Assessment of measurable residual disease (MRD) is an independent prognostic indicator in acute myeloid leukemia (AML) [1]. Multiple techniques with various sensitivities are currently utilized: polymerase chain reaction (PCR), multiparameter flow cytometry (MFC), and next-generation sequencing (NGS) [2]. AML MRD positivity prior to allogeneic hematopoietic cell transplantation (pre-allo-HCT) is associated with worse leukemia-free-survival (LFS), overall survival (OS) and cumulative incidence of relapse (CIR) [3]. In comparison to previously published studies, Maffini and colleagues now report on the impact of MRD testing prior to a first allo-HCT in secondary AML (sAML), in a multicenter, retrospective registry-based analysis of the European Society for Blood and Marrow Transplantation (EBMT) [4]. A total of 318 adult patients ≥18 years with sAML in first complete remission (CR) transplanted between 2006 and 2018 were included, among whom 110 patients had detectable MRD compared to 208 patients with undetectable MRD. The majority (69%) of sAML evolved from an antecedent myelodysplastic syndrome and 59% had intermediate risk cytogenetics. Sixty-nine percent underwent matched unrelated allo-HCT, 31% underwent matched related allo-HCT, and 94% received unmanipulated peripheral blood stem cells. As expected in this cohort of relatively older patients with AML (median age: 60 years, range, 18–75), only 31% and 37% of patients underwent myeloablative conditioning regimen (MAC) in the MRD positive and MRD negative groups, respectively. There were no substantial differences in LFS, OS and graft-versus-host disease-free relapse-free survival (GRFS) between patients with and without MRD. These findings are consistent with a recent large study of 3,653 patients with AML: those with sAML (n = 601) had no differences in relapse-incidence between the MRD positive and MRD negative patients [5].

As the authors note, information on MRD testing methods was only available for 94 patients, and PCR or PCR combined with NGS constituted >90% of the used MRD testing methods. Unfortunately, no additional data for threshold levels, timing and performance characteristics for these MRD testing methods was available. Furthermore, retrospective registry-based studies looking at MRD for hematological malignancies are inherently confounded by the differences in MRD assessment techniques, their sensitivities, the timing of MRD assessment and whether one or multiple techniques were utilized. This variation underscores the need for standardization and harmonization for MRD assays among different centers, particularly when these tests are performed outside of clinical trials.

Nevertheless, Maffini and colleagues shed a light on the significance of MRD testing for sAML in a real-world setting. As sAML constitutes roughly 22% (range 5.7–42.1%) of patients with AML [6], it is important to identify additional prognostic factors that would predict response rates and survival outcomes in this patient population. The optimal treatment for patients with sAML is not well established and it’s often extrapolated from clinical trials evaluating therapy options for other AML subtypes. sAML is characterized by high risk factors including older age, poor performance status, co-morbidities, high risk cytogenetics and genetic mutations. As such, prospective data for sAML is limited and registry-based studies constitute the best available data to further understand and characterize sAML.

What is yet to be determined is how we can enhance the treatment of sAML. For instance, future studies should focus on the difference in MRD kinetics, whether intensive chemotherapy or combination of a hypomethylating agent with venetoclax have the same effect on MRD eradication, particularly prior to allo-HCT. Since allo-HCT is the only curative therapy for sAML [7], it is important to optimize the transplant outcomes with the appropriate selection of the conditioning regimen, GVHD prophylaxis therapy, and graft sources. For example, Hourigan and colleagues have shown that MAC, rather than reduced intensity (RIC), can improve the relapse and survival rates in patients with AML and genomic evidence of MRD prior to allo-HCT [8]. However, Gilleece and colleagues showed that in patients with AML in first CR and a detectable pre-allo-HCT MRD, RIC and non-myeloablative conditioning regimen were only inferior to MAC for patients <50 years old [9]. In addition, Murdock and colleagues showed that genomic evidence of MRD positivity, at the time of CR in patients with AML aged ≥60 undergoing allo-HCT, had no independent impact on post-transplant risks [10]. Whether these observations hold true for sAML is yet to be determined, although the use of MAC in this older patient population is challenging.

Patients with detectable pre-allo-HCT who are ineligible for MAC, may be more likely to benefit from post-allo-HCT maintenance therapy. Nevertheless, exploratory data from the SORMAIN trial showed that patients with undetectable pre-allo-HCT and those with detectable post-allo-HCT derived the strongest benefit from sorafenib maintenance therapy [11]. Two upcoming clinical trial (AMADEUS NCT04173533 and BMT-CTN 1506 NCT02997202) will hopefully address the value of maintenance therapy in patients with AML undergoing allo-HCT.

In conclusion, since Maffini and colleagues showed no effect of pre-allo-HCT MRD in patients with sAML, the transplant outcomes in this patient population might be primarily driven by baseline genetics rather than the MRD at the time of transplantation. Nevertheless, incorporating standardized MRD testing in the treatment of sAML, preferably in the context of prospective clinical trials, would be optimal. However, until then, harmonizing MRD testing in the context of registry-based retrospective studies could be a viable alternative in the hope of improving the treatment of and optimizing allo-HCT for sAML.