The presence of minimal residual disease predicts relapse in patients with NMP1-mutated acute myeloid leukemia and is superior to currently used molecular genetic markers in determining whether these patients should be considered for stem cell transplantation, a new study has found.

At 3 years, patients with minimal residual disease (MRD) had a significantly greater risk of relapse than those with no MRD (82% vs. 30%; univariate hazard ratio, 4.80; P less than .001) and a lower rate of survival (24% vs. 75%; univariate HR, 4.38; P less than .001), Adam Ivey  of King’s College London reported ( N Engl J Med. 2016; doi:10.1056/NEJMoa1507471 ).

In an editorial that accompanied the study Dr. Michael J. Burke from the Children’s Hospital of Wisconsin in Milwaukee wrote, “Time will tell, but this moment may prove to be a pivotal one in the assessment of minimal residual disease to assign treatment in patients with AML” ( N Engl J Med. 2016; doi:10.1056/NEJMe1515525 ).

In adult AML, assessment of MRD has taken a back seat to analyses of cytogenetic and molecular lesions in determining a patient’s risk and treatment strategy. Typically, allogeneic stem cell transplantation is used for patients with high-risk features such as chromosome 3, 5, or 7 abnormalities or the FLT3-internal tandem duplication (ITD) mutation, while chemotherapy alone is used for low-risk disease.

The role of transplantation is unclear, however, for cytogenetically standard-risk patients, which includes those with a mutation in the gene encoding nucleophosmin (NPM1).

To address this issue, the investigators used a reverse-transcriptase quantitative polymerase chain reaction assay to evaluate 2,569 bone marrow and peripheral-blood samples from 346 patients with NPM1 mutations who had completed two cycles of induction chemotherapy in the U.K. National Cancer Research Institute AML17 trial .

MRD, defined as persistence of NPM1-mutated transcripts in peripheral blood, was present in 15% of patients after the second chemotherapy cycle.

Patients with MRD were significantly more likely than those without MRD to have a high U.K. Medical Research Council clinical risk score and to carry the FLT3-ITD mutation.

On univariate analysis, the risk of relapse was significantly higher with the presence of MRD in peripheral blood, an increased white cell count, and with the DNMT3A and FLT3-ITD mutations.

Only the presence of MRD and an elevated white cell count significantly predicted survival, Mr. Ivey reported.

“We could find no specific molecular subgroup consisting of 10 patients or more that had a rate of survival less than 52%; in contrast, the rate in the group with the presence of minimal residual disease was 24%,” he observed.

In multivariate analysis, the presence of MRD was the only significant prognostic factor for relapse (HR, 5.09; P less than .001) or death (HR, 4.84; P less than .001).

The results were validated in an independent cohort of 91 AML17 study patients. It confirmed that MRD in peripheral blood predicts worse outcome at 2 years than the absence of MRD, with a cumulative incidence of relapse of 70% vs. 31% (P = .001) and overall survival rates of 40% vs. 87% (P = .001), reported the investigators, including senior author Professor David Grimwade , also from King’s College London.

The clinical implications of these results “are substantive” because NPM-1 mutated AML is the most common subtype of AML and because of the uncertainty over the best treatment strategy for patients typically classified as standard risk, editorialist Dr. Burke observed.

“Now with the ability to reclassify standard-risk or low-risk patients as high-risk on the basis of the persistent expression of mutant NPM1 transcripts, it may be possible that stem-cell transplantation is a better approach in patients who otherwise would be treated with chemotherapy alone and that transplantation may be avoidable in high-risk patients who have no evidence of minimal residual disease,” he wrote. “Such predictions will need to be tested prospectively.”

The presence of MRD is also known to be an important independent prognostic factor in acute lymphoblastic leukemia, but since AML has a greater molecular heterogeneity, routine MRD assessment has not been as quickly adopted in AML, Dr. Burke noted.

The Children’s Oncology Group, however, recently adopted MRD assessment by flow cytometry to further stratify children with newly diagnosed AML after first induction therapy into low-risk or high-risk groups.

The study was supported by grants from Bloodwise and the National Institute for Health Research. Mr. Ivey and Dr. Burke reported having no disclosures.