MRD Assessment Gains Momentum in Hematologic Malignancies

Jerald P. Radich, MD

Measurable residual disease (MRD) serves as an important biomarker to determine how a patient with a hematologic malignancy is responding to treatment, according to Jerald P. Radich, MD.

“MRD allows you to get a more sensitive determination of how your patient is responding to therapy. In most diseases, it can really help you benchmark response, predict what's going to happen in the future, and help you change therapies,” said Radich. “This biomarker can really [inform] how you treat patients [when their disease burden] is several orders of magnitude less than at frank hematological relapse; [when that happens, patients] are much harder to [manage].”

In an interview with OncLive, Radich, a member of the Clinical Research Division and the Kurt Enslein Endowed Chair at Fred Hutchinson Cancer Center, a professor at the University of Washington School of Medicine, and a physician in the Seattle Cancer Care Alliance, discussed the growing importance of MRD assessment in the treatment paradigm for patients with hematologic malignancies.

OncLive: What is the role of MRD in hematologic malignancies?

Radich: MRD measures disease burden. We’re using MRD to define response to therapy, not only to predict who is going to relapse but also as a way to understand how some therapies may be more active than other therapies. This has become an issue that has really garnered the attention of the FDA and other regulatory agencies: the idea that you can use MRD as an early marker of disease response in drug development trials. There is a lot of work being done now—in the acute leukemias in particular, but also in some of the chronic leukemias—to use MRD as end points in early phase II trials because it's so strongly associated with later outcomes.

What are the ways to measure MRD?

It depends on what disease you're dealing with. In general, one way to [measure MRD] is via flow cytometry; this technique measures aberrant antigen presentation on cells. The sensitivity [with this approach] is usually somewhere between 1 leukemia cell in 1,000 to 1 in maybe 10⁴. polymerase chain reaction (PCR)—based [detection], which will look at either translocations or point mutations. Next-generation sequencing (NGS) is a way to look in a more wholesale way at both translocations and point mutations in 1 assay; however, that approach that suffers a little bit from sensitivity and cost issues. Now, we're starting to use [methods] like single-cell genomics to understand the complexity of clonal evolution and how a Darwinian selection paves the way for residual disease in relapse.

How does MRD predict outcomes in patients?

Regarding the acute leukemias, we have yet to find a level of MRD positivity that is not associated with bad outcomes. In acute myeloid leukemia (AML) and acute lymphoblastic leukemia (ALL), at the end of induction or consolidation therapy, if a patient is MRD positive, no matter how you measure it, their risk of relapse is usually anywhere from 3- to 5-fold more than those who are MRD negative. In the chronic leukemias, like myeloma and chronic myeloid leukemia (CML), MRD doesn't seem to have as much of an impact as much [on outcomes]. Some patients can have low levels of disease and not relapse. What I mean by “low levels” is generally around 1 leukemia cell on a background of 104. Patients with levels that low or below that can still have presence of MRD but may not be at a high risk of relapse.

In what diseases is MRD most useful right now?

MRD is most useful in the acute leukemias; in those cancers, [MRD is] able to forecast long-term measures. That being said, if a patient with AML or ALL is MRD positive at the end of induction or consolidation therapy, they are also at very high risk of relapse after transplant. Even if they have small MRD positivity, even transplant cannot overcome that level of disease, [despite that] it’s very low. [MRD] is really just a biomarker of sensitivity. In patients whose disease is still present, MRD just shows you that their disease is relatively resistant compared with someone whose disease becomes negative.

In diseases like CML, which is the model for other chronic diseases like CLL and myeloma, we use MRD for measuring depth where we think patients are safe to eventually discontinue therapy; this has been shown to be safe in a certain cohort of patients with CML with about half of people relapsing and half not. That's going to be the model for what will eventually happen in CLL and myeloma, when they get a fair amount of people at those low levels of disease for a long period of time. When do we discontinue maintenance in these patients as well? That’s probably how it would be used in those diseases.

What are the challenges with using MRD?

There's a few. One is that many of these methods cannot be standardized across areas; this is especially true with flow cytometry, where it's done very differently with various levels of sensitivity from lab to lab. When you consider [approaches like] NGS, the problem has been turnaround time, which is relatively very slow compared with other PCR methods. Cost is another issue. Although [NGS costs] will eventually come down, [these tests] are currently relatively expensive. Availability [of these tests] is becoming less of an issue, but for many community doctors, finding ready access to reliable assays that can turn results around quickly while still having good quality control can be an issue.

Is there anything else that you would like to add?

It’s important to remember that as time goes on, patients are likely going to demand these assays. We saw that happen in the CML world with PCR-based testing. PCR methods were not frequently done by physicians in the past, but eventually advocacy groups and the like really started to demand that [the test] become part and parcel of patient care. That came despite the fact that National Comprehensive Cancer Network guidelines asked for this stuff. It wasn’t until patients got involved that it put pressure on everyone to come up with these standards. I think that will also be true of MRD in most of these other hematologic malignancies.