Unlike acute lymphoblastic leukemia (ALL), where the use of measurable residual disease (MRD)–guided therapies and the MRD “eraser” blinatumomab has led to dramatic success, similar MRD-driven therapeutic accomplishments in acute myeloid leukemia (AML) have been elusive. The incorporation of targeted FLT3 inhibitors into the therapeutic paradigm in FLT3-ITD mutated AML has transformed what was historically a dreaded diagnosis into a frequently curable AML subtype. Based on results from randomized controlled trials, FLT3 inhibitors are now routinely combined with intensive chemotherapy during induction and consolidation1,2 and restarted as maintenance therapy following allogeneic hematopoietic cell transplantation (HCT). However, tolerability in the HCT setting has challenged the universal uptake of the initially studied FLT3 inhibitors, sorafenib and midostaurin, as post-HCT maintenance in the clinic. Thus, despite early evidence,3,4 critical questions remain—most notably, do all patients with FLT3-ITD mutated AML derive benefit from post-HCT FLT3 inhibitor maintenance, and can we use recently developed deep MRD technologies to select patients who truly require this additional therapeutic intervention?
These are the questions that the BMT-CTN1506/MORPHO intended to answer through an international phase III randomized controlled trial examining gilteritinib versus placebo as post-HCT maintenance in adults with FLT3-ITD mutated AML.5 With a primary endpoint of relapse-free survival (RFS), patients were randomized to begin gilteritinib 120mg daily or placebo between 30 and 90 days following HCT with maintenance continued for two years. Importantly, MRD was serially measured beginning pre-HCT using an ultra-sensitive polymerase chain reaction and next-generation sequencing (PCR-NGS) hybrid assay with a sensitivity of as low as 1 x 10-6 to detect FLT3-ITD mutations, and a key pre-specified secondary endpoint focused on the effect of gilteritinib versus placebo in patients with and without peri-HCT MRD. Of 488 patients registered pre-HCT, 356 patients underwent post-HCT randomization; most patients were treated in North America but approximately one-quarter of the study cohort was treated in Europe and another quarter were treated in Asia/Pacific countries. Gilteritinib was relatively well tolerated, with myelosuppression and liver enzyme elevation representing the primary side effects; 31 patients in the gilteritinib arm (versus 10 in the placebo arm) were unable to complete two years of maintenance due to adverse events. While the two-year RFS was 77.2% (95% CI, 70.1-82.8) in patients receiving gilteritinib and 69.9% (95% CI, 62.4-76.2), the trial just barely failed to meet the primary RFS endpoint (hazard ratio [HR] 0.679, 95% CI, 0.459-1.005; p=0.0518).
However, the story does not end there. Using the highly sensitive MRD assay, approximately half of the patient cohort was found to have detectable MRD either pre-HCT or early post-HCT, which was associated with decreased RFS and overall survival, regardless of treatment arm. Perhaps the most interesting and clinically relevant finding was that patients with detectable MRD pre-HCT or early post-HCT experienced striking improvements in RFS with the use of gilteritinib versus placebo (p=0.0065), while RFS among MRD-negative patients was essentially identical for patients regardless of maintenance assignment (p=0.575). An additional noteworthy finding was that, relative to placebo, post-HCT gilteritinib maintenance led to significantly improved outcomes in patients treated in North America, whereas the benefit was not seen for patients who were treated at other international sites.
In Brief
MORPHO’s results are important and immediately practice-changing. Due to the inability of the study to meet its primary endpoint, statisticians and regulators may forever brand this a “negative study.” This perhaps misses the mark in terms of the clinical importance of its observations. The data confirm a biologic hypothesis that post-HCT FLT3 inhibitor therapy allows for effective control of disease in the highest-risk population (MRD-positive patients) until such time that a graft-versus-leukemic effect can clear residual disease, ultimately contributing to cure. Conversely, patients without detectable FLT3-ITD MRD peri-HCT showed no demonstrable benefit from gilteritinib maintenance; thus, they can potentially be spared from any added toxicity. It should also be noted that MORPHO’s lack of restrictions on age, prior therapy, transplant preparative regimen, or graft-versus-host disease prophylaxis, as well as its worldwide enrollment, introduced substantial heterogeneity in transplant delivery that complicates data interpretation. Indeed, the study demonstrated marked differences in the RFS of patients randomly assigned to the placebo arm based upon the location in which they enrolled. These differences translated into regional variations in the clinical benefit of maintenance gilteritinib relative to placebo.
For years, hematologists have sought an “MRD eraser” for AML. Numerous studies have confirmed that peri-HCT MRD in AML is prognostic, but MORPHO now shows that peri-HCT MRD is actionable. Increased access to FLT3-ITD MRD testing, as well as reimbursement for MRD testing and FLT3 inhibitors, remain issues to be solved in the field. Overall, data from MORPHO move the field forward and help us make rational treatment decisions for patients with AML by showing when we should augment therapy for patients in need and limit excess treatment for those who would derive no clear benefit.
Competing Interests
Dr. Perl has received consulting fees, research funding, and travel support from Astellas and Daiichi Sankyo, and he serves on their advisory boards. He also serves on an advisory board for Aptose and has received research funding from Fujifilm. Dr. Muffly has received consulting funding from Astellas.
