Skip to main content
NCBI home page
Blood Advances logoLink to Blood Advances
. 2025 Oct 8;10(3):917–928. doi: 10.1182/bloodadvances.2025016444

FLT3-ITD measurable residual disease from the QuANTUM-First trial

Mark J Levis 1,∗, Harry P Erba 2, Pau Montesinos 3, Hee-Je Kim 4, Radovan Vrhovac 5, Elżbieta Patkowska 6, Pavel Žák 7, Po-Nan Wang 8, Jaime E Connolly Rohrbach 9, Ken C N Chang 9, Li Liu 9, Yasser Mostafa Kamel 9, Karima Imadalou 9, Arnaud Lesegretain 9, Jorge Cortes 10, Mikkael A Sekeres 11, Hervé Dombret 12, Sergio Amadori 13, Jianxiang Wang 14, Richard F Schlenk 15, Alexander E Perl 16
PMCID: PMC12905609  PMID: 41052406

Key Points

  • •

    The addition of quizartinib to induction chemotherapy results in deeper remissions.

  • •

    FLT3-ITD MRD after induction treatment correlates with relapse and survival.

Visual Abstract

graphic file with name BLOODA_ADV-2025-016444-ga1.jpg

Open in a new tab

Abstract

QuANTUM-First was a randomized trial that demonstrated that the addition of quizartinib, a potent and selective FMS-like tyrosine kinase 3 (FLT3) inhibitor, to induction and consolidation chemotherapy, followed by monotherapy maintenance, improved the survival for patients with newly diagnosed FLT3–internal tandem duplication (FLT3-ITD)–mutated acute myeloid leukemia. We conducted a post hoc analysis of the trial data focusing on measurable residual disease (MRD) as assayed using an amplicon-based next-generation sequencing assay, and on the impact of molecular biomarkers such as FLT3-ITD insertion length and comutations. This is, to our knowledge, the first prospective, randomized trial of an FLT3 inhibitor in newly diagnosed patients in which FLT3-ITD MRD data were collected throughout therapy. We established that quizartinib induces deeper remissions with respect to FLT3-ITD MRD vs placebo, and that the amount of MRD at the completion of induction correlates with relapse and survival. We found that longer FLT3-ITD insertion mutations correlated with worse outcome, quizartinib was beneficial irrespective of insertion mutation length, and the FLT3-ITD MRD assay was more sensitive when bone marrow was used vs peripheral blood. Regardless of the presence of NPM1 (nucleophosmin 1) comutation, quizartinib increased the rates of MRD negativity at the end of induction vs placebo. Finally, comparison of the FLT3-ITD mutation length between the polymerase chain reaction (PCR) with capillary electrophoresis assay obtained at screening and the PCR next-generation sequencing MRD assay performed at the end of induction showed a 96.2% concordance with the exact ITD length. This trial was registered at www.clinicaltrials.gov as #NCT02668653.

Introduction

FMS-like tyrosine kinase 3–internal tandem duplication (FLT3-ITD) mutations are one of the most common mutations in acute myeloid leukemia (AML)1, 2, 3 and are associated with an adverse impact on prognosis.1,2,4, 5, 6, 7 A number of small-molecule FLT3 inhibitors have been developed, have received regulatory approval, and have become a standard component of the treatment regimens for FLT3-ITD AML.8, 9, 10, 11, 12

Retrospective studies of banked DNA samples collected from clinical trial participants established a number of variables that influenced the outcome of patients with FLT3-ITD–mutated AML, and measurable residual disease (MRD) has emerged as one of the most important of these.13, 14, 15, 16 To date, the prognostic impact of FLT3-ITD MRD in newly diagnosed patients undergoing intensive therapy has been established from retrospective analyses. There are no data from randomized trials testing intensive chemotherapy and an FLT3 inhibitor compared with placebo to confirm whether any of these biomarkers continue to exert a prognostic effect with current therapies.14,16, 17, 18

QuANTUM-First (ClinicalTrials.gov identifier: NCT02668653) was a phase 3 randomized trial of induction, consolidation, and maintenance therapy combined with either placebo or quizartinib for patients with newly diagnosed FLT3-ITD–mutated AML.12 Based on the trial results, quizartinib received regulatory approval for the treatment of this patient population in the United States, European Union, United Kingdom, Switzerland, Hong Kong, and Japan. Bone marrow aspirates and peripheral blood samples were collected from patients on this trial at enrollment and throughout the course of treatment at prespecified time points after achieving remission to therapy. These samples were analyzed at baseline for FLT3-ITD length by polymerase chain reaction (PCR) and capillary electrophoresis and for comutations by next-generation sequencing (NGS), and during the course of treatment for FLT3-ITD MRD by a highly sensitive and validated amplicon-based approach specifically designed for this study.19 We conducted this post hoc analysis of FLT3-ITD MRD at different time points during treatment to better understand how quizartinib affected the clinical outcomes of remission and survival in the QuANTUM-First trial, and to potentially provide guidance on the clinical management of this disease going forward.

Methods

Clinical trial

A detailed description of the QuANTUM-First trial has been published previously.12

MRD assay and molecular studies

MRD was analyzed at the following prespecified time points during therapy, only in patients achieving remission: (1) after completion of induction; (2) after completion of consolidation before entering maintenance; and (3) before allogeneic hematopoietic cell transplantation (allo-HCT, if performed; supplemental Figure 1). Trial participants entering the induction phase (539 total) received either 1 (429/539 [79.6%]) or 2 (110/539 [20.4%]) courses of induction based on the results of bone marrow aspirates collected on or around day 21 of induction. Genomic DNA was prepared from bone marrow aspirates and/or peripheral blood from patients achieving remission. If the aspirate was dry, peripheral blood was used. This included patients with complete remission (CR) and CR with incomplete neutrophil or platelet recovery (CRi). Composite complete remission (CRc) was defined as CR + CRi. These induction samples were collected at the time of remission assessment (before any further therapy) and analyzed using an amplicon-based MRD assay developed specifically for the QuANTUM-First trial and performed by Navigate BioPharma Services, Inc (Carlsbad, CA), as previously described.19 Technical details on the MRD assay are provided in the supplemental Methods. The cutoff date for MRD data was September 30, 2022.

Statistical analysis

Rates of CR with FLT3-ITD MRD negativity and rates of CRc with FLT3-ITD MRD negativity were prespecified secondary end points of QuANTUM-First. Comparisons of CR/CRi/CRc rates by the end of induction, including those with FLT3-ITD MRD negativity, were made using a stratified Cochran-Mantel-Haenszel test. Comparison of the FLT3-ITD MRD variant allele frequency (VAF) by the end of induction and by the end of consolidation, between treatment arms, was made using a Wilcoxon rank-sum test. The medians of best FLT3-ITD MRD VAF during induction and during consolidation were calculated by collecting the maximum values for each patient at each available time point during induction (day 21, cycle 1 [C1]; day 56, C1; day 21, C2; and day 56, C2) or during consolidation (day 21, C1; day 21 of the last cycle; and after allo-HCT), and then using the lowest of these maximal values from each patient to calculate the median across patients up to the end of induction or up to the end of consolidation. Comparisons of overall survival (OS) and relapse-free survival (RFS) by FLT3-ITD length were made using unstratified Cox regression analysis. All P values were not adjusted for multiplicity.

The study was conducted in compliance with the Declaration of Helsinki and approved by institutional review boards at each site. All patients provided informed consent.

Results

Of 539 patients in the QuANTUM-First trial, 368 (68.3%) achieved CRc and 297 (55.1%) achieved CR after 1 or 2 courses of induction (Figure 1). Of 368 patients in CRc, 321 (87.2%) had samples available for MRD analysis, with 162 (84.4%) in the quizartinib arm and 159 (90.3%) in the placebo arm. Characteristics of patients achieving CRc are listed in Table 1. Overall, there was a broad range of FLT3 mutation burden at diagnosis: 67% of the patients had an FLT3-ITD VAF of >25%, corresponding to an allelic ratio of 0.33; with 54% having an FLT3-ITD VAF ranging between >25% and ≤50% (approximating allelic ratios of 0.33-1), and 13% having an FLT3-ITD VAF of >50%. FLT3-ITD MRD data were not available for 47 patients achieving CRc after 1 or 2 courses of induction because samples were not collected, or for technical reasons, such as inadequate DNA yield to perform the assay, or insufficient reads to make a meaningful assessment. The baseline characteristics of the 47 patients in CRc with no MRD data available were overall similar to those of the 321 patients in CRc with MRD data (supplemental Table 1; compare with Table 1).

Figure 1.

Figure 1.

Open in a new tab

Diagram for MRD assessment and analysis in patients who achieved CRc and CR per independent review committee (IRC) after 1 or 2 courses of induction. MRD data were not available for 47 patients achieving CRc (39 of whom with CR) because the sample was not collected, or the quantity was not sufficient to perform the assay, or insufficient read lengths were obtained from the assay to make a meaningful assessment. ∗After 1 or 2 courses of induction. MRD data not available for 47 patients achieving CRc (39 of whom with CR) because sample was not collected, or quantity was not sufficient to perform the assay, or insufficient read lengths were obtained from the assay to make a meaningful assessment. †Include protocol-specified allo-HCT and non–protocol-specified allo-HCT. ITT, intent-to-treat.

Table 1.

Baseline characteristics of patients with CRc per IRC by the end of induction and available MRD data

Patient characteristics Patients who achieved CRc by the end of induction with available MRD data (N = 321)∗
Quizartinib (n = 162) Placebo (n = 159)
Age
 Median (range), y 56 (23-75) 55 (20-75)
 <60 years, n (%) 99 (61.1) 96 (60.4)
 ≥60 years, n (%) 63 (38.9) 63 (39.6)
 60-64 years, n (%) 20 (12.3) 25 (15.7)
 ≥65 years, n (%) 43 (26.5) 38 (23.9)
Sex, n (%)
 Male 73 (45.1) 65 (40.9)
 Female 89 (54.9) 94 (59.1)
ECOG PS, n (%)
 0 53 (32.7) 56 (35.2)
 1 83 (51.2) 82 (51.6)
 2 26 (16.0) 21 (13.2)
Mutated NPM1, n (%) 99 (61.1) 102 (64.2)
FLT3-ITD/total FLT3, n (%)
 ≥3% to ≤25% 57 (35.2) 50 (31.4)
 >25% to ≤50% 85 (52.5) 88 (55.3)
 >50% 20 (12.3) 21 (13.2)
 >25% 105 (64.8) 109 (68.6)
FLT3-ITD length, median (range), bp 51 (6-243) 54 (6-231)
MRD sample collection, n (%)
 Peripheral blood 16 (9.9) 11 (6.9)
 Bone marrow aspirate 161 (99.4) 158 (99.4)
Open in a new tab

ECOG PS, Eastern Cooperative Oncology Group performance status; IRC, independent review committee.

∗

MRD data are available for 321 of 368 patients achieving CRc after 1 or 2 courses of induction (47 patients had no MRD data).

Quizartinib added to chemotherapy results in deeper remissions

In this study, FLT3-ITD mutations from MRD samples were detected with VAFs as low as 2 × 10−6. Among 258 patients who achieved CR by the end of induction and had MRD assessment during the first or second induction cycle, the median best FLT3-ITD MRD VAF was threefold lower (0.0084% vs 0.0246%; nominal P = .0159) in the quizartinib arm vs the placebo arm (Figure 2A), consistent with a previous observation that the addition of an FLT3 inhibitor to induction chemotherapy results in deeper remissions.19 The percentage of patients in CR with MRD-negative status during the first or second induction cycle was numerically higher in the quizartinib arm (22.4%) vs the placebo arm (13.5%; Figure 2B; MRD cutoff of 0; nominal P = .0739) with a similar trend when analyzing MRD by induction cycle using either the 10−4 cutoff or the 0 cutoff (supplemental Figure 2). A similar trend was found when analyzing MRD by induction cycle. Specifically, after the first induction cycle, the median best FLT3-ITD MRD VAF was 2.8-fold lower (0.0086% vs 0.0242%; nominal P = .0439) with the quizartinib arm vs placebo (Figure 2C), and after the second induction cycle, it was eightfold lower (0.0066% vs 0.0553%; nominal P = .1345). It should be noted that the second induction cycle MRD data in patients with CR correspond to only 23.6% of the overall patients who received 2 induction cycles. The MRD analysis conducted in patients with CRc (supplemental Figures 3 and 4) yielded similar results as in patients with CR (Figure 2; supplemental Figure 2). Among 270 patients who achieved CR by the end of induction and had MRD assessment at the end of consolidation, the median best FLT3-ITD MRD VAF was lower (0% vs 0.0015%; nominal P = .0006) in the quizartinib arm vs the placebo arm (supplemental Figure 5). The same MRD analysis conducted in patients with CRc (supplemental Figure 6) yielded similar results as for patients with CR (supplemental Figure 5). Generally, more patients in the quizartinib arm, with either CR or CRc by the end of induction, had MRD-negative status during consolidation vs the placebo arm using either the 10−4 cutoff or the 0 cutoff (supplemental Figures 7A,C and 8A,C). The MRD-negative status during maintenance is shown in supplemental Figures 7B,D and 8B,D.

Figure 2.

Figure 2.

Open in a new tab

Best FLT3-ITD MRD VAF during induction by treatment arm in patients who achieved CR per IRC by the end of induction. Scatterplots of best FLT3-ITD MRD VAF by the end of induction (A), rates of MRD positivity and negativity using either the 10−4 cutoff or the 0 cutoff (B), and scatterplots of best FLT3-ITD MRD VAF by induction cycle (C). ∗Wilcoxon rank-sum test. †Percentage of patients with FLT3-ITD MRD VAF of >0 among patients in CR with MRD data. ‡Percentage of patients with FLT3-ITD MRD VAF of 0 among patients in CR with MRD data. §Fisher exact test. ||There were 9 patients in the quizartinib arm and 6 patients in the placebo arm with baseline MRD data who did not have MRD data during induction. IQR, interquartile range; NA, not applicable.

Presence of NPM1 comutation and MRD levels

A subgroup analysis of FLT3-ITD MRD levels assessed at the end of induction by NPM1 (nucleophosmin 1) comutation status in patients with CRc showed that quizartinib provides deeper remissions over placebo, regardless of the presence of NPM1 comutation. Among patients with wild-type NPM1, the median best FLT3-ITD MRD VAF was twofold lower (0.0216% vs 0.0494%; nominal P = .8413) in the quizartinib arm vs the placebo arm, and among patients with mutated NPM1, the median best FLT3-ITD MRD VAF was threefold lower (0.0077% vs 0.0258%; nominal P = .0216) in the quizartinib arm vs the placebo arm (supplemental Figure 9). The percentage of patients in CRc with MRD-negative status at the end of induction was higher in the quizartinib arm (25.3% in patients with mutated NPM1 and 17.2% in patients with wild-type NPM1) vs the placebo arm (12.7% in patients with mutated NPM1, and 12.5% in patients with wild-type NPM1; supplemental Figure 9).

Comparison of MRD results from bone marrow aspirates vs peripheral blood

Of all the bone marrow and peripheral blood samples analyzed for MRD in QuANTUM-First, there were a total of 27 cases in which MRD was analyzed from the bone marrow and peripheral blood at the same time point, and in which an FLT3-ITD clone was detected in 1 or both or the paired samples. These 27 paired samples were derived from a total of 18 patients over different time points during therapy. The mean and median VAFs of clones detected in marrow samples were 6.2- and 6.6-fold higher, respectively, compared with the corresponding VAFs derived from the paired peripheral blood samples (P = .0011; Figure 3). In 18 of 27 cases (66.7%), a clone was detected in both samples (Figure 3). There were 8 sample pairs (all collected during the induction phase) in which an FLT3-ITD clone detected in the bone marrow was not detected in the peripheral blood. These 8 sample pairs corresponded to 6 patients (quizartinib, n = 5; placebo, n = 1), 2 of whom had 2 sample pairs meeting these criteria. One patient underwent allo-HCT within 3 months. Four patients treated with quizartinib were still alive without relapses at the time of the data cutoff (>2 years from the relevant sample dates for all 4 patients). Of 2 patients who died, 1 patient treated with quizartinib died because of COVID-19, ∼10 months after the relevant sample date, with the last response assessment conducted 2 years before death showing no relapse; and 1 patient treated with placebo died because of an adverse event related/secondary to AML, ∼28 months after the relevant sample date, with the last response assessment conducted 3 months before death also showing no relapse. Conversely, in 1 sample pair collected during induction (corresponding to 1 patient treated with quizartinib), a clone detected in the peripheral blood was not detected in the bone marrow (Figure 3). This patient started allo-HCT ∼7 months after the relevant sample date and had not relapsed for ∼3 years after the relevant sample date.

Figure 3.

Figure 3.

Open in a new tab

Comparison of MRD VAFs from blood vs BM. Of all MRD samples analyzed in QuANTUM-First, there were 27 samples for which both marrow and blood were analyzed at the same time point (from a total of 18 different patients) and a FLT3-ITD clone was detectable in either or both paired samples. (A) Individual VAF values for each of these 27 samples according to the source, marrow (left) or blood (right). (B) Each individual VAF from the BM, with a line connecting it to corresponding VAF from the blood for that paired sample. In 8 cases, a clone detected in the BM was not detected in the blood. Conversely, in 1 case, a clone detected in the blood was not detected in the BM. BM, bone marrow; PB, peripheral blood.

Prognostic impact of the level of FLT3-ITD MRD

To explore the prognostic impact of the level of FLT3-ITD MRD, we used 2 different cutoff values for FLT3-ITD MRD. The first level used was the limit of quantification for this assay, which is 1 × 10−4. The second cutoff was no detectable clone, that is, 0. Among patients who achieved CR by the end of induction, those who were MRD negative experienced a longer survival benefit than those who were MRD positive, regardless of treatment arm (Figure 4A,D). For all patients irrespective of treatment arm, the 10−4 cutoff was more robust prognostically than the 0 cutoff, with a hazard ratio (HR) of 0.627 (95% confidence interval [CI], 0.427-0.922; Figure 4A) vs 0.789 (95% CI, 0.470-1.323; Figure 4D). However, when examined by treatment arm, quizartinib provided a numerical survival benefit over placebo, regardless of MRD status, with either cutoff (Figure 4B-C,E-F). A similar survival benefit by MRD status conferred by quizartinib over placebo was found among patients who achieved CRc by the end of induction (supplemental Figure 10).

Figure 4.

Figure 4.

Open in a new tab

OS by FLT3-ITD MRD status at the end of induction in patients who achieved CR per IRC by the end of induction. Kaplan-Meier plots of OS in patients who achieved CR by the end of induction, based on MRD status, using either the 10−4 cutoff (A-C) or the 0 cutoff (D-F). ∗Unstratified Cox regression analysis. NR, not reached.

To address the impact of MRD on the likelihood of relapse after completion of consolidation therapy, we examined RFS for patients with CRc in the quizartinib arm and in the placebo arm, who did not undergo allo-HCT in the first CRc (CRc1; Figure 5). In patients treated with quizartinib, MRD negativity assessed by the end of induction translated in improved RFS over MRD positivity using the 10−4 cutoff, with an HR of 0.791 (Figure 5A), whereas the 0 cutoff showed an HR of 1.050 (Figure 5B). MRD negativity assessed within 30 days before starting maintenance showed improved RFS over MRD positivity, as expected, using either the 10−4 cutoff, with an HR of 0.140 (Figure 5E), or the 0 cutoff, with an HR of 0.353 (Figure 5F). In patients treated with placebo, MRD negativity assessed by the end of induction translated in improved RFS over MRD positivity, as expected, using either the 10−4 cutoff, with an HR of 0.529 (Figure 5C), or the 0 cutoff, with an HR of 0.435 (Figure 5D). Similarly, MRD negativity assessed within 30 days before starting maintenance translated into improved RFS over MRD positivity, as expected, using either the 10−4 cutoff, with an HR of 0.404 (Figure 5G), or the 0 cutoff, with an HR of 0.496 (Figure 5H).

Figure 5.

Figure 5.

Open in a new tab

RFS by FLT3-ITD MRD status. RFS by FLT3-ITD MRD status by the end of induction (A-D) and within 30 days before receiving maintenance (E-F) in patients who achieved CRc per IRC and did not undergo allo-HCT in CRc1. Kaplan-Meier plots of RFS in patients who achieved CRc by the end of induction, who did not undergo allo-HCT based on MRD status assessed by the end of induction (A-D) and within 30 days before receiving maintenance (E-H), in the quizartinib (A-B, E-F) and placebo (C-D, G-H) arms, using either the 10−4 cutoff (A, C, E, G) or the 0 cutoff (B, D, F, H). ∗Unstratified Cox regression analysis. NR, not reached.

The RFS analysis in patients with CR by MRD status at the end of induction and by treatment arm showed that among patients with MRD-negative status (when assessed using the 10−4 cutoff), those who received quizartinib had a numerically longer RFS vs those who received placebo (supplemental Figure 11A,C). Among patients with MRD-positive status (when assessed by either cutoff), those who received quizartinib had a numerically longer RFS vs those who received placebo (supplemental Figure 11B,D). Further analysis of cumulative incidence of relapse in patients with CRc shows that those with MRD positivity, based on the 10−4 cutoff or the 0 cutoff, had lower rates of cumulative incidence of relapse at 1, 2, and 3 years in the quizartinib arm vs the placebo arm (supplemental Figure 12).

FLT3-ITD MRD analysis by allo-HCT

We analyzed the MRD status shift from the end of induction to consolidation/maintenance in patients with CR (supplemental Table 2), as well as in patients with CRc (supplemental Table 3), according to allo-HCT status. The rate of patients with CR/CRc who were MRD positive by the end of induction and became MRD negative in consolidation/maintenance was higher with quizartinib vs placebo using either the 10−4 cutoff or the 0 cutoff, regardless of whether they received allo-HCT in the first CR (CR1)/CRc1 or not (supplemental Tables 2 and 3). This difference in favor of quizartinib was particularly evident among patients who did not undergo allo-HCT (supplemental Tables 2 and 3).

We then explored the prognostic impact of the MRD status at the end of induction by allo-HCT in patients with CR (supplemental Figures 13 and 14) as well as in patients with CRc (supplemental Figures 15 and 16). Although the number of patients was low in some of these subgroups, the survival benefit provided by quizartinib over placebo was generally more pronounced among patients who did not undergo allo-HCT in CR1/CRc1, regardless of MRD status (supplemental Figures 13-16).

To assess OS from the time of allo-HCT, according to the latest pre–allo-HCT MRD status and treatment arm, further analyses were conducted in 151 patients who underwent allo-HCT in CR1 as well as in 190 patients who underwent allo-HCT in CRc1 (supplemental Figures 17 and 18). Among patients who received allo-HCT in CR1, quizartinib provided a numerically longer OS vs placebo, irrespective of pre–allo-HCT MRD status (supplemental Figure 17). Among patients who received allo-HCT in CRc1, quizartinib provided a numerically longer OS vs placebo in patients with MRD-positive status, with no differences in those with MRD-negative status (supplemental Figure 18).

FLT3-ITD mutation length and number of ITDs

In QuANTUM-First, the median FLT3-ITD length was 54 base pairs (bp). We compared FLT3-ITD mutation length between the results of the Navigate clinical trial assay (CTA) obtained at screening and the results of the PCR-NGS MRD assay obtained at the end of induction. Among 342 patients (with 417 corresponding ITD sequences) achieving CRc by the end of induction, tested by the CTA and the PCR-NGS MRD assay at screening, concordance in FLT3-ITD detection of exact ITD length was revealed in 329 (96.2%) patients (with 393 corresponding ITD sequences; supplemental Figure 19A). There were 69 (20.2%) patients in whom >1 ITD was detected by both assays (supplemental Figure 19B; 63 patients had 2 ITDs detected by both assays, and 6 patients had 3 ITDs detected by both assays). Additional details on the concordance between the CTA and NGS MRD assay in detecting the same FLT3-ITD sequences at screening and longitudinally are provided in the supplemental Results.

To examine whether ITD length was associated with remission rates and MRD status, we calculated the rates of CR/CRc and the rates of MRD negativity by the end of induction according to FLT3-ITD length assessed at enrollment. We found that patients with ITD length equal or above the median of 54 bp had lower rates of CR and CRc than patients with ITD length below the median, regardless of treatment arm (supplemental Table 4). Within each ITD length subgroup, the patients who received quizartinib generally had numerically higher rates of CR/CRc and rates of MRD negativity (supplemental Table 4). Previous studies have established that mutations longer than 40 to 50 bp in length are associated with both a more distal insertion location in the gene and with worse outcome.20 In most of these studies, patients were treated with chemotherapy without FLT3 inhibition. Therefore, to confirm that longer FLT3-ITD mutations had a similar negative prognostic effect in QuANTUM-First, we first focused on patients in the placebo arm. As shown in Figure 6A, FLT3-ITD length equal or above the median of 54 bp was associated with worse OS for patients treated with chemotherapy only (HR, 0.727; 95% CI, 0.530-0.997), consistent with previous studies examining this issue. When we examined OS for the quizartinib compared with placebo arms, there was a trend for quizartinib to be more effective than placebo, irrespective of FLT3-ITD length (Figure 6B-C). Among patients with CRc, treatment with placebo was associated with longer OS in patients with a single ITD compared with those with multiple ITDs, whereas treatment with quizartinib was associated with longer OS in patients with multiple ITDs than those with a single ITD (supplemental Figure 20).

Figure 6.

Figure 6.

Open in a new tab

OS by FLT3-ITD length at diagnosis. (A) Kaplan-Meier plot of OS in patients on the placebo arm according to FLT3-ITD length. (B) Kaplan-Meier plot of OS comparing placebo vs quizartinib arms for patients with FLT3-ITD lengths below the median of 54 bp in length. (C) Kaplan-Meier plot of OS comparing placebo vs quizartinib arms for patients with FLT3-ITD lengths equal or above the median of 54 bp in length. ∗Median ITD length (54 bp) is calculated based on enrollment assay data (Navigate BioPharma FLT3-ITD mutation assay). †Patients may have only 1 ITD length or >1 ITD length. ‡Unstratified Cox regression analysis. §One patient was randomized in the quizartinib arm based on positive FLT3-ITD results per local laboratory testing, which were not confirmed by central laboratory testing owing to lack of availability of screening sample; for this patient receiving quizartinib, the ITD length was unknown. Therefore, the number of patients on study in panel B could be 264 or 265; similarly, the number of patients on study in panel C could be 274 or 275. NR, not reached.

Discussion

These data represent, to our knowledge, the first validation of the utility of FLT3-ITD MRD monitoring in a prospective, randomized trial of patients with AML undergoing intensive therapy in the context of FLT3 inhibitor therapy. The MRD assay we used was developed specifically for QuANTUM-First, with the intention of testing it prospectively at multiple time points throughout the entire course of treatment. The European LeukemiaNet MRD working group has recommended that AML MRD be assessed after 2 cycles of intensive chemotherapy.21 However, QuANTUM-First was designed in 2015, well before these recommendations. Regardless, it is clear from multiple retrospective studies, and now 2 recent prospective studies, that FLT3-ITD MRD will be an important component of patient management going forward.12,14,16,19,22, 23, 24, 25, 26

The quantitative aspect of the MRD assay clarified the nature of the benefits conferred by quizartinib in this trial. The addition of quizartinib to induction chemotherapy resulted in a deeper CR with respect to the level of FLT3-ITD MRD, and that deeper remission itself was associated with prolonged survival. We also showed that quizartinib treatment can lower the levels of MRD in patients with CRc, regardless of the presence of NPM1 comutation. Data from this analysis are limited to the patients who achieved remission, because MRD analysis was not performed in the intent-to-treat population. Our data demonstrate that patients on the quizartinib arm achieved FLT3-ITD eradication more frequently than those on the placebo arm. Consensus statements from working groups and regulatory authorities have suggested that a cutoff of 10−3 is most appropriate for MRD in AML in general, regardless of methodology or mutation being tracked.21,27 However, FLT3-ITD mutations represent a unique and specific molecular signature for any patient. A recent analysis of the Blood and Marrow Transplant Clinical Trials Network 1506/MORPHO study using a similar MRD detection platform in the context of allo-HCT of patients with FLT3-ITD AML found that any level of MRD detected in the bone marrow negatively affected survival.28 Similarly, recent reports of the Pre-MEASURE study found that the detection of MRD in the blood of patients with FLT3-ITD AML who achieved CR1 before allo-HCT was associated with an increased risk of relapse and death in the posttransplant clinical outcomes.29,30 Specifically, any detectable level of MRD was associated with an increased risk of relapse and death, with VAFs of ≥0.01% (corresponding to ≥10−4 level) having the highest risk.30

FLT3-ITD mutations are rapidly being recognized as a validated MRD marker, and it is imperative that the method used to detect them as MRD be standardized internationally, in the same way as BCR-ABL1 fusion detection is now standardized.31 Any detectable level of FLT3-ITD MRD in this study appeared to influence prognosis; therefore, any FLT3-ITD clone detected by this assay should be viewed by clinicians as indicating an increased risk of relapse. Although the number of paired samples in this study was small, our results from comparing the VAFs of the 27 paired marrow and blood samples indicate that the MRD assay is more sensitive when marrow is used as the source of DNA. Whether or not the additional degree of sensitivity conferred by marrow over blood is necessary for optimal management of patients is unclear at this time.

Per protocol, samples for MRD analysis were collected by the end of induction, and additional sampling during consolidation or maintenance was not required. Consequently, the available MRD data in consolidation or maintenance were limited. Nevertheless, the overall MRD analysis presented here shows the depth of MRD negativity in the quizartinib arm (vs the placebo arm), which may explain the remarkable durability of remission recorded in the quizartinib arm (38.6 months) vs the placebo arm (12.4 months), as previously published.12 Another limitation of the study is that the time point “by the end of induction” during which most patients were evaluated for MRD might have been a too early point to show a strong prognostic discrimination. In retrospect, MRD assessment after 2 courses of treatment (eg, 1 induction plus 1 consolidation, or 2 courses of induction) might have been a more appropriate choice, in line with current recommendations.21

Longer FLT3-ITD insertions have repeatedly been reported to be associated with worse outcome.20,32 The mechanism of this is unclear, but some data suggest that longer insertions result in greater FLT3 autophosphorylation.33 Interestingly, post hoc analysis from the RATIFY trial indicated that patients with AML harboring FLT3-ITD mutations extending into the tyrosine kinase domain (and therefore usually longer) did not benefit from the addition of midostaurin to chemotherapy.17 However, in this work, we found that quizartinib was effective regardless of ITD length. We also found that quizartinib was effective in patients with multiple ITDs and provided longer OS in patients with multiple ITDs than in those with a single ITD.

In summary, this post hoc study of the QuANTUM-First data has provided new insights into how quizartinib, as a potent, selective FLT3 inhibitor, induces deeper remissions and prolongs survival for patients with FLT3-ITD AML. This work provides prospective validation of the use of FLT3-ITD mutations as markers of MRD in the context of intensive chemotherapy in combination with potent FLT3 inhibition for the treatment of FLT3-ITD–mutated AML.

Conflict-of-interest disclosure: M.J.L. reports consulting or advisory role with Daiichi Sankyo, Amgen, Astellas Pharma, Bristol Myers Squibb, AbbVie/Genentech, GlaxoSmithKline, Syndax, and Takeda; provided expert testimony for Novartis; reports research funding from Astellas Pharma (to institution); and reports funds for travel, accommodations, and expenses from Astellas Pharma. H.P.E. reports consulting or advisory role with AbbVie, Agios Pharmaceuticals, Astellas Pharma, Bristol Myers Squibb, Celgene, Daiichi Sankyo, Genentech, GlycoMimetics, ImmunoGen, Incyte, Jazz Pharmaceuticals, Kura Oncology, MacroGenics, Novartis, Pfizer, Servier, Syros Pharmaceuticals, Takeda, and Trillium Therapeutics; reports speakers bureau participation with AbbVie, Agios Pharmaceuticals, Bristol Myers Squibb, Celgene, Incyte, Jazz Pharmaceuticals, Novartis, and Servier; reports research funding from AbbVie, Agios Pharmaceuticals, ALX Oncology, Amgen, Ascentage, Celgene, Daiichi Sankyo, Forma Therapeutics, Forty Seven, Gilead, GlycoMimetics, ImmunoGen, Jazz Pharmaceuticals, Kura Oncology, MacroGenics, Novartis, PTC Therapeutics, Servier, and Sumitomo Dainippon Pharma; and served as steering committee member for GlycoMimetics, chair of the Myeloid Neoplasms Repository study for Bristol Myers Squibb and Celgene, and chair of the independent review committee of the VIALE A and VIALE C studies for AbbVie. P.M. reports consulting or advisory role with AbbVie, Astellas Pharma, BeiGene, Bristol Myers Squibb, Gilead, Incyte, Jazz Pharmaceuticals, Kura Oncology, Menarini, Stemline Therapeutics, Nerviano Medical Sciences, Novartis, Otsuka Pharmaceutical, Pfizer, Ryvu Therapeutics, and Takeda; reports speakers bureau participation with AbbVie, Astellas Pharma, Bristol Myers Squibb, Gilead, Jazz Pharmaceuticals, and Pfizer; and reports research funding from AbbVie, Bristol Myers Squibb, Jazz Pharmaceuticals, Menarini, Stemline Therapeutics, Novartis, Pfizer, and Takeda. R.V. reports consulting or advisory role with AbbVie, Astellas Pharma, Pfizer, and PharmaS; and received payment for lectures from AbbVie, Astellas Pharma, Merck Sharp & Dohme, Novartis, Pfizer, PharmaS, Servier, and Teva. E.P. reports consulting or advisory role with KCR US; received payment for lectures from Amgen, Angelini, Astellas Pharma, Novartis, Pfizer, and Servier; and received funds for travel, accommodations, and expenses from Angelini, Astellas Pharma, Bristol Myers Squibb, Jazz Pharmaceuticals, Novartis, Pfizer, and Servier. H.-J.K. received honoraria from AbbVie, AML Hub, Bristol Myers Squibb, Hando, Novartis, Aston Sci, Amgen, Takeda, Green-Cross Biopharma, AIM BioSciences, Astellas Pharma, Jazz Pharmaceuticals, Janssen, LG Chemical, Pfizer, ViGen Cell, Ingenium, Sanofi, Meiji Pharm, and Merck Sharpe & Dohme; reports consulting or advisory role with Jazz Pharmaceuticals, Novartis, AbbVie, Astellas Pharma, Merck Sharpe & Dohme, Bristol Myers Squibb, Takeda, Sanofi, Handok, and AML Hub; reports speakers bureau participation with Jazz Pharmaceuticals, Takeda, and Novartis; and reports consulting or advisory role with AbbVie, Astellas Pharma, BeiGene, Bristol Myers Squibb, Gilead, Incyte, Jazz Pharmaceuticals, Kura Oncology, Menarini, Stemline Therapeutics, Nerviano Medical Sciences, Novartis, Otsuka Pharmaceutical, Pfizer, Ryvu Therapeutics, and Takeda. J.C. reports consulting or advisory role with AbbVie, Bio-Path, Daiichi Sankyo, Gilead, Forma Therapeutics, Novartis, Pfizer, and Takeda; received payment for lectures from Novartis, Pfizer, and Takeda; reports research funding from AbbVie, Daiichi Sankyo, Novartis, Sun Pharma, and Pfizer; and reports stock options with Bio-Path. M.A.S. reports consulting or advisory role with Bristol Myers Squibb, Kurome Therapeutics, and Novartis; and reports stock options with Kurome Therapeutics. H.D. reports consulting or advisory role with Incyte and Servier. J.W. reports consulting or advisory role with AbbVie; and reports participation on a data safety monitoring committee for AstraZeneca. R.F.S. reports consulting or advisory role with Daiichi Sankyo (for participation on a steering committee) and with AbbVie, Jazz Pharmaceuticals, and Pfizer (for participation on advisory boards); reports payment for lectures from Daiichi Sankyo, Novartis, and Pfizer; serves on a data safety monitoring board or advisory board for BerGenBio and Novartis; and reports research funding from AbbVie, AstraZeneca, Boehringer Ingelheim, Daiichi Sankyo, PharmaMar, Pfizer, and Roche. A.E.P. reports honoraria from Astellas Pharma and Daiichi Sankyo; reports consulting or advisory role with Astellas Pharma, Actinium Pharmaceuticals, Daiichi Sankyo, AbbVie, Forma Therapeutics, Sumitomo Dainippon, Celgene/Bristol Myers Squibb, Syndax, Genentech, BerGenBio, Immunogen, Foghorn Therapeutics, Rigel, and Curis; reports research funding from Astellas Pharma, Bayer, Daiichi Sankyo, Fujifilm, AbbVie, and Syndax (all to institutions); and reports funds for travel, accommodations, and expenses from Daiichi Sankyo. J.E.C.R., K.C.N.C., L.L., K.I., and A.L. are employees of Daiichi Sankyo. Y.M.K. is a former employee of Daiichi Sankyo; reports grants or contracts from Bexon Clinical Consulting and Stemline Therapeutics; and reports consulting fees from Stemline Therapeutics. The remaining authors declare no competing financial interests.

Acknowledgments

The authors thank the patients, their families, and caregivers for their participation in the QuANTUM-First study. The authors further thank the QuANTUM-First steering committee members, the investigators, the study staff, and independent review committee and data monitoring committee members for their important contributions.

This study was sponsored by Daiichi Sankyo, Inc. Medical writing support was provided by Mohamed Abdelmegeed, Emanuela Marcantoni, and Francesca Balordi, of The Lockwood Group (Stamford, CT), in accordance with Good Publication Practice (2022) guidelines, with funding by Daiichi Sankyo, Inc.

Authorship

Contribution: M.J.L. contributed to study design, analyzed the data, and wrote the manuscript; and all other authors contributed to study design, helped analyze the data, and edited the manuscript.

Footnotes

Presented at the 65th annual meeting and exposition of the American Society of Hematology, San Diego, CA, 9 to 12 December 2023.

Anonymized individual participant data from completed studies and applicable supporting clinical study documents are available on request at https://vivli.org/. Data sharing access is provided through the Vivli data sharing portal (https://vivli.org/ourmember/daiichi-sankyo/) after approval of a research proposal and signed data agreement from the research, legal, and intellectual property reviewers of the funder of this study and an independent review panel from the Vivli platform. If clinical study data and supporting documents are provided pursuant to company policies and procedures, Daiichi Sankyo will continue to protect the privacy of the company and clinical study patients.

The full-text version of this article contains a data supplement.

Supplementary Material

Supplemental Methods, Reference, Results, Tables, and Figures
BLOODA_ADV-2025-016444-mmc1.pdf (13.6MB, pdf)

References

  • 1.Nakao M, Yokota S, Iwai T, et al. Internal tandem duplication of the FLT3 gene found in acute myeloid leukemia. Leukemia. 1996;10(12):1911–1918. [PubMed] [Google Scholar]
  • 2.Kottaridis PD, Gale RE, Frew ME, et al. The presence of a FLT3 internal tandem duplication in patients with acute myeloid leukemia (AML) adds important prognostic information to cytogenetic risk group and response to the first cycle of chemotherapy: analysis of 854 patients from the United Kingdom Medical Research Council AML 10 and 12 trials. Blood. 2001;98(6):1752–1759. doi: 10.1182/blood.v98.6.1752. [DOI] [PubMed] [Google Scholar]
  • 3.Kiyoi H, Towatari M, Yokota S, et al. Internal tandem duplication of the FLT3 gene is a novel modality of elongation mutation which causes constitutive activation of the product. Leukemia. 1998;12(9):1333–1337. doi: 10.1038/sj.leu.2401130. [DOI] [PubMed] [Google Scholar]
  • 4.Kiyoi H, Naoe T, Nakano Y, et al. Prognostic implication of FLT3 and N-RAS gene mutations in acute myeloid leukemia. Blood. 1999;93(9):3074–3080. [PubMed] [Google Scholar]
  • 5.Fröhling S, Schlenk RF, Breitruck J, et al. Prognostic significance of activating FLT3 mutations in younger adults (16 to 60 years) with acute myeloid leukemia and normal cytogenetics: a study of the AML Study Group Ulm. Blood. 2002;100(13):4372–4380. doi: 10.1182/blood-2002-05-1440. [DOI] [PubMed] [Google Scholar]
  • 6.Schnittger S, Schoch C, Dugas M, et al. Analysis of FLT3 length mutations in 1003 patients with acute myeloid leukemia: correlation to cytogenetics, FAB subtype, and prognosis in the AMLCG study and usefulness as a marker for the detection of minimal residual disease. Blood. 2002;100(1):59–66. doi: 10.1182/blood.v100.1.59. [DOI] [PubMed] [Google Scholar]
  • 7.Thiede C, Steudel C, Mohr B, et al. Analysis of FLT3-activating mutations in 979 patients with acute myelogenous leukemia: association with FAB subtypes and identification of subgroups with poor prognosis. Blood. 2002;99(12):4326–4335. doi: 10.1182/blood.v99.12.4326. [DOI] [PubMed] [Google Scholar]
  • 8.Pratz K, Cho E, Karp JE, et al. Phase I dose escalation trial of sorafenib as a single agent for adults with relapsed and refractory acute leukemias. J Clin Oncol. 2009;27(suppl 15):7065. [Google Scholar]
  • 9.Burchert A, Bug G, Fritz LV, et al. Sorafenib maintenance after allogeneic hematopoietic stem cell transplantation for acute myeloid leukemia with FLT3-internal tandem duplication mutation (SORMAIN) J Clin Oncol. 2020;38(26):2993–3002. doi: 10.1200/JCO.19.03345. [DOI] [PubMed] [Google Scholar]
  • 10.Stone RM, Mandrekar SJ, Sanford BL, et al. Midostaurin plus chemotherapy for acute myeloid leukemia with a FLT3 mutation. N Engl J Med. 2017;377(5):454–464. doi: 10.1056/NEJMoa1614359. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 11.Perl AE, Martinelli G, Cortes JE, et al. Gilteritinib or chemotherapy for relapsed or refractory FLT3-mutated AML. N Engl J Med. 2019;381(18):1728–1740. doi: 10.1056/NEJMoa1902688. [DOI] [PubMed] [Google Scholar]
  • 12.Erba HP, Montesinos P, Kim HJ, et al. Quizartinib plus chemotherapy in newly diagnosed patients with FLT3-internal-tandem-duplication-positive acute myeloid leukaemia (QuANTUM-First): a randomised, double-blind, placebo-controlled, phase 3 trial. Lancet. 2023;401(10388):1571–1583. doi: 10.1016/S0140-6736(23)00464-6. [DOI] [PubMed] [Google Scholar]
  • 13.Döhner K, Thiede C, Jahn N, et al. Impact of NPM1/FLT3-ITD genotypes defined by the 2017 European LeukemiaNet in patients with acute myeloid leukemia. Blood. 2020;135(5):371–380. doi: 10.1182/blood.2019002697. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 14.Grob T, Sanders MA, Vonk CM, et al. Prognostic value of FLT3-internal tandem duplication residual disease in acute myeloid leukemia. J Clin Oncol. 2023;41(4):756–765. doi: 10.1200/JCO.22.00715. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 15.Hess CJ, Feller N, Denkers F, et al. Correlation of minimal residual disease cell frequency with molecular genotype in patients with acute myeloid leukemia. Haematologica. 2009;94(1):46–53. doi: 10.3324/haematol.13110. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 16.Loo S, Dillon R, Ivey A, et al. Pretransplant FLT3-ITD MRD assessed by high-sensitivity PCR-NGS determines posttransplant clinical outcome. Blood. 2022;140(22):2407–2411. doi: 10.1182/blood.2022016567. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 17.Rücker FG, Du L, Luck TJ, et al. Molecular landscape and prognostic impact of FLT3-ITD insertion site in acute myeloid leukemia: RATIFY study results. Leukemia. 2022;36(1):90–99. doi: 10.1038/s41375-021-01323-0. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 18.Othman J, Potter N, Ivey A, et al. Postinduction molecular MRD identifies patients with NPM1 AML who benefit from allogeneic transplant in first remission. Blood. 2024;143(19):1931–1936. doi: 10.1182/blood.2023023096. [DOI] [PubMed] [Google Scholar]
  • 19.Levis M, Shi W, Chang K, et al. FLT3 inhibitors added to induction therapy induce deeper remissions. Blood. 2020;135(1):75–78. doi: 10.1182/blood.2019002180. [DOI] [PubMed] [Google Scholar]
  • 20.Polak TB, Van Rosmalen J, Dirven S, et al. Association of FLT3-internal tandem duplication length with overall survival in acute myeloid leukemia: a systematic review and meta-analysis. Haematologica. 2022;107(10):2506–2510. doi: 10.3324/haematol.2022.281218. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 21.Heuser M, Freeman SD, Ossenkoppele GJ, et al. 2021 Update on MRD in acute myeloid leukemia: a consensus document from the European LeukemiaNet MRD Working Party. Blood. 2021;138(26):2753–2767. doi: 10.1182/blood.2021013626. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 22.Thol F, Kölking B, Damm F, et al. Next-generation sequencing for minimal residual disease monitoring in acute myeloid leukemia patients with FLT3-ITD or NPM1 mutations. Genes Chromosomes Cancer. 2012;51(7):689–695. doi: 10.1002/gcc.21955. [DOI] [PubMed] [Google Scholar]
  • 23.Bibault JE, Figeac M, Hélevaut N, et al. Next-generation sequencing of FLT3 internal tandem duplications for minimal residual disease monitoring in acute myeloid leukemia. Oncotarget. 2015;6(26):22812–22821. doi: 10.18632/oncotarget.4333. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 24.Levis MJ, Perl AE, Altman JK, et al. A next-generation sequencing-based assay for minimal residual disease assessment in AML patients with FLT3-ITD mutations. Blood Adv. 2018;2(8):825–831. doi: 10.1182/bloodadvances.2018015925. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 25.Blätte TJ, Schmalbrock LK, Skambraks S, et al. getITD for FLT3-ITD-based MRD monitoring in AML. Leukemia. 2019;33(10):2535–2539. doi: 10.1038/s41375-019-0483-z. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 26.Levis MJ, Hamadani M, Logan B, et al. Gilteritinib as post-transplant maintenance for AML with internal tandem duplication mutation of FLT3. J Clin Oncol. 2024;42(15):1766–1775. doi: 10.1200/JCO.23.02474. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 27.Hematologic Malignancies: Regulatory Considerations for Use of Minimal Residual Disease in Development of Drug and Biological Products for Treatment. https://www.fda.gov/media/134605/download
  • 28.Levis MJ, Hamadani M, Logan BR, et al. Measurable residual disease and posttransplantation gilteritinib maintenance for patients with FLT3-ITD-mutated AML. Blood. 2025;145(19):2138–2148. doi: 10.1182/blood.2024025154. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 29.Dillon LW, Gui G, Page KM, et al. DNA sequencing to detect residual disease in adults with acute myeloid leukemia prior to hematopoietic cell transplant. JAMA. 2023;329(9):745–755. doi: 10.1001/jama.2023.1363. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 30.Dillon LW, Gui G, Ravindra N, et al. Measurable residual FLT3 internal tandem duplication before allogeneic transplant for acute myeloid leukemia. JAMA Oncol. 2024;10(8):1104–1110. doi: 10.1001/jamaoncol.2024.0985. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 31.White HE, Salmon M, Albano F, et al. Standardization of molecular monitoring of CML: results and recommendations from the European treatment and outcome study. Leukemia. 2022;36(7):1834–1842. doi: 10.1038/s41375-022-01607-z. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 32.Stirewalt DL, Kopecky KJ, Meshinchi S, et al. Size of FLT3 internal tandem duplication has prognostic significance in patients with acute myeloid leukemia. Blood. 2006;107(9):3724–3726. doi: 10.1182/blood-2005-08-3453. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 33.Liu SB, Dong HJ, Bao XB, et al. Impact of FLT3-ITD length on prognosis of acute myeloid leukemia. Haematologica. 2019;104(1):e9–e12. doi: 10.3324/haematol.2018.191809. [DOI] [PMC free article] [PubMed] [Google Scholar]

Associated Data

This section collects any data citations, data availability statements, or supplementary materials included in this article.

Supplementary Materials

Supplemental Methods, Reference, Results, Tables, and Figures
BLOODA_ADV-2025-016444-mmc1.pdf (13.6MB, pdf)

Articles from Blood Advances are provided here courtesy of The American Society of Hematology

RESOURCES