Next-Generation Sequencing MRD Assessment Valuable Tool in AML Management

Assessing minimal residual disease (MRD) using next-generation sequencing (NGS) in patients with acute myeloid leukemia (AML) prior to undergoing allogeneic hematopoietic cell transplantation (alloHCT) was highly predictive of post-transplant relapse and survival. This technique could help identify patients who should undergo alloHCT, according to Felicitas Thol, MD, from the department of hematology, hemostasis, oncology, and stem cell transplantation at Hannover Medical School in Germany, and researchers.

“Patients with intermediate-risk AML [may be] transplanted if they are in first complete remission,” study coauthor Michael Heuser, MD, also from the Hannover Medical School, told ASH Clinical News. “Our diagnostic approach identifies patients with AML who have an excellent prognosis with alloHCT and a low risk of relapse, which may help better select those patients who would benefit most from transplant.”

The study included 116 adult patients with AML who were in complete remission (CR) and underwent an alloHCT between 1996 and 2016. Patients also had an available DNA sample at the time of diagnosis and during CR prior to undergoing alloHCT.

To validate the prognostic effect of NGSbased MRD assessment, the researchers first evaluated MRD at diagnosis, in remission, and at molecular relapse in four patients with NPM1-mutated AML using an established polymerase chain reaction testing method and the NGS method to establish concordance between the two methods.

Investigators collected and quantified MRD using a 46-gene sequencing panel in samples from bone marrow and peripheral blood. Of the entire cohort of 116 patients, 20 were excluded because they did not have any appropriate mutation in the myeloid panel analysis or the variant allele frequency was >5 percent in the selected MRD mutation before alloHCT.

Ninety-six patients were evaluable for the main analysis: 43 patients (45%) were MRD positive and 53 (55%) were MRD negative prior to transplant. Patients who were MRD positive more often had an adverse cytogenetic profile and complex karyotype, the authors noted, and, “importantly, the frequency of MRD positivity was similar whether MRD was determined in peripheral blood or bone marrow (43% and 48%, respectively).”

Over a median follow-up of 6.2 years (range not provided), 27 MRD-positive patients (63%) and eight MRD-negative patients (15%) experienced a post-alloHCT relapse. This translated to a five-year cumulative incidence of relapse (CIR) of 66 percent and 17 percent, respectively. In univariate competing-risk analysis, MRDpositive patients had a significantly higher risk of relapse (hazard ratio [HR] = 5.58; 95% CI 2.47-12.59; p<0.001).

Survival also was shorter among MRDpositive patients than MRD-negative patients:

  • 5-year overall survival (OS): 41% vs. 78% (HR=3.06; 95% CI 1.53-6.12; p=0.002)
  • 5-year relapse-free survival (RFS): 31% vs. 74% (HR=3.56; 95% CI 1.86- 6.81; p<0.001)

However, there was no significant difference in nonrelapse mortality between the two MRD groups (HR=0.60; p=0.47).

MRD-positivity also was found to be an independent predictor for CIR, OS, and RFS in multivariate analyses adjusting for mutation status, conditioning regimen, and age (TABLE).

“[This] NGS-based MRD monitoring can be applied to a large proportion of AML patients using almost any available molecular aberration and … is highly predictive of relapse and survival when assessed in CR prior to alloHCT,” the authors conclude. Also, it overcomes limitations of available MRD techniques for AML, which rely on monitoring NPM1 markers that are present only in a minority of patients with AML.

“NGS-MRD methods need to be standardized, and this is now being addressed by the European LeukemiaNet MRD Working Party,” Dr. Heuser added. “Next, we need to establish the clinical benefit of MRD-guided treatment decisions in prospective trials.”

The study was limited by its relatively small sample size and the inclusion of patients receiving care from only one institution. In addition, the investigators suggested that the NGS-based MRD technique is limited by the clonal evolution of cells and concerns about the assay’s sensitivity.

The authors report no conflicts of interest.


Thol F, Gabdoulline R, Liebich A, et al. Multi-gene measurable residual disease (MRD) monitoring by next-generation sequencing in AML is highly predictive for outcome after allogeneic hematopoietic cell transplantation. 2018 September 6. [Epub ahead of print]

“When considering the clinical application of MRD in any disease, there are multiple levels to the story: The first is whether MRD can predict post-treatment outcomes; the second is determining what method is best; and the third – and most important and most difficult hurdle to clear – centers on intervening to modify risk and change the course of disease in patients at the greatest risk of relapse.

This particular study looked at the role MRD plays in AML. There are many methods to detect MRD in AML, including metaphase cytogenetics, fluorescence in situ hybridization, multicolor flow cytometry, and molecular analyses (e.g., next-generation sequencing) for known mutations. Each of these methods has advantages and drawbacks, but most hinge on identifying aberrations that set the leukemia cells apart from normal myeloblasts. So far, NPM1 mutations have been identified as an ideal candidate marker for MRD, but these are present in approximately 35 percent of cases of newly diagnosed AML.

This study used NGS to detect MRD in patients undergoing alloHCT. The key here is that the samples were paired, meaning that each patient had a sample from the time of diagnosis and at the time of MRD assessment.

The innovative portion of this study is the group’s use of bioinformatics. To improve the performance of mutation detection by NGS, the authors used an error-corrected sequencing approach to identify small amounts of MRD and differentiate that from sequencing errors. Using this method, they were able to conduct reliable MRD assessment on a wide swath of AML patients.

While this is certainly a step forward, there are several unanswered questions. Mutation detection for MRD assessment is hampered by the presence of clonal hematopoiesis and ancestral clones. These not only predispose a patient to the development of AML but also can be present at the time of MRD assessment and influence the interpretation of this testing in a multifaceted manner. Ultimately, as our detection of MRD in AML is refined, future studies must turn to interventions that will reduce the risk of relapse and improve the outcomes for our patients.”

Aaron T. Gerds, MD, MS
Assistant Professor in Medicine
Cleveland Clinic Taussig Cancer Institute
Cleveland, OH