In this issue of Blood Advances, Levis et al1 report a post hoc analysis of data from the Blood and Marrow Transplant Clinical Trials Network 1506 (MORPHO) clinical trial (NCT02997202), demonstrating the importance of NPM1 comutation status in determining the role of allogeneic hematopoietic cell transplantation (HCT) conditioning intensity and the utility of post-HCT gilteritinib maintenance therapy in acute myeloid leukemia (AML) with fms-like tyrosine kinase 3 internal tandem duplication (FLT3-ITD). In the MORPHO study, patients with AML with FLT3-ITD were randomly assigned to receive the FLT3 inhibitor gilteritinib or placebo, after HCT was performed in first complete remission (CR).2,3 The study demonstrates that gilteritinib improved the primary end point of relapse-free survival (RFS) exclusively in patients with detectable FLT3-ITD–based measurable residual disease (MRD) either before or after HCT.2,3
The new analysis reported here considers 2 further variables, conditioning regimen intensity and NPM1 comutation at diagnosis, in addition to peri-HCT MRD. Preparative regimens, chosen by investigators, were categorized as myeloablative conditioning (MAC) or reduced intensity conditioning (RIC) using Center for International Blood and Marrow Transplantation guidelines. MRD was defined as detection of FLT3-ITD alleles in DNA by tandem polymerase chain reaction (PCR) and next-generation sequencing (NGS). There were 3 salient findings. (1) MAC was not better than RIC at eradicating pre-HCT FLT3-ITD MRD. (2) The previously demonstrated benefit of gilteritinib in patients with FLT3-ITD MRD peri-HCT was driven almost entirely by patients who had NPM1 comutations at diagnosis. (3) MAC was superior to RIC in preventing relapse only in patients who did not have NPM1 comutations at diagnosis or detectable FLT3-ITD MRD peri-HCT. Despite the inherent caveats associated with a retrospective subgroup analysis, the data make a compelling case that both HCT conditioning and post-HCT maintenance can be optimized by considering AML biology and disease burden.
FLT3-ITD causes constitutive activation and aberrant signaling of the FLT3 receptor tyrosine kinase. It is present in AML in ∼30% of patients, most commonly with a normal karyotype. Outcomes in patients with AML with FLT3-ITD were historically poor due to rapid relapses after initial treatment responses, but are significantly improved with HCT in first CR.4 Incorporation of the FLT3 inhibitors midostaurin or quizartinib into cytotoxic chemotherapy–based treatment regimens has had limited impact on CR rate, but has significantly improved RFS and overall survival in placebo-controlled randomized trials.5,6 Arguably, the primary benefit of adding FLT3 inhibitors to cytotoxic chemotherapy in AML with FLT3-ITD may be to delay relapse and allow time to proceed to HCT in first CR. The MORPHO study then provided data supporting rational use of the FLT3 inhibitor gilteritinib as maintenance therapy after HCT by demonstrating that gilteritinib improved the primary end point of RFS exclusively in patients with detectable MRD either before or after HCT.2,3
AML with NPM1 mutation (NPM1m) has become recognized as a distinct entity in recent years. Mutations in NPM1, encoding the nucleocytoplasmic shuttling protein nucleophosmin, are present in AML in approximately a third of patients at diagnosis. They are characteristically mutations in exon 12 that cause a frameshift in the C-terminal region of the protein, resulting in its aberrant restriction to the cytoplasm ("NPM1c"), with consequent upregulation of homeobox genes. AML with NPM1m is associated with favorable treatment responses. Although NPM1 mutations rarely co-occur with adverse-risk cytogenetic abnormalities or with MDS-related adverse risk mutations (eg, RUNX1, ASXL1), they commonly co-occur with FLT3 mutations, and ameliorate treatment response in patients with FLT3-ITD,7 providing the rationale for choosing NPM1m status as a determinative covariate in the current study.
In the present study, patients who had both NPM1 mutations and FLT3-ITD saw a large and sustained benefit from post-HCT gilteritinib maintenance. In contrast, patients who did not have NPM1 comutations had only limited benefit from post-HCT gilteritinib. AML with NPM1m appears unable to escape inhibition of comutant FLT3-ITD following HCT. The next steps in approaching relapse risk in non-NPM1m AML with FLT3-ITD could be informed by knowing how many patients who relapsed despite gilteritinib relapsed with FLT3-ITD (suggesting capacity to overcome FLT3 inhibition) and how many relapsed with wild-type FLT3 (suggesting capacity to find a FLT3-independent treatment resistance pathway), as well as characterizing new mutations present at relapse.
The finding here that higher conditioning intensity fails to benefit patients with FLT3-ITD MRD is mostly in agreement with prior studies. Hourigan et al8 found that patients with FLT3-ITD MRD had near-universal dismal outcomes, which were not improved by MAC. Similarly, Loo et al9 showed that pre-HCT detection of FLT3-ITD MRD, even at very low levels, using high-sensitivity PCR-NGS-based approaches was associated with a dismal prognosis which was not ameliorated by MAC HCT. Dillon et al10 found a dose-dependent relationship between FLT3-ITD burden and overall survival. Although they found that MAC and RIC plus melphalan mitigated some of the FLT3-ITD MRD–associated risk, the bigger picture of mostly poor outcomes in this population was unchanged. It should be noted that the studies addressing this issue have used different methods of MRD detection, timing of assessment and sample sources. Given differences in approaches to MRD assessment, it is important to specify all of these details to optimize interpretation of MRD status. Despite this caveat, based on data to date, it is possible that FLT3-ITD is a powerful enough chemotherapy resistance mechanism (at least in the absence of NPM1 comutation) that the primary benefit of HCT is a graft-versus-leukemia effect and that transplant conditioning intensity is less important.
The initial data from the MORPHO study were a significant advance in that they offered high-quality evidence for MRD-based post-HCT therapy.2 A first post hoc analysis provided further support for the role of serial quantitative FLT3-ITD MRD detection in guiding post-HCT gilteritinib maintenance therapy.3 The current second post hoc analysis provides important data that, particularly if confirmed in a randomized prospective fashion, make NPM1 comutation status a major variable to consider when weighing both SCT conditioning intensity and the utility of posttransplant gilteritinib.
Conflict-of-interest disclosure: The authors declare no competing financial interests.
