Is Minimal Residual Disease a Better Prognostic Factor than Genetic Markers in AML?

In patients with high-risk acute myeloid leukemia (AML), often defined by cytogenetic and/or molecular factors, allogeneic hematopoietic cell transplantation (AlloHCT) is recommended in first remission. For patients with cytogenetically standard-risk AML– who are often younger adults – the benefits of AlloHCT are less clear.

The NPM1 mutation appears in approximately one-third of this patient population, representing a potential leukemia-specific prognostic marker that could, when combined with monitoring of minimal residual disease (MRD) by reverse-transcriptase quantitative polymerase chain reaction assays, help determine whether patients would benefit from AlloHCT, according to the authors of a recent report published in The New England Journal of Medicine.1

“We wanted to see which method was more powerful – the up-front profiling technique or the MRD testing – and, in terms of predicting patient outcome, the MRD testing was much more powerful,” Prof. David Grimwade, a co-author of the study told ASH Clinical News.

Adam Ivey, MSc, from the Molecular Oncology Unit and Cancer Genetics Laboratory in the Department of Medical and Molecular Genetics at Guy’s Hospital in London, and colleagues evaluated outcomes according to molecular profiles and sequential monitoring of MRD in a large series of patients with NPM1-mutated AML who had undergone intensive treatment in the National Cancer Research Institute AML17 trial. Patients were enrolled between April 6, 2009, and December 31, 2014.

“People decide whether or not to do a transplant based largely on the characteristics of the AML at diagnosis – relying on genetic markers and cytogenetics,” Prof. Grimwade said. For the large portion of AML patients who are considered “standard risk” according to cytogenetics, he continued, meaning “they do not have low- or high-risk cytogenetics. … They account for a good half of younger adults with AML, and people just don’t know whether or not they are best served by having a transplant. It’s a contentious area [and] it’s uncertain what to do.”

Researchers used a real-time quantitative polymerase-chain-reaction assay to detect MRD in 2,569 samples from 346 patients with NPM1-mutated AML. They then developed a custom 51-gene panel to perform targeted sequencing of 223 samples at the time of diagnosis and 49 samples obtained at the time of relapse. “Molecular profiling highlighted the complexity of NPM1-mutated AML, with segregation of patients into more than 150 subgroups, thus precluding reliable outcome prediction,” the researchers wrote. MRD status “was more informative,” they added, and was able to detect trace levels of leukemia cells after chemotherapy.

Fifteen percent of patients had persistent NPM1-mutated transcripts following the second chemotherapy cycle, according to analyisis of the blood samples. These patients experienced a significantly higher risk of relapse at three years, compared with patients without such transcripts (82% vs. 30%; hazard ratio [HR] = 4.80; 95% CI 2.95-7.80; p<0.001). Persistent NPM1-mutated transcripts were also associated with a lower rate of overall survival at three years (24% vs. 75%; HR=4.38; 95% CI 2.57-7.47; p<0.001).

In multivariate analysis, which included results of the targeted sequencing analyses, MRD was the only independent prognostic factor for mortality (HR=4.84; 95% CI 2.57-9.15; p<0.001) and relapse (HR=5.09; 95% CI 2.84-9.13; p<0.001).

“Compared with crude genetic information [gathered via targeted sequencing at baseline], testing whether a patient has a response via MRD is so much more powerful,” Prof. Grimwade said.

In addition, though mutations associated with pre-leukemic clones continued to be detectable following chemotherapy, even when patients were in remission, detection of NPM1 mutations were more associated with frank relapse, and were found in 69 of 70 patients at the time of relapse. Assessing the leukemia-specific NPM1 mutation in AML patients can be used for sequential monitoring of MRD to identify impending relapse and to help guide treatment decisions, the authors added. “The message is not that profiling should not be performed – because it could identify particular mutations that could be useful for targeted therapies – but that we cannot put all our eggs in that basket,” Prof. Grimwade told ASH Clinical News. “These data have shown that MRD testing provides independent prognostic information and needs to be on the agenda.”

While 21 MRD-positive patients (46%) and 61 MRD-negative patients (26%) underwent AlloHCT, the authors observed “no significant effect of transplantation among patients according to MRD status.” This analysis was limited by small numbers, though, so “the question of whether outcomes might be improved by more rapid deployment of transplantation is being studied in the ongoing NCRI AML19 trial.” Also, rates of MRD detection were higher in bone marrow–derived samples than peripheral blood samples, suggesting that risk assessment could differ depending on the source of the samples.

Though MRD has been proven as an independent prognostic factor for the duration of remission and survival in acute lymphocytic leukemia (ALL), Michael J. Burke, MD, from the Department of Pediatrics at the Children’s Hospital of Wisconsin, wrote in an accompanying editorial, “since AML has a greater molecular heterogeneity than ALL, the routine assessment of MRD has not been as quickly adopted in patients with AML.”2

While the study has limitations (including the fact that its results do not apply to patients without NPM1-mutated AML and that the impact of modifying interventions based on MRD status need to be studied formally), the “provocative results” raise some questions. “Will the addition of the assessment of MRD to risk stratification achieve better outcomes in adults with AML? Can we reassign treatment and thereby reduce the risk of relapse among patients who are now classified as being either at high risk or low risk? … Time will tell, but this moment may prove to be a pivotal one in the assessment of MRD to assign treatment in patients with AML,” Dr. Burke concluded.


  1. Ivey A, Hills RK, Simpson MA, et al. Assessment of Minimal residual disease in standard-risk AML. N Engl J Med. 2016;374:422-33.
  2. Burke MJ. Minimal residual disease in NPM1-mutated AML. N Engl J Med. 2016;374:481-82.