Clinical practice recommendation on hematopoietic stem cell transplantation for acute myeloid leukemia patients with FLT3-internal tandem duplication: a position statement from the Acute Leukemia Working Party of the European Society for Blood and Marrow Transplantation

Abstract

The FMS-like tyrosine kinase 3 (FLT3) gene is mutated in 25-30% of patients with acute myeloid leukemia (AML). Because of the poor prognosis associated with FLT3-internal tandem duplication mutated AML, allogeneic hematopoietic stem-cell transplantation (SCT) was commonly performed in first complete remission. Remarkable progress has been made in frontline treatments with the incorporation of FLT3 inhibitors and the development of highly sensitive minimal/measurable residual disease assays. Similarly, recent progress in allogeneic hematopoietic SCT includes improvement of transplant techniques, the use of haploidentical donors in patients lacking an HLA matched donor, and the introduction of FLT3 inhibitors as post-transplant maintenance therapy. Nevertheless, current transplant strategies vary between centers and differ in terms of transplant indications based on the internal tandem duplication allelic ratio and concomitant nucleophos-min-1 mutation, as well as in terms of post-transplant maintenance/consolidation. This review generated by international leukemia or transplant experts, mostly from the European Society for Blood and Marrow Transplantation, attempts to develop a position statement on best approaches for allogeneic hematopoietic SCT for AML with FLT3-internal tandem duplication including indications for and modalities of such transplants and on the potential optimization of post-transplant maintenance with FLT inhibitors.

Introduction

FMS-like tyrosine kinase 3 (FLT3) is a transmembrane ligand-activated receptor tyrosine kinase that is normally expressed by hematopoietic stem cells and early myeloid and lymphoid progenitor cells, and is involved in the proliferation, differentiation and apoptosis of hematopoietic cells1 through various signaling pathways, including phosphatidylinositol 3-kinase (PI3K) and rat sarcoma (RAS) signal-transduction cascades.72 FLT3 is mutated in about 25-30% of newly diagnosed cases of acute myeloid leukemia (AML),108 either by internal tandem duplications (FLT3-ITD) of the juxtamembrane domain (19-25%), and/or by a point mutation, usually involving the tyrosine kinase domain (TKD) at D835 or I836 in the activating loop (7-10%).1311 Both mutations are more frequent in cytogenetically normal AML and both constitutively activate FLT3 causing dimerization in a ligand-independent manner, resulting in proliferation and survival of leukemia cells.1514

FLT3-ITD mutations in newly diagnosed AML are associated with a greater disease burden, manifesting as an elevated white blood cell count and a high percentage of blasts at the time of diagnosis as well as a tendency to early relapse and a poor overall prognosis.171612108 Both European LeukemiaNet (ELN) recommendations and National Comprehensive Cancer Network (NCCN) guidelines incorporate FLT3-ITD mutations in risk-stratifying patients based on allelic burden and nucleophosmin-1 (NPM1) co-mutation.1918 In cytogenetically normal patients, FLT3-ITD mutations in the presence of a concomitant NPM1 mutation, mainly when the FLT3-ITD allele ratio is low (<0.5), fare better than those with wild-type NPM1.22201716108 Despite the great effort to harmonize and cross-validate the FLT3 assays within clinical trials,23 there is still no consensus on the FLT3-ITD allele ratio threshold and there is considerable variability between centers in the assessment of the FLT3-ITD ratio according to the technique used, if one is available. Furthermore, in addition to NPM1 mutations, a significant overlap with other mutations (WT1, IDH1, DNMT3A) as well as NUP98/NSD1 fusions modify outcome as well as response to therapy. Although patients with FLT3-ITD AML respond to conventional induction chemotherapy with remission rates similar to those seen in other subtypes of AML, they are much more likely to relapse and to relapse quickly.28241211 The prognostic impact of FLT3-TKD is less clear,3229 but it, too, is influenced substantially by NPM1 co-mutation which confers a better prognosis.3533

The availability of active FLT3 inhibitors that are able to disrupt the oncogenic signaling initiated by FLT3 has improved the overall survival (OS) of patients with FLT3-mutated AML.36 Midostaurin, a multikinase inhibitor, was granted Food and Drug Administration (FDA) and European Medicines Agency (EMA) approval for the treatment of patients with newly diagnosed FLT3-mutated AML, in combination with intensive chemotherapy, and by the EMA in addition as maintenance treatment after conventional consolidation therapy. This approval was based on the results of the RATIFY trial, which demonstrated that the combination of midostaurin with standard induction therapy resulted in significantly prolonged OS (not censored for transplant) for AML with either FLT3-ITD or FLT3-TKD mutations.37 The benefit was particularly remarkable in patients who went on to receive allogeneic hematopoietic stem cell transplantation (allo-SCT) in first complete remission (CR1). Following the results of the ADMIRAL trial, gilteritinib, a second-generation FLT3 inhibitor, was recently approved for relapsed/refractory FLT3-mutated AML with FLT3-ITD and FLT3-TKD mutations.38 Promising data were also reported for quizartinib and crenolanib.4039 Finally, because of its long-time availability, sorafenib has been tested, alone or in combination, in various settings in FLT3-ITD AML, such as first-line therapy4241 or for the treatment of disease relapse,4543 including after failure of allo-SCT.5745 However, recent data appear to support incorporating sorafenib into the treatment of patients with FLT3-mutated AML, possibly with induction therapy595841 as well as maintenance therapy after allo-SCT.656043

Because of the diversity in FLT3-mutated AML, which depends on the type of FLT3 mutation, FLT3-ITD allelic burden, insertion site and co-occurring mutations, the decision regarding whether to perform allo-SCT in CR1 is becoming more challenging.7566 With the use of more effective therapies, especially with the incorporation of FLT3 inhibitors, deeper responses are being achieved. The assessment of minimal/measurable residual disease (MRD) at the time of response has enabled prediction of outcomes in AML, and tailoring of post-remission therapeutic strategies accordingly.7876 Additionally, substantial progress has been made in allo-SCT in recent years, including improvement of transplant techniques, the use of haploidentical donors in patients lacking an HLA-matched donor,8179 and post-transplant preventive strategies, such as prophylactic or preemptive use of FLT3 inhibitors.858263 Nevertheless, current transplant strategies vary between centers and differ in terms of indications for the transplants and treatments following them. This review provides a consensus from European Society for Blood and Marrow Transplantation (EBMT) experts on best approaches to allo-SCT in AML with FLT3-ITD including the indications for and modalities of allo-SCT and on potential optimization of post-transplant maintenance therapy with FLT3 inhibitors.

The consensus process

Two chairpersons (AB and MM) appointed a panel of 32 physicians (hereafter referred to as the Panel) selected mostly from the EBMT) for their expertise in research and clinical practice in AML and allo-SCT. A physician with expertise in clinical epidemiology (ML) ensured the methodological correctness of the process. The objective of the Panel was to identify practical issues pertinent to all physicians involved in the therapeutic management of patients undergoing allo-SCT for AML with FLT3 mutations and to generate best practice recommendations on indications for and modalities of allo-SCT and on potential optimization of post-transplant maintenance with FLT3 inhibitors. This was done through a number of questions according to the Delphi technique.86 A search for relevant literature in English was performed in the MEDLINE, EMBASE and PubMed databases (up to August 2019). Most of the studies used for these recommendations are retrospective cohort studies or phase II trials, with only a few prospective randomized trials. Three panelists drafted statements that addressed the key questions identified, and the remaining panelists scored their agreement with those statements and provided suggestions for rephrasing them.

The evaluation of evidence and the subsequent recommendations were graded according to the system used by Couriel.87 The strength of the recommendations (Online Supplementary Table S1) and evidence levels (Online Supplementary Table S2) were rated by all participants of the consensus process.

Overview of prognosis and current indications for allogeneic stem-cell transplantation in FLT3-mutated acute myeloid leukemia

The indication for allo-SCT in FLT3-ITD AML depends largely on FLT3 variables (allelic burden, insertion site and co-occurring mutations), on disease status (including MRD), and on the use of FLT3 inhibitors during induction/consolidation treatment, in addition to other patient-, donor- and graft-related factors. Unfortunately, there are no prospective randomized trials evaluating the best post-remission therapeutic strategy in FLT3-mutated AML, taking in consideration all the diverse combinations.

Several recent reports have suggested that allele burden might affect prognosis of FLT3-ITD AML treated with standard induction chemotherapy.89882217 Indeed, the presence of a high allelic burden of FLT3-ITD mutations (≥0.5) confers a poor prognosis.91902712 Several studies have demonstrated that allo-SCT significantly improves survival outcomes in this category959269 and that the negative impact of high allele burden might be overcome when patients undergo allo-SCT in CR1.17 Therefore, all patients with FLT3-ITD should be considered for allo-SCT in CR1.96926966 These patients still face higher rates of early relapse and poor responses to further therapy and eventually poor long-term survival.9792 The worst prognosis is observed in patients who relapse after allo-SCT, who have predicted 1-year OS rates below 20%.98 However, a subcategory of patients with FLT3-ITD/NPM1 mutation of the ELN intermediate-risk group treated with FLT3 inhibitors, and who achieve MRD negativity, may be offered the possibility of post-remission consolidation with longitudinal MRD monitoring of NPM1.91 This approach should be undertaken with caution, and preferably within a clinical trial, since recent data suggest the possible extinction of the NPM1 clone after chemotherapy while the FLT3-ITD clone persists.

Additional mutations may, however, influence the prognosis of AML with FLT3-ITD. For example, the co-existence of NPM1 mutation with FLT3-ITD is associated with improved outcomes, particularly in patients with a low FLT3 allelic ratio (<0.5).2016108 According to the 2017 ELN recommendations, this subcategory is stratified as favorable risk, advocating against the need for allo-SCT.91 Nonetheless, the good prognosis of a low allelic ratio is not universally recognized, with data suggesting better outcome for allografted patients regardless of NPM1 mutation status.99 A threshold for FLT3 allelic burden is also controversial and differs according to studies. It was mainly based on the median of the mutant-to-wildtype ratio found in different retrospective studies. For example, in one study evaluating the prognostic factors of newly diagnosed AML, a FLT3 ratio above 0.78 was associated with worse survival, whereas in another study the threshold was 0.51.1711 Therefore, the allelic burden has a continuous effect on survival outcomes and a ratio of 0.5 is a chosen threshold based on maximum clinical prognostic data. With the advent of FLT3 inhibitors in the frontline treatment of FLT3-mutated AML, the OS has improved regardless of the allelic burden and the use of allo-SCT. Whether NPM1-mutant FLT3-ITD AML warrants post-remission allo-SCT in CR1 or not is still debatable. Although some studies analyzing the effect of allo-SCT in patients with NPM1-mutant FLT3-ITD found no improvement in OS or relapse risk, we must take into consideration the retrospective nature of the analysis and the small number of patients with a non-statistically significant improvement in OS and relapse risk.2217 Interestingly, patients with newly diagnosed AML with NPM1-mutant FLT3-ITD treated with frontline midostaurin and intensive chemotherapy, had a 3-year OS rate of around 75%. In a retrospective subgroup analysis, the benefit of allo-SCT was only seen in the adverse ELN subgroup [hazard ratio (HR)=0.39; P=0.003], but not in the favorable (HR=0.78; P=0.62) and intermediate risk subgroups (HR=0.81; P=0.53).91 These findings should, however, be interpreted with caution as the RATIFY trial was not powered to demonstrate a difference of benefit of allo-SCT among diverse FLT3-ITD/NPM1 genotypes. For example, the total number of patients in the favorable ELN subgroup was 85 and these patients were divided into four small groups according to whether they did or did not receive midostaurin and/or allo-SCT in CR1.91

The deleterious effect of FLT3-ITD was most clinically relevant in patients with concomitant NPM1 and DNMT3A mutations, suggesting that AML patients with NPM1, FLT3-ITD and DNMT3A mutations (triple-positive AML) should be transplanted regardless of the FLT3-ITD allelic ratio.8 A recent study conducted on 147 patients found that NPM1-positive AML with low allelic FLT3-ITD still had an unfavorable outcome, with an OS rate of only 41%, but with significant improvements in both relapse-free survival (RFS) and OS for those allografted in CR1.99 This challenges the notion of withholding transplant for patients with supposedly favorable outcomes. In that sense, a recent study from the MD Anderson Cancer Center showed that allo-SCT improved leukemia-free survival (LFS) and OS independently of the FLT3-ITD allelic ratio and NPM1 mutation status.100 This fits with recent NCCN guidelines still offering allo-SCT for all patients with FLT3-ITD mutations regardless of allelic ratio or NPM1 mutation status.18

On the other hand, patients with a low allelic ITD ratio lacking an NPM1 mutation (and lacking other adverse risk mutations) are currently considered intermediate risk, hence in a gray prognostic area with no proper consensus on optimal treatment strategy. There is conflict regarding the current practice between proceeding to allo-SCT for these patients or limiting allo-SCT only to those who do not achieve MRD negativity by multiparametric flow cytometry. Indeed, technical limitations prevent the use of FLT3 mutation for assessment of MRD which must therefore rely on multiparametric flow cytometry.101 Finally, Versluis et al. reported that in patients with wildtype NPM1 AML without FLT3-ITD or with a low allelic ratio of FLT3-ITD, reduced intensity conditioning allo-SCT resulted in better OS and RFS rates as compared with chemotherapy or autologous SCT.89

Overall, limitations to the universal incorporation of FLT3-ITD allelic ratio into routine clinical practice and the treatment algorithm include the lack of a clear cut-off (0.5 in the ELN recommendations, 0.7 in the RATIFY study) and the potential variability of the allelic ratio over time. A global effort is needed to standardize the technique for determining the FLT3-ITD allelic ratio, making it universal with calibration of all laboratories, reminiscent of the global exercise the world did for BCR/ABL1. Similarly, the definition of high and low allelic ratio should also be standard-ized with a clear consensus on a cut-off level. Until these technical challenges are addressed, the transplant indication remains controversial in patients with FLT3-ITD who belong to the ELN favorable risk group (low allelic ratio <0.5 with concomitant NPM1 mutation) and who achieve MRD negativity. Many European cooperative groups follow the ELN algorithm, deferring allo-SCT in patients with NPM1-mutant FLT3-ITD, unless there is molecular persistence of NPM1. Thus, performing MRD assessment regularly to decide on allo-SCT timing is crucial when selecting this approach. Conversely, the NCCN guidelines are still advocating allo-SCT in CR1 in this setting.

Finally, data on the prognosis of FLT3-TKD AML remain conflicting, with some studies suggesting a negative impact of TKD mutations on LFS and OS,302511 while others suggesting no prognostic effect, or even a benefit when a NPM1 mutation is present.35343229

Hematopoietic stem cell transplantation and factors predictive of outcome

As stated above, because of the poor prognosis associated with FLT3-ITD mutated AML, allo-SCT was most frequently performed in patients in CR11027466 including fit patients ≥60 years of age.103 In most studies, the LFS rate at 2 years ranges between 50 to 60% in that setting,104979266 although a wide variation from 20%10570 to 70%69 has been reported. There are knowledge gaps about the factors that can predict outcome after allo-SCT.

A previous EBMT study97 reported that patients with FLT3-ITD mutated AML with concomitant mutated NPM1 had better post-transplant outcomes compared to those with wildtype NPM1. Similarly, other studies reported that the presence of active disease or MRD before allo-SCT results in poor post-transplant outcomes.107106

A recent, large EBMT registry study assessed outcomes in 462 allografted FLT3-mutated AML patients with a median follow-up of 39 months for alive patients.63 Forty percent received allo-SCT from matched related donors, 49% from matched unrelated donors and 11% from haploidentical donors. Two-year cumulative incidence of relapse (CIR) and non-relapse mortality rates were 34% and 15%, respectively, whereas LFS, OS and graft-versus-host disease (GvHD)-free. relapse-free survival (GRFS) rates were 51%, 59% and 38%, respectively. On multivariable analysis, the need for more than one induction treatment negatively affected outcome, while prescribing an allo-SCT in CR1 resulted in improved CIR, LFS and OS. Presence of an NPM1 mutation was also associated with better outcomes, including better CIR, LFS, OS and GRFS. Post-transplant maintenance therapy with sorafenib significantly reduced the CIR and improved LFS, OS and GFRS. Outcomes were not affected by the type of donor or conditioning intensity. An important finding from this study was that in vivo T-cell depletion with antithymocyte globulin decreased chronic GvHD and significantly improved LFS, OS and GRFS, without an apparent increase in the risk of relapse. This indicates that, even in the setting of FLT3-mutated AML, in vivo T-cell depletion does not appear to abrogate the graft-versus-leukemia effect. Finally, the use of haploidentical donors was associated with improved GRFS compared to that achieved with other types of donors. Given the high risk of rapid relapse of patients with FLT3-mutated AML in CR1 and the poor outcome of allo-SCT in CR2 or beyond,1081211 these results and those of a recent EBMT study suggest that, in the absence of a matched sibling donor, performing haploidentical transplants in CR1 may be considered.109 Furthermore, in another large EBMT study on more than 6,500 adult AML patients allografted in CR1, multivariate analysis confirmed the lack of a statistically significant difference in OS following transplants from matched related donors or 10/10 matched unrelated donors, or haplo-SCT.110 Finally, the results of a CIBMTR, EUROCORD and EBMT collaborative analysis demonstrated that outcomes after umbilical cord blood transplantation are similar to those after allo-SCT from sibling donors for patients with FLT3-ITD AML.110

Post-transplant maintenance in FLT3-mutated acute myeloid leukemia

Even after allo-SCT, FLT3-mutated AML is associated with a higher risk of early relapse (30%-59%) compared to FLT3-wildtype AML.9282 Indeed, in a CIBMTR analysis of 511 patients (158 with FLT3 mutations), there was an increase in relapse rates in FLT3-mutated AML (38% vs. 28%; P=0.04; relative risk 1.60; 95% CI: 1.15-2.22).74 Satisfactory treatment of patients with FLT3-mutated AML who relapse or progress after allo-SCT, is an unmet need. Chemotherapy or FLT3 inhibitors alone or combined with donor lymphocyte infusions are rarely effective in the long term,5045 even though a small proportion of patients who relapse after allo-SCT can achieve long-lasting responses with sorafenib.57555452 A second allo-SCT can be offered to only a small percentage of patients and is associated with a rather high non-relapse mortality rate.111 Several studies have, therefore, investigated the use of post-transplant maintenance with FLT3 inhibitors as a strategy aimed to reduce relapse after allo-SCT.112

Midostaurin was not offered as maintenance therapy to recipients of allo-SCT in the RATIFY study,113 but the RADIUS phase II randomized trial compared post-transplant midostaurin maintenance with standard care in 60 adult patients.114 Estimated relapse rates at 18 months were 24% in the standard care group and 11% in the midostaurin group (P=0.27).114 In another prospective phase II study, maintenance midostaurin was also offered to FLT3-mutated AML patients undergoing allo-SCT in CR1. In a landmark analysis in patients who were event free at day 100 after transplant (n=116), those who started maintenance therapy within 100 days after their transplant (n=72) had a significantly better OS than those who did not.115 The main cause of early discontinuation of maintenance midostaurin after allo-SCT (23%) was poor tolerability, mainly as a result of gastrointestinal toxicity.114

Sorafenib has been studied as maintenance therapy following allo-SCT, demonstrating benefit with regards to survival and improved outcomes in a phase I study, a pilot study, a single-center study, a multicenter study, a registry study and a randomized study.1166560 A phase I trial (NCT01398501) was conducted in which 22 FLT3-ITD AML patients received twelve 28-day cycles of sorafenib 45-120 days after allo-SCT.61 The maximum tolerated dose was established at 400 mg twice daily. The 1-year progression-free survival (PFS) rate was 85% with a corresponding 1-year OS of 95%. In a pilot study, six patients with FLT3-ITD AML received sorafenib (n=5 maintenance, n=1 salvage) after allo-SCT with similarly encouraging results.116 Five of these patients developed cutaneous corticosteroid-sensitive GvHD within a few days after sorafenib initiation, suggesting a possible immunomodulatory effect, and remarkably all patients had sustained molecular remissions.

In a single-institution, observational study on FLT3-ITD AML patients transplanted in CR1, 26 patients who received sorafenib as maintenance treatment after allo-SCT were compared to 55 historical controls who did not.62 The sorafenib cohort had a better 2-year OS rate (81% vs. 62%), improved PFS (82% vs. 53%), and lower relapse incidence (8% vs. 38%).

In a multicenter study, 27 FLT3-mutated AML patients (aged 15-57 years) received maintenance therapy with sorafenib as a single agent after allo-SCT.60 At a median follow-up of 18 months, 25 patients were in complete remission with full donor chimerism, with 1-year PFS and OS rates reaching 92%. Updated results after a median follow-up of 40 months further demonstrated favorable long-term outcomes in patients receiving sorafenib maintenance therapy, with 2-year PFS and OS rates reaching 73% and 80%, respectively, with an acceptable toxicity profile.65

A recent large EBMT registry study assessed outcomes in 462 allografted FLT3-mutated AML patients over a median follow-up of 39 months for surviving patients.63 Twenty-eight patients received post-transplant sorafenib maintenance treatment, initiated at a median of 55 days after transplantation (range, 1-173) at a median dose of 800 mg/day (range, 200-800 mg/day). Thirteen patients in the sorafenib group had chronic GvHD at a median time of 76 days after the initiation of sorafenib (range, 9-194 days). Chronic GvHD was limited in seven patients and extensive in six. On multivariate analysis, post-transplant maintenance with sorafenib significantly reduced the relapse incidence (HR=0.39; P=0.05), and improved LFS (HR=0.35; P=0.01), OS (HR=0.36; P=0.03) and GFRS (HR=0.44; P=0.02). Matched-pair analysis was also performed on 52 patients (26 in the sorafenib group and 26 controls) who engrafted and survived after allo-SCT with no relapse or grade II-IV acute GvHD until sorafenib initiation. The 2-year LFS and OS rates were 79% and 83%, respectively, in the sorafenib group (P=0.02) versus 54% and 62%, respectively, in the controls (P=0.007).

More recently, preliminary conclusions of a double-blind, prospective trial (SORMAIN) that randomized patients to either maintenance treatment with sorafenib or placebo introduced during the first 60-100 days after allo-SCT provided further support for the use of this drug in this high-risk setting.64 Eighty transplanted FLT3-ITD adult AML patients were randomized 1:1 to receive either sorafenib (up to 400 mg twice daily) or placebo for up to 24 months. After a median follow-up of 42 months, the median RFS was 31 months in the placebo group whereas it was ‘not reached’ in the sorafenib group (corresponding to a 2-year RFS of 53% vs. 85%: HR 0.39; P=0.01). Sorafenib was well-tolerated with toxicities that were generally manageable, mostly by dose reduction. These findings build on previously reported data and confirm that sorafenib maintenance therapy after allo-SCT in FLT3-ITD AML patients is both safe and efficient in significantly reducing CIR and improving survival.

In addition to sorafenib’s direct anti-leukemic effect, a possible synergism between the drug and alloreactive donor T cells in facilitating long-term disease control has been suggested,117 and has also been proposed in murine models in which sorafenib apparently exacerbated GvHD.118 A recent study demonstrated that sorafenib promotes graft-versus-leukemia activity in mice and humans through interleukin-15 production in FLT3-ITD leukemia cells.119

Gilteritinib is also currently being prospectively assessed for maintenance use in FLT3-ITD AML after allo-SCT in a phase III, randomized, double-blind, placebo-controlled multicenter trial (NCT02997202).120 This study aims to enroll 346 adult patients with AML in CR1, randomized 1:1, to receive either gilteritinib 120 mg or placebo for 2 years. In addition, a large phase III randomized study (NCT04027309) by a consortium of several cooperative study groups, including HOVON, AMLSG, SAKK, ALFA, CETLAM, PETHEMA, FILO and ALLG, is anticipated to start enrolling by the end of 2019: patients will be randomized to midostaurin or gilteritinib added to standard induction and consolidation treatment. Patients who achieve complete remission will continue maintenance with either midostaurin or gilteritinib.

Finally, the recent approval of midostaurin for frontline treatment of FLT3-mutated AML in the USA and Europe may challenge the role of post-transplant maintenance therapies, including sorafenib. Accordingly, new data should be generated in this setting.122121 Most FLT3-mutat ed AML patients, however, are not currently receiving midostaurin, at least outside the USA and some other countries; therefore, for the foreseeable future, patients may still benefit from sorafenib maintenance treatment after allo-SCT.

Summary of position statement (Table 1)

Table 1.Summary of the European Society for Blood and Marrow Transplantation position statement on allogeneic hematopoietic stem-cell transplantation in FLT3-internal tandem duplication acute myeloid leukemia.

1- Indications for allogeneic stem-cell transplantation in FLT3-internal tandem duplication acute myeloid leukemia

• The indication for allo-SCT is controversial in patients with FLT3-ITD who belong to the ELN favorable risk group (low allelic ratio <0.5 with concomitant NPM1 mutation) and who achieve MRD negativity. Allo-SCT may be delayed until first relapse as recommended by the ELN or performed in CR1 as allowed by NCCN guidelines. Grade level C-II
• In general, all other patients with FLT3-ITD AML should be considered for allo-SCT in CR1 if feasible. Grade level B-II

2- Modalities of hematopoietic stem cell transplantation

• Donors should be selected according to EBMT general guidelines⁸³ including the potential use of cord blood grafts whenever indicated. Grade level B-II
• In vivo T-cell depletion decreases the risk of chronic GvHD, without apparently increasing the risk of relapse, in FLT3-ITD AML and is therefore an option in this setting. Grade level B-II
• The choice of conditioning has no direct link with FLT3-ITD mutation and should be adapted to other individual risk factors such as age, disease status at transplant, and donor type. Grade level B-II

3- Post-transplant maintenance for FLT3-internal tandem duplication acute myeloid leukemia

• Post-transplant maintenance therapy with a FLT3 inhibitor for patients who have undergone allo-SCT for FLT3-ITD AML is recommended (except for patients with active acute GvHD). In the absence of an appropriate clinical trial, sorafenib could be considered as the preferred option, but other FLT3 inhibitors are attractive and warrant further investigation in larger prospective studies. Grade level B-II
• Maintenance therapy should be initiated as soon as possible after disease evaluation, including MRD assessment (whenever feasible), especially in patients with MRD-positive AML before or after allo-SCT, provided there is adequate hematologic reconstitution. Grade level B-II
• Sorafenib should be transiently discontinued in the case of GvHD requiring systemic treatment with corticosteroids, but may be cautiously resumed once remission of GvHD is documented. Grade level B-III
• If choosing sorafenib, the recommended post-transplant maintenance dose is 400 mg/day in two divided doses. Patients with MRD-positive disease may receive 800 mg/day in two divided doses, to be adapted according to tolerance. Grade level B-III
• One potential challenge is the lack of approval of sorafenib for AML and its off-label use may not be reimbursed in many/most countries. Ongoing studies will determine the role and modalities of use of midostaurin, gilteritinib or other FLT3 inhibitors in this setting.
• The duration of maintenance therapy is not firmly established, but a minimum of 2 years is recommended, depending on tolerance. Grade level B-III
• Monitoring is recommended for potential drug-drug interactions and long-term side effects.

Aspects to be resolved

• Standardization of FLT3-ITD allelic ratio in terms of technique and cut-off level
• Indication for allo-SCT in patients with FLT3-ITD AML who belong to the ELN intermediate risk group (high allelic ratio ≥0.5 with concomitant NPM1 mutation) and who achieve MRD negativity.
• Time of withdrawal of immunosuppression
• Pre-emptive versus prophylactic donor lymphocyte infusion
• Post-transplant maintenance with FLT3 inhibitors outside FLT3-ITD AML (immunomodulatory and off-target effects)
• Impact of post-transplant maintenance therapy on immune reconstitution and environment
• Combination of post-transplant FLT3 inhibitors with other drugs such as hypomethylating agents
• Monitoring of patients receiving post-transplant FLT3 inhibitors for potential extramedullary relapse or aggressive clone selection.

Footnotes

• Check the online version for the most updated information on this article, online supplements, and information on authorship & disclosures: www.haematologica.org/content/105/6/1507

• Received November 19, 2019.
• Accepted March 19, 2020.