Initial prognostication systems for younger, fit patients with AML receiving intensive chemotherapy included the FLT3 mutation among karyotypes associated with negative prognostic outcomes.¹ The development of small-molecule agents inhibiting FLT3-mutant tyrosine kinase has led to significant improvements in outcomes for these patients. With incorporation of FLT3 inhibitors into the AML treatment approach, the 2022 European Leukemia Net classification now considers FLT3-mutated disease to carry intermediate risk.²

FLT3 inhibitors are divided into classes based on their underlying biology and ability to block different types of FLT3 mutations. Type I inhibitors bind both active and inactive mutant FLT3 tyrosine kinase receptors and therefore are active against FLT3-mutated AML bearing both FLT3 ITD and FLT3 TKD mutations. By contrast, type II inhibitors bind inactive mutant FLT3 receptors and therefore do not demonstrate any inhibition of FLT3-TKD–mutated disease. Among FLT3 inhibitors that currently have FDA approval for the treatment of AML, midostaurin and gilteritinib are type I inhibitors and quizartinib is a type II inhibitor. Sorafenib, another type II FLT3 inhibitor, is also used off-label for certain patients with FLT3-mutated AML.³

The FLT3 inhibitors are characterized by different toxicity profiles, with those agents targeting a more limited number of kinases thought to carry fewer adverse effects (AEs). Of most importance, clinical use of quizartinib has demonstrated a risk of corrected QT interval (QTc) prolongation and nausea, vomiting, and myelosuppression.⁴ By contrast, midostaurin is primarily associated with gastrointestinal (GI) toxicities such as nausea, vomiting, and diarrhea, as well as fatigue and pulmonary toxicities.⁵ Gilteritinib is characterized by myelosuppression with long count recovery, particularly seen with use in combination therapy. Liver AEs and a small risk of differentiation effects are also seen with gilteritinib.⁶ The safety profile of sorafenib established with its use in certain solid tumor types and AML is characterized by hand–foot syndrome, GI toxicities, and rash.⁷

Midostaurin was the first FLT3 inhibitor approved for use in AML with FLT3 mutations based on results of the phase III RATIFY trial. Younger, fit patients between the ages of 18-59 were randomized to treatment with cytarabine and anthracycline-based chemotherapy in combination either with midostaurin or placebo.⁸ Although more than 7000 patients were screened, only 200 patients were eventually enrolled on each arm of the study. Patients received either midostaurin or placebo in combination with 7+3 induction and high-dose cytarabine consolidation. Patients achieving first complete remission (CR1) could continue with midostaurin or placebo as maintenance monotherapy for 12 months, and those without CR1 could proceed to allogeneic stem cell transplant (SCT).

Results of the RATIFY trial published in the New England Journal of Medicine in 2017 demonstrated a significant difference in OS between patients receiving midostaurin vs placebo (median 74.7 vs 25.6 months; 1-sided P = .009), which corresponded with a 7.2% difference in survival at 4 years. CR rates were not significantly different (59% vs 54%). Presented at American Society of Hematology 2024, there is now 10-year follow-up survival results for the RATIFY trial showing durable event-free survival benefit.⁹

Midostaurin benefit was also apparent in patients who went on to allogeneic SCT after CR1. Allogeneic SCT benefited patients regardless of midostaurin or placebo treatment, emphasizing its importance to improve long-term outcomes independent of targeted treatment. However, the best outcome was seen in those who received both FLT3 inhibitor and transplantation, which became standard of care in this patient population.

QuANTUM-First repeated the trial design used in RATIFY to investigate the newer-generation FLT3 inhibitor quizartinib, again using placebo control in combination with a 7+3 induction backbone and high-dose cytarabine consolidation.¹⁰ The patient population was slightly different, including patients 75 years of age or younger, whereas the upper cutoff for RATIFY was 59 years. A second difference was that QuANTUM-First enrolled only patients with FLT3-ITD mutations because of the selective activity of quizartinib, whereas the efficacy of midostaurin against FLT3-TKD mutations meant that patients with mutations of either type were included in the RATIFY trial.

QuANTUM-First also provided a longer duration of maintenance therapy—up to 3 years after completion of upfront chemotherapy—compared with the 2 years of maintenance given in RATIFY. 

The primary endpoint, OS, was statistically improved with quizartinib vs placebo (median: 31.9 vs 15.1 months; 2-sided P = .032).

The newer QuANTUM-First trial used endpoints beyond OS and remission to more thoroughly characterize the benefit of quizartinib addition to chemotherapy, including measurable residual disease testing. MRD analysis can detect multiple FLT3-ITD mutation sequences indicative of disease in as little as 800-1100 ng of genomic DNA per sample, using highly sensitive next-generation sequencing. Variant allele frequency was also used to quantify FLT3-ITD–mutated AML cells.

Thus, the QuANTUM-First trial was able to demonstrate deeper responses with quizartinib, than standard 7+3 chemotherapy, meaning that statistically significant differences in detectable MRD were seen in patients treated with quizartinib and standard 7+3 chemotherapy vs patients treated with 7+3 and placebo.¹¹ Furthermore, treatment arm differences in MRD became more pronounced with additional treatment cycles, and reduction in FLT3 MRD levels was predictive of OS. These new biomarker data with improved sensitivity show that long-term survival of patients with FLT3-ITD–mutated AML derives from an early and sustained reduction in FLT3-ITD leukemic burden with quizartinib treatment.

Quizartinib, like all drugs, has some toxicity. On the QuANTUM-First trial, there was more neutropenia in the quizartinib arm vs the placebo arm (20% vs 10% all grades; 18% vs 9% grade ≥3) and more febrile neutropenia (44% vs 42% all grades; 43% vs 41% grade ≥3).^10,12

These are some potential cardiac AEs specific to quizartinib. There was more QTc prolongation with quizartinib compared with placebo. In total, 13.6% of patients receiving quizartinib had any form of treatment-emergent QTc prolongation compared with 4.1% of patients receiving placebo. Approximately 2.3% of patients receiving quizartinib had QTc that was >500 ms, for example, and 10% had a QTc increase from baseline of >60 ms. Other cardiac treatment-emergent AE reported with quizartinib include cardiac arrest/ventricular fibrillation and ventricular tachycardia, which occurred in <1% of patients.

In this trial, 2 patients had cardiac arrest with recorded ventricular fibrillation in the setting of severe hypokalemia while receiving quizartinib, and 1 patient who received quizartinib died in their sleep.

QTc prolongation led to treatment discontinuation for 2 patients receiving quizartinib.

Because of the known risk of cardiac AEs associated with quizartinib, patients should be counseled and evaluated for cardiac risk under the FDA-mandated Risk Evaluation and Mitigation Strategy program. This program ensures that healthcare professionals prescribing quizartinib receive education regarding the mechanism responsible for QTc prolongation with quizartinib and proper patient selection and treatment based on risk factors for QTc prolongation. 

RATIFY and QuANTUM-First showed that both drugs consistently improve OS in comparison with the control arms, but the different age criteria included in each trial makes it impossible to compare the outcomes. No study directly comparing quizartinib and midostaurin in a younger patient population (younger than 60 years of age) has been performed, but subgroup analysis of QuANTUM-First shows results comparable with those of RATIFY in patients of this age group.^8,10 On the other hand, in a post hoc analysis, older patients (age 60-75 years) treated with quizartinib in QuANTUM-First did not experience enough survival benefit to recommend this regimen for that patient population (17.5 vs 14.2 months with placebo; HR: 0.911; 95% CI: 0.658-1.263.) This may be because of the presence of molecular aberrations in addition to FLT3-ITD in these older patients, who often have secondary AML, in contrast with younger patients who are more likely to have underlying disease driven by the single FLT3 mutation.^8,13

Considering the best therapy for older, unfit patients with FLT3-mutated AML, numerous clinical trials have investigated addition of an FLT3 inhibitor to lower-intensity venetoclax–azacitidine chemotherapy rather than the standard 7+3 backbone.¹⁴ This regimen is tolerated well by older patients and has led to very high response rates in FLT3-mutated disease. Ongoing trials are evaluating triplets of an FLT3 inhibitor with venetoclax and a hypomethylating agent.¹⁵ Surprisingly, patients in these trials have demonstrated such high response rates that triplet therapy may challenge the efficacy of the 7+3 plus FLT3 inhibitor regimen, even for younger patients. Many questions remain regarding the optimal upfront therapy for all patients with FLT3-mutated AML.

FLT3 inhibitors are also beneficial in treatment of relapsed/refractory (R/R) FLT3-mutated AML. The ADMIRAL trial, first published in 2019, demonstrated that gilteritinib, a type I inhibitor efficacious in both FLT3-ITD–mutated and FLT3-TKD–mutated disease, provides improved outcomes compared with salvage chemotherapy (either intensive or less intensive) in the R/R setting¹⁶; 371 adult patients with R/R FLT3-mutated AML were randomized 2:1 to receive single-agent gilteritinib 120 mg/day or investigator-selected salvage chemotherapy (options including intensive mitoxantrone/etoposide/cytarabine, FLAG-IDA regimens, or less intensive low-dose cytarabine or azacitidine).

Patients with FLT3-mutated R/R AML receiving gilteritinib monotherapy achieved a median OS of 9.3 months, compared with 5.6 months in those receiving salvage chemotherapy (HR: 0.665; P = .0013).¹⁷ The 1-year OS rate for gilteritinib-treated patients was 37%, in comparison with 19% in patients receiving salvage chemotherapy. It should be noted that the QuANTUM-R trial, which compared quizartinib with salvage chemotherapy, found a much smaller OS benefit (median: 6.2 vs 4.7 months) and a lack of significant event-free survival benefit that along with other concerns resulted in the FDA not approving this agent in the R/R setting.¹⁸

In the ADMIRAL trial, patients who had previously received an upfront FLT3 inhibitor showed less benefit with gilteritinib treatment in the R/R setting than did FLT3 inhibitor–naive patients, potentially because of acquired resistance.¹⁹ However, all patients benefited from single-agent gilteritinib in comparison with salvage chemotherapy, regardless of prior TKI exposure.

The safety profile of gilteritinib is well-documented, with the most notable AEs being increased liver function test abnormalities and myelosuppression, along with a few rare toxicities including posterior reversible encephalopathy. Overall, these are considered to be rare and counterbalanced by clear clinical benefit with this agent. Long-term follow-up data published in 2022 has continued to demonstrate improved survival in gilteritinib-treated patients who received subsequent allogeneic SCT with 15% still alive at 3 years following gilteritinib vs  10% or fewer patients treated with salvage chemotherapy.¹⁷

Patients who went on to subsequent allogeneic SCT in the ADMIRAL trial also had the option of resuming gilteritinib monotherapy after the allogeneic SCT. Those who received posttransplant gilteritinib experienced extended survival vs those who did not (median: 16.2 vs 8.4 months).²⁰

The demonstrated benefit of FLT3 inhibitor treatment as posttransplant maintenance in the R/R setting led to the question of whether a posttransplant FLT3 inhibitor could also benefit patients undergoing allogeneic SCT in first remission or response.

This was first investigated in the SORMAIN trial with the first-generation inhibitor sorafenib²¹; 83 patients with FLT3-ITD AML who achieved complete hematologic remission with allogeneic SCT were randomized to either sorafenib up to 800 mg/day or placebo for up to 24 months, with a primary endpoint of RFS.

Statistically significant benefits in both RFS and OS were seen for patients receiving sorafenib vs placebo in this posttransplant setting. At 24 months, 85% of sorafenib-treated patients remained relapse free in comparison with 53% of those receiving placebo (P = .002), and 91% vs 66% were still alive (P = .007).

Positive results were seen with a similar regimen in an open-label randomized trial in China.^22,23 In 202 patients with FLT3-ITD AML, 5-year survival with sorafenib maintenance after allogeneic SCT was significantly increased compared with placebo (HR: 0.55; P = .011). RFS and leukemia-free survival were also significantly improved. The most common AEs seen with sorafenib in this trial were infections, GI toxicity, electrolyte alterations, and skin toxicities including rash.

The definitive trial establishing the role of posttransplant FLT3 inhibitor therapy was the Blood & Marrow Transplant Clinical Trials Network global, double-blind phase III MORPHO trial.²⁴ Gilteritinib maintenance following allogeneic SCT was compared with placebo in adults with FLT3-ITD AML, with the inclusion of sensitive FLT3-ITD MRD assessment as a biomarker. Patients with FLT3-ITD were randomized between Day 30 and Day 90 following allogeneic SCT to receive gilteritinib 120 mg/day or placebo for 24 months. Of note, this trial included assessment of MRD at multiple timepoints pre transplant and post transplant to investigate the potential utility of MRD status in selecting patients for transplant.

Not unexpectedly, patients receiving gilteritinib post transplant had higher rates of grade ≥3 AEs than those receiving placebo, particularly decreased neutrophil count, decreased platelet count, anemia, and liver function test abnormalities.

The primary endpoint of RFS was not statistically different between the gilteritinib and placebo arms (HR: 0.679; P = .0518). The secondary endpoint of OS also demonstrated no difference between treatment arms  (HR: 0.846; P = .4394.)

Although the primary endpoint was not met, the MORPHO trial provided a notable finding regarding the impact of FLT3 MRD, which was analyzed using a polymerase chain reaction next-generation sequencing assay in 98% of patients prior to transplant and 97% of patients post transplant. MRD positivity (defined as FLT3-ITD variant allele frequency of 1x10^-6 or higher) rates were 46.1% prior to transplant and 19.9% following transplant. Surprisingly, a highly significant difference in RFS was seen between the gilteritinib and placebo arms in MRD-positive patients (HR: 0.515; P = .0065), although no benefit of gilteritinib treatment had been seen in the overall patient population. Patients who were MRD negative prior to transplant did not see improvement in RFS with gilteritinib vs placebo. This finding demonstrates the potential of peritransplant FLT3 MRD to guide posttransplant treatment strategy for individual patients and suggests that gilteritinib maintenance should be considered standard of care for those who exhibit peritransplant MRD-positive status.

As expected, patients who were MRD positive going into transplant had significantly worse survival outcomes than those who were MRD negative, regardless of the treatment arm.

Upfront FLT3 inhibitor therapy has been found to affect patterns of clonal evolution seen in subsequent clinical relapse.²⁵ The primary distinction is between patients with loss of the FLT3-ITD clone and recurrent disease characterized as FLT3 wild-type AML in contrast with those who have persistent FLT3 mutation with the possible addition of treatment-resistant mutations. For example, approximately 46% of patients who developed recurrent AML after receiving midostaurin in combination with 7+3 chemotherapy in the RATIFY trial were seen to have FLT3‑ITD wild-type disease, which suggests that clonal selection and evolution have taken place between the time of diagnosis and recurrence.

Subsequent FLT3 inhibitor therapy has been associated with progressively reduced survival with each additional line of treatment.²⁶ The best outcomes for FLT3 inhibition are found when these agents are used in the frontline setting or as early as possible in sequential therapy.

Certain FLT3 mutations have been associated with resistance to FLT3 inhibitor treatment and persistence of FLT3-mutated disease.²⁷ Patients treated with the type II inhibitors sorafenib and quizartinib frequently develop resistance derived from TKD mutations. The FLT3 F691L mutation, which confers resistance to almost all known TKIs, is frequently seen following treatment with midostaurin, sorafenib, or gilteritinib.

In addition, patients with FLT3 wild-type disease at the time of clonal evolution are often characterized by the appearance of mutations involving the RAS/MAPK kinase pathway, particularly after type I gilteritinib and midostaurin therapy.²⁸ Other mutations such as BCR-ABL, although rare, also occur in patients after treatment with FLT3 inhibitors.