Minimal Residual Disease in Acute Myeloid Leukemia: Mutation Matters

A Conversation With David P. Steensma, MD

Get Permission

A RECENT article in The New England Journal of Medicine explored the nuances of minimal/measurable residual disease testing after induction treatment of acute myeloid leukemia (AML)1 and David P. Steensma, MD, and Benjamin L. Ebert, MD, PhD, of Dana-Farber Cancer Institute and Harvard Medical School in Boston, penned an editorial discussing minimal residual disease in AML in the context of clonal hematopoiesis, which is frequent in older patients.2 The ASCO Post recently asked Dr. Steensma to put the findings from this important study into perspective for our readers. 

David P. Steensma, MD

David P. Steensma, MD

Importance of Minimal Residual Disease in AML 

What is the importance of minimal residual disease in AML? 

When a patient with leukemia is in complete remission according to conventional clinicopathologic criteria, the burden of malignant cells that are still present after induction therapy is a strong predictor of relapse risk. Minimal residual disease is the term we most often use to describe the persistence of these cells below the level of detection with standard light microscopy, supplemented by immunohistochemical stains. Although assessment of minimal residual disease has gained widespread acceptance in other leukemias, it is also very important in AML—yet such testing is currently uncommonly used in clinical practice. 

Why not look for minimal residual disease in patients with AML? 

Many hematologist-oncologists are still not using minimal residual disease testing, even in diseases such as acute lymphoid leukemia (ALL), where it is well established as the most important predictor of relapse, and testing is recommended by guidelines. In fact, when my leukemia group colleagues and I have done teaching or participated in advisory boards, we’ve noticed a degree of confusion about how minimal residual disease assessment for ALL is done and what its implications are. 

“Minimal residual disease testing in AML is more difficult and more nuanced than it is, for example, in ALL.”
— David P. Steensma, MD

Tweet this quote

In AML, it’s not been common to perform minimal residual disease mutation testing for two reasons. First, even if we document minimal residual disease, although we know it is a risk factor for relapse, we can’t necessarily do anything about it. If you transplant patients with persistent minimal residual disease, they don’t necessarily do any better. 

Second, minimal residual disease testing in AML is more difficult and more nuanced than it is, for example, in ALL. In ALL, cells with an abnormal phenotype are relatively easy to detect by special high-sensitivity flow-cytometry assays. In AML, malignant cells may not have a distinct immunophenotype to clearly distinguish them from normal progenitors, so they might be missed on flow tests, and next-generation sequencing tests can be difficult to interpret. 

I hope the publication of this article will be informative and lead to more testing—and, more important, prompt more trials focused specifically on the high-risk minimal residual disease–positive group, for whom we need better options. 

Key Study Findings 

What were the main findings of the published study? 

On behalf of the HOVON and Swiss cooperative groups, Jongen-Lavrencic and colleagues presented the results of targeted sequencing-based assessments for minimal residual disease in 430 patients with AML treated with intensive induction therapy.1 They detected minimal residual disease in 51% of the patients, and its presence was, as has previously been shown by others, a strong relapse predictor. 

However, there was a subset of patients whose only detectable mutation was in one of three genes encoding epigenetic regulators: DNMT3A, TET2, or ASXL1, which the authors called “DTA mutations.” These patients had a lower likelihood of relapse, at least after 4 years of follow-up. 

DTA mutations are three of the most common mutations associated with what Ben Ebert and I and several of our colleagues defined in 2015 as “clonal hematopoiesis of indeterminate potential” (CHIP).3 CHIP is not a universal consequence of aging but is present in many elderly persons—10% by age 65 to 70 and more than 35% after age 85. It is a risk factor both for developing a hematologic malignancy and, interestingly, for cardiovascular mortality. It can be thought of as a myeloid or stem cell equivalent of monoclonal gammopathy of undetermined significance (MGUS), which is a precursor for myeloma in the same way that CHIP can be a precursor for AML or myelodysplastic syndromes. 

We have known that minimal residual disease is a risk factor for relapse in AML, but the distinction between minimal residual disease due to mutations that are not CHIP-associated such as FLT3 or NPM1, vs those that are, is new information for clinicians. So we now know that DNA mutations present after AML induction have different implications, depending on which mutation type is present. Understanding that nuance is important. 

“CAR T-cell therapy is well established in hematologic cancers. We just need to figure out how to do it in myeloid diseases.”
— David P. Steensma, MD

Tweet this quote

More on DTA Mutations 

Should oncologists be concerned, for other reasons, about the detection of DTA mutations alone, or can they simply be dismissed? 

I think they are still important. With longer follow-up, DTA mutations may turn out to be a risk factor for relapse. It’s just that, for these patients, it may take a while for additional new mutations to accumulate, if all they have after treatment are age-related genetic changes, which are the initiators of clonal expansion but not the driver of the acute leukemia. 

Minimal residual disease in AML is analogous to trimming a few branches off an invasive tree; the tree is still very much alive. DTA minimal residual disease, in contrast, is like chopping down the unwanted tree but leaving a stump; most stumps will never regrow into a tree, but, given time, maybe a few will. Minimal residual disease negativity is like chopping down the bad tree and digging up the stump and grinding it up. 

At this point, I do think patients with CHIP-like minimal residual disease can be told their risk for relapse may be lower than that for the other group of patients—not zero but certainly lower. Also, we don’t know the cardiovascular risk in this specific population, but I would guess it is elevated like CHIP. 

Ways to Measure Minimal Residual Disease 

For assessment, the study used both next-generation sequencing and flow cytometry, and the concordance between the two was just 69%. They each contributed independent prognostic information. Does this mean both tests should be ordered together? 

These are two ways to measure minimal residual disease in AML, and they provide slightly different results. The study authors suggest that maybe we need to do both, but I would probably not do that clinically now. Next-generation sequencing would be my first choice for assessment, if I had to choose one, because you get a bit more information. Flow cytometry adds to the expense without providing a lot of additional information. However, there are fans of flow cytometry who will disagree with me! 

New Treatment Approaches 

When non-DTA mutations are detected, can anything be done to reduce a patient’s risk for relapse? 

At the moment, these patients would typically be sent to transplant, but they don’t necessarily do better, even then. We definitely need new treatment approaches for this situation, in particular, ways to eradicate minimal residual disease. There are clinical trials looking at new drugs for this. Probably the most promising approaches are with immunotherapy. 

One agent being evaluated in AML minimal residual disease is the monoclonal antibody/drug-conjugate SL-401, which targets CD123 and is not mutation-specific. Unlike their impressive efficacy in solid tumors, immune checkpoint inhibitors have not been very successful so far in hematologic malignancies, at least as monotherapies. On the other hand, chimeric antigen receptor (CAR) T-cell therapy is well established in hematologic cancers. We just need to figure out how to do it in myeloid diseases. It’s already a game-changer in lymphoid and plasma cell disorders. Perhaps someday we’ll have a safer CAR T-cell product we can use in the minimal residual disease–positive population. That would be an important advance. ■

DISCLOSURE: Dr. Steensma reported no conflicts of interest. 


1. Jongen-Lavrencic M, Grob T, Hanekamp D, et al: Molecular minimal residual disease in acute myeloid leukemia. N Engl J Med 378:1189-1199, 2018

2. Steensma DP, Ebert BL: Clonal hematopoiesis after induction chemotherapy for acute myeloid leukemia. N Engl J Med 378:1244-1245, 2018

3. Steensma DP, Bejar R, Jaiswal S, et al: Clonal hematopoiesis of indeterminate potential and its distinction from myelodysplastic syndromes. Blood 126:9-16, 2015.