessed in a phase I/ II, first-in-human, dose-escalation and dose-expansion study (CHRYSALIS, NCT02014558) [55]. This multi-center, interna- tional, open-label study enrolled and treated 252 adults ≥ 18 years old with AML that was relapsed or refractory to at least 1 cycle of induction chemotherapy [55]. The presence of a FLT3 mutation was not required for entry on the study, although 191/252 (76%) patients tested FLT3mut+ at study entry. Gilteritinib monotherapy was given orally in continuous 28-day cycles at dose levels ranging from 20 mg/day up to 300 mg/day, which was established as the maximum tolerated dose based on the dose limiting toxicities of grade 3 elevated AST (n = 1) and grade 3 diarrhea (n = 1) [55].

Clinical responses were observed at all dose levels and in subjects with both FLT3-WT and FLT3mut+ AML, although the proportion of patients achieving a response was highest among those who received a dose of at least 80 mg/day and in those with FLT3 mutations [55]. Among the 169 FLT3mut+ patients who received a dose of gilteritinib of at least 80 mg/ day, the overall response rate was 52% (95% confidence inter- val (CI) 44%-60%) and the composite complete remission rate was 41% (95% CI 33%-49%), including 11% (18/169) who achieved a complete morphologic remission (CR) and 30% (51/169) who achieved a complete remission with incomplete platelet or hematological recovery [55]. In contrast, among the 58 FLT3WT participants who were treated with gilteritinibatall dose levels, the overall response rate was 12% (95% CI, 5%- 23%) and the composite complete remission rate was 9% (95% CI, 3%-19%) [55]. Based on the observed clinical activity and tolerability and the results of pharmacodynamic assays performed in this study, a gilteritinib dose of 120 mg/day was selected for further study in phase III trials [55].

4.2. Phase III trials
The initial results of the ADMIRAL trial(NCT02421939) were recently presented [43]. ADMIRAL was an international, open- label, randomized phase III study that compared gilteritinib to salvage chemotherapy in adults with FLT3-ITD and/or FLT3-TKD mutant AML that was in first relapse or was refractory to induc- tion chemotherapy [43]. The vast majority of participants enrolled in ADMIRAL (335/371, 90%) had FLT3-ITD mutations, including 7 subjects (2%) with both FLT3-ITD and TKD mutations, while 31 participants (8%) had FLT3-TKD mutations only [43]. A total of 371 participants were enrolled and randomized 2:1 to gilteritinib 120 mg/day versus salvage chemotherapy (investiga- tor’s choice of MEC (mitoxantrone, etoposide, cytarabine), FLAG- Ida (fludarabine, cytarabine, idarubicin, and G-CSF), azacitidine, or low-dose cytarabine) [43]. Among the 124 subjects rando- mized to the salvage chemotherapy arm, 68 were treated with intensive re-induction chemotherapy (MEC or FLAG-Ida), while 41 received low intensity therapy (azacitidine or low-dose cytar- abine) and 15 did not receive treatment on-study [43]. Allogeneic hematopoietic stem cell transplantation (HSCT) was allowed, and participants on the gilteritinib arm could resume gilteritinib following successful donor engraftment.
Based on an intention-to-treat analysis, the median OS was significantly longer in the gilteritinib arm than in the salvage chemotherapy arm (9.3 months vs. 5.6 months; hazard ratio (HR) 0.637, 95% CI 0.49–0.83) [43].

The compo- site CR (CRc) rate, which included both morphologic CRs as well as CRs with partial or incomplete count recovery, was 54% (134/247) in the gilteritinib group and 22% (27/124) in the salvage chemotherapy group [43]. The CR rate was 21% (52/247) in the gilteritinb arm and 11% (13/125) in the salvage chemotherapy arm. The median duration of remis- sion was 11 months in the gilteritinib group but not estim- able in the chemotherapy arm due to frequent censoring of these patients after response [43]. Of note, the CR rate appeared to be similar among participants with FLT3-TKD mutations alone (19%) compared to those with FLT3-ITD mutations (21%), although the small number of subjects with FLT3-TKD-only mutations who received gilteritinib was small (n = 21), precluding definitive conclusions as to the survival benefits among FLT3-TKD patients across the ther- apeutic arms [43]. Overall, this study established the super- iority of gilteritinib as compared to salvage chemotherapy in adults with relapsed and/or refractory FLT3mut+ AML with respect to both OS and clinical response rates. The interim results of this study, together with the data from the phase I/II CHRYSALIS trial, led to the approval of gilteritinib mono- therapy for the treatment of relapsed and/or refractory FLT3mut+ AML by the U.S. FDA in November of 2018, and the label was expanded in May 2019 to include the OS data, which provide more definitive evidence of clinical benefit.

4.3.Ongoing and planned clinical trials
A number of clinical trials will evaluate gilteritinib in other patient populations and in combination with additional agents, as outlined in Table 2 and briefly summarized here. A phase I trial (NCT02236013) evaluating the safety of gilter- itinib in combination with standard induction and consolida- tion chemotherapy has almost completed accrual, and a randomized phase II trial (NCT03836209) will open soon in which patients with newly diagnosed FLT3mut+AML will be randomly assigned to receive gilteritinib vs. midostaurin in combination with standard induction and consolidation che- motherapy. Gilteritinib is also being studied in combination with azacitidine (NCT02752035), venetoclax(NCT03625505), and atezolizumab (NCT03730012).The Blood and Marrow Transplant Clinical Trials Network(BMT CTN) 1506 study (NCT02997202) is a randomized, placebo-controlled phase III trial evaluating gilteritinib maintenance therapy following allo- geneic HSCT. Another randomized phase III trial is evaluating gilteritinib maintenance (NCT02997202) in adults with FLT3mut + AML in a first complete remission who are not candidates for allogeneic HSCT. Together, these trials will provide data regarding the safety and efficacy of combining gilteritinib with frontline intensive chemotherapy and with lower inten- sity agents, as well as information regarding the role of gilter- itinib in the maintenance setting.

5. Safety and tolerability
5.1. Common and serious adverse events
In the phase III ADMIRAL study, common treatment-emergent adverse events (AEs) (occurring in ≥ 25% of patients) in the gilteritinib arm (n = 246) included an increase in transaminases,fatigue, neutropenic fever, nausea, diarrhea, constipation, head- ache, cough, hypokalemia, anemia, and thrombocytopenia [43]. The most common severe AEs (grade ≥ 3) included neutropenic fever (46%), anemia (41%), thrombocytopenia (23%), increased transaminases (ALT increased in 14%, AST increased in 15%), and hypokalemia (13%) [43]. Elevated creatine phosphokinase (CPK) levels have also been reported with gilteritinib (grade ≥ 3 in 4%) [55]. In general, for any grade 3 or higher toxicity that is felt to be related to gilteritinib, it is recommended that gilter- itinib be held until resolution or improvement to grade 1, at which time gilteritinib may be restarted at a dose of 80 mg daily [54]. This approach may be overly cautious for asymptomatic elevations of CPK in the absence of findings of myositis or rhabdomyolysis, given these elevations may be due not to muscle injury but via impaired hepatic turnover of CPK, which occurs commonly with a number of FLT3 inhibitors and is most likely due to off-target inhibition of the structurally similar CSF1R (encoded by FMS) [56]. Dose reduction is generally not neces- sary for cytopenias that are quite commonly encountered with gilteritinib therapy and may be ongoing during response, as approximately half of patients responding to gilteritinib mono- therapy in the relapsed/refractory setting have ongoing transfu- sion requirements. In cases of marrow hypoplasia thought to be due to gilteritinib, a brief drug hold of 1 to 2 weeks to allow blood count recovery may be desirable.

5.2. Warnings and precautions
The U.S. FDA label for gilteritinib contains several important warnings and precautions, including a boxed warning regard- ing the risk of differentiation syndrome, which has been reported in 1% to 3% of patients treated with gilteritinib [54,57]. If differentiation syndrome is suspected based on the development of pulmonary infiltrates, hypoxia, fever, hypoten- sion, fluid retention, or other symptoms, treatment with corti- costeroids is recommended. Rare cases of posterior reversible encephalopathy syndrome (PRES) have also been reported in patients on gilteritinib, which resolved after holding the drug. Symptoms of PRES may include headache, altered mental status, visual changes, seizures, and hypertension; if sus- pected, a brain MRI should be obtained to confirm the diag- nosis [58]. If PRES occurs, gilteritinib should be permanently discontinued. Pancreatitis has also been reported. If pancrea- titis occurs, gilteritinib should be held until resolution. Once signs and symptoms of pancreatitis resolve, consideration can be given to cautiously restarting gilteritinib at a lower dose (80 mg per day).

Prolongation of the corrected QT interval (QTc) has been reported from trials with gilteritinib, although concurrent medications that also prolong QT (e.g. ondansetron, quinolone antibiotics, and azole antifungals) were not restricted if felt to be essential for appropriate medical management. In the phase III ADMIRAL trial, a prolonged QTc interval occurred in 12/246 participants (5%), although only 1 subject had an increase in the QTc to > 500 msec [43]. Due to the risk of QTc prolongation, monitoring of electrolytes throughout gil- teritinib therapy is recommended, with repletion of potassium and magnesium levels as needed [54]. An electrocardiogram should be performed at baseline, on days 8 and 15 of cycle 1,and prior to the start of cycles 2 and 3 [54]. If the QTc interval increases to > 500 msec, gilteritinib should be held [54]. When the QTc improves to ≤ 480 msec or within 30 msec of baseline, gilteritinib may be restarted at a lower dose (80 mg/day) [54]. If the QTc increases by > 30 msec, the increase should be confirmed by repeat EKG and, if genetic breeding confirmed, dose reduction considered [54].

5.3. Drug-drug interactions
If taken concomitantly with gilteritinib, strong CYP3A inhibitors (i.e. voriconazole, posaconazole) have been shown to increase gilteritinib concentration and therefore have the potential to increase the risk of toxicity [54]. If alternative options are not available and concurrent therapy with a strong CYP3Ainhibitor is necessary, then patients should be monitored for toxicity more frequently [54]. Dietary furanocoumarins such as grapefruit and its juice strongly inhibit CYP3A4 and should be avoided during gilteritinib therapy due to unpredictable increases in gilteritinib concentration that could magnify toxicity risk. Additionally, con- comitant use of gilteritinib with combined P-gp and strong CYP3A inducers (i.e. St. John’s Wort) is not recommended as this may decrease gilteritinib exposure (and thus, efficacy of gilteritinib could theoretically be impaired) [54]. Gilteritinib has the potential to reduce the efficacy of drugs that target sigma non-specific receptor and/or 5HT2B, such as escitalopram, fluox- etine, or sertraline. Use of alternative medications is generally recommended unless these drugs are considered essential.

6.Conclusion
In summary, gilteritinib is an orally available tyrosine kinase inhibitor for the treatment of relapsed and/or refractory FLT3mut+ AML that is approved in the U.S. and Japan and is now under review at the European Medicines Agency. This agent is a potent and selective inhibitor of FLT3 that is able to inhibit both FLT3-ITD and FLT3-TKD mutations. Gilteritinib Ko143 is the first FDA-approved agent for the treatment of relapsed/ refractory FLT3mut+ AML and has been shown in a randomized phase III trial to have an overall survival benefit compared to salvage chemotherapy.

7. Expert opinion
Gilteritinib is the first selective FLT3 inhibitor to be approved by the U.S. FDA for the treatment of FLT3mut+ relapsed and/or refractory AML. As described in this article, the initial approval was based on interim response rates and tolerability data from the gilteritinib arm of a randomized phase III trial conducted in patients with first relapse or primary refractory FLT3mut+ AML [43]. This trial has subsequently been presented and showed an improvement in response and OS among gilteritinib- treated patients compared to salvage chemotherapy [43]. Importantly, gilteritinib is generally well-tolerated and, as an oral agent, is frequently administered on an outpatient basis. As a result, gilteritinib monotherapy is rapidly becoming the standard of care for patients with relapsed and/or refractory FLT3mut+ AML. Of note, quizartinib has also been shown to lead to improved OS compared to salvage chemotherapy in patients with relapsed/refractory FLT3-ITD+ AML [44] and was recently approved for relapsed/refractory FLT3-ITD+ AML in Japan. There are several important differences to note in comparing gilteritinib and quizartinib. First, although quizartinib is also a potent and selective inhibitor of FLT3, it is a type II inhibitor and therefore has clinical activity against FLT3-ITD mutations but not TKD mutations. Indeed, although responses to quizar- tinib in relapsed/refractory FLT3-ITD+ patients occur at least as frequently as gilteritinib, prior studies have demonstrated the rapid development of secondary FLT3-D835 resistance muta- tions in quizartinib-treated patients [59,60]. Selection for on- target resistance mutations in the TKD may be one reason why the median duration of response to quizartinib in this context appears relatively short [44].

In contrast, gilteritinib is a type I inhibitor that is able to inhibit both FLT3-ITD and TKD muta- tions [40]. These differences in clinical activity may contribute to modestly longer duration of clinical responses in patients treated with gilteritinib, although this remains speculative as gilteritinib and quizartinib have not been directly compared and the patient populations included in each drug’s phase 3 trial were not entirely overlapping. The investigational agent crenolanib is another type I kinase inhibitor that, like gilteriti- nib, has activity against both FLT3-ITD and also FLT3-D835 mutations [38,39]. Of note, crenolanib has a short half-life requiring dosing three times a day, whereas gilteritinib has a longer half-life and is taken once daily [46,55].Gilteritinib and quizartinib also differ in terms of their kinase inhibitory and toxicity profiles. In particular, gilteritinib is only a weak inhibitor of Kit, but does inhibit Axl, which has been hypothesized to contribute to its clinical activity, though to date this has not been extensively validated from clinical specimens. Quizartinib inhibits both FLT3 and Kit at relatively similar potencies, which might contribute to antileukemic activity but also has the potential to contribute to the cyto- penias that are commonly observed with this agent [41,61].

Regardless of Kit inhibition, gilteritinib is also associated with cytopenias, which are prominent during the first 1 to 2 months of therapy and then often improve or resolve in a substantial number of responding patients. Both agents can prolong the QT interval. However, QT prolongation was a particularly pro- minent toxicity in phase I/II testing of quizartinib and indeed was the drug’s dose-limiting toxicity [62]. Mitigation of this toxicity largely rested upon avoidance of concurrent QT prolonging agents, ensuring normal potassium and magne- sium levels with supplementation, and the use of lower qui- zartinib doses. In a phase IIB study of quizartinib, the rate of QTc prolongation by > 60 msec was 19.4% in patients who received quizartinib at a dose of 60mg daily and 5.3% in patients who received quizartinib at a dose of 30mg daily [63]. By contrast, severe QT prolongation was rare with gilter- itinib and only weakly dose-dependent. In the CHRYSALIS phase I/II study of gilteritinib, QT prolongation occurred in 3% participants (8/252) [55]. It should be noted that both of these drugs have almost exclusively been tested in patients with baseline QTcF measurements that were squarely in the normal range, and EKGs were extensively tested with central cardiology readings in realtime during the trials. Caution and close monitoring should be exercised when these drugs are administered to patients with known prior heart disease.

One current challenge in treating patients with the selective FLT3 inhibitors is that standard clinical response criteria may incompletely characterize clinical responses to these novel agents. We recently conducted a clinicopathologic correlative study that included a subset of patients who were treated with gilteritinib on the phase I/II CHRYSALIS study [57]. We found that gilteritinib induces terminal myeloid differentiation in approxi- mately half of clinical responders, which is characterized by a stable to increased total marrow cellularity and a persistently elevated FLT3 mutant allelic burden despite a significant reduc- tion in blast percentage [57]. In the remainder of responders, gilteritinib induces a response without differentiation character- ized by a reduction in total marrow cellularity along with a reduction in marrow blast percentage and FLT3 mutant allelic burden [57]. We and others have also noted differentiation responses in a subset of quizartinib-treated patients [64,65]. Additionally, a recent study by Levis et al. found that mutational clearance of FLT3-ITD, as measured by a sensitive next- generation sequencing assay, was associated with improved OS in patients treated with gilteritinib [66]. Together, these studies suggest that responses as defined by the International Working Group criteria [67] may not fully characterize clinical responses to gilteritinib and the other selective FLT3 inhibitors and thatfurther study to better understand responses is needed.

Despite the promising clinical trial results with gilteritinib and the other FLT3 inhibitors, these agents alone are not curative and responding patients typically develop secondary resistance and progression of disease [38,42–44,55]. An important area of ongoing research involves understanding mechanisms of second- ary resistance to the selective FLT3 inhibitors. Although secondary FLT3 mutations are common inpatients who develop resistance to quizartinib and sorafenib, on-target resistance to gilteritinib is decidedly uncommon, occurring in approximately 10% of patients [68]. Instead, we have identified Ras pathway mutations, most commonly NRAS and KRAS, as a common mechanism of secondary resistance to gilteritinib [68]. Secondary mutations in FLT3 at the F691L gatekeeper domain and BCR-ABL fusions have also been identified at the time of progression in patients treated with gilteritinib [68]. Consistent with this data, Zhang and colleagues have evaluated mechanisms of resistance to crenolanib and found that patients with Ras pathway mutations at baseline are less likely to respond to crenolanib [69]. They also identified several other mechanisms of resistance to crenolanib, including the persistence or expansion of clones containing IDH1, TET2,and TP53 mutations [69]. Importantly, complex patterns of clonal selection and expan- sion have been identified inpatients treated with FLT3 inhibitors, suggesting that FLT3 inhibitor monotherapy – even with those that are highly potent and selective – is inadequate to fully eradicate leukemic clones [59,68,69].

Current clinical trials are evaluating the tolerability and clinical activity of gilteritinib in combination regimens. An exciting ongoing phase Ib trial (NCT03625505) is evaluating gilteritinib in combination with venetoclax, another oral agent that has been reported to have synergistic anti-tumor effects with FLT3 inhibitors in vitro. It is hoped that trials such as this may generate higher response rates and/or longer remission duration with development of less secondary resistance than gilteritinib monotherapy. Ultimately, a goal in the field is to develop highly effective yet generally less toxic outpatient regimens for the treatment of FLT3mut+ AML.Concurrent with studies of combinations of novel agents, planned or ongoing randomized trials will also establish whether frontline intensive chemotherapy combined with gil- teritinib increases response rates compared to intensive che- motherapy combined with midostaurin. Whether the addition of a selective FLT3 inhibitor such as gilteritinib rather than a multikinase inhibitor like midostaurin impacts the level of measurable residual disease (MRD) following induction or at the time of transplant is also a key endpoint for these studies, and it is conceivable that this potential surrogate endpoint for clinical benefit will be available years before final overall sur- vival data are mature. Additional randomized studies are also examining the role of prolonged maintenance therapy with gilteritinib after chemotherapy or transplant (NCT02927262, NCT02997202). If positive, this would represent a major shift in the standard treatment duration for AML, which in the U.S. currently does not use prolonged maintenance therapy for any subgroup of patients.

We expect the above trials will yield mature data in the next 2–5 years which will inform clinical strategies. As largely has been the case for BCR-ABL1+ acute lymphoblastic leukemia following the incorporation of ABL1 kinase inhibitors into standard che- motherapy, it is hoped that FLT3-ITD eventually will no longer be considered a high-risk molecular lesion in AML due to substantial improvements in cure rates following integration of selective, potent FLT3 inhibitors and intensive chemotherapy. With such improvements in the treatment ofFLT3mut+ AML, one could envi- sion that novel and low-intensity agents could someday replace intensive chemotherapy for many patients as the preferred drugs to combine with FLT3 inhibitors. This mirrors the de-escalation of chemotherapy approaches in acute promyelocytic leukemia fol- lowing the optimization of targeted therapy for that particular AML subtype.In summary, gilteritinibis an important newly-approved agent in the AML arsenal and has become the standard of care for patients with relapsed and/or refractory FLT3mut+ AML. Gilteritinib is now being studied in combination with traditional chemotherapeutic agents in the frontline setting, with novel agents in relapsed/refractory patients, and as a single-agent main- tenance therapy after completion of consolidation or after transplant.