Acute Myeloid Leukemia: The Path to Personalized Medicine - Episode 8

MRD Status in AML


Harry Erba, MD, PhD: So you brought up MRD [minimal residual disease]. The depth of remission might be very important, and a lot of us are looking at MRD, even though I don’t think we really know how to measure it and what to do with it. I know some of our transplant groups know what to do with it. If we have a patient who’s in remission but they have evidence of MRD, based on the prognostic significance, I’ve had patients sent back to me to get more consolidation. So I guess it begs the question, help me here, Mark, when they’re sent back to me, what consolidation to get rid of MRD?

Alexander Perl, MD: And is that right?

Harry Erba, MD, PhD: And is that right?

Mark J. Levis, MD, PhD: And does that help them?

Harry Erba, MD, PhD: And how would you test?

Mark J. Levis, MD, PhD: I’m ambivalent in MRD. I’m actively involved in research in MRD, but the problem is sitting in the room with a patient and you’d say, “Okay, great news, you have no MRD.” “Okay, so that means I’m cured?” “No, you have a relapse risk of about 30%.” Or, “Terrible news, you have MRD.” “Oh, I’m gonna die.” “Well, no, you have a 30% chance of not relapsing.” It’s like the Fibonacci sequence. It’s a magic 70/30. Everything is 70/30. So we’re not quite at the point where we know what to do with this.

We all recognize that if you’ve got detectable MRD by some method, it’s a bad thing. We haven’t figured out how to address it. And again, you run into the problems of, “Okay, how did you assess the MRD and how sensitive was this?” “Well, I went down to 1 cell in a million.” “Okay, well that sounds like a pretty good remission if you’re detecting 1 cell in a million.” On the other hand, if it’s 1% by my fancy next-generation sequencing assay, well you know I can actually do that by flow in my laboratory here. What’s so great about that? And I know that’s bad. So the modality of testing, flow cytometry, hot spot PCR [polymerase chain reaction], focused PCR, RT [real-time] PCR, next-generation sequencing, modified assays, this is an evolving world. We don’t yet know how to do it, but I think the important thing is virtually for every clinical trial going forward, MRD is a key component because we need it as a regulatory point.

Harry Erba, MD, PhD: We do. We can’t have 10-year RATIFY trials for drug approvals. We need some surrogate endpoint earlier on that translates into benefit. And the other thing is you’ll get back a next-generation sequencing panel after a remission, and there might still be mutations there. Is it always bad? Sasha?

Alexander Perl, MD: It depends on what the mutations are, because some mutations people walk around with and it’s not anything more than just their bone marrow is getting older. They have clinical hematopoiesis of indeterminate significance. And many, if not the vast majority, of people just walking down the street who could be demonstrably shown to have a mutation in their blood and clonal hematopoiesis never go on to a myeloid malignancy. But some of them do. And so if they get treated for that myeloid malignancy, they may go back to that state they were in before. And if you follow those patients, their risk of relapse is substantially lower than if you see a mutation that’s more strongly associated with AML [acute myeloid leukemia], or at least progression beyond CHIP [clonal hematopoiesis of indeterminate potential].

So these aging-related mutations, and the most common ones we see are DNMT3A, ASXL1, and TET2. So common that they were lumped together for analysis on a large study that was not really looking at MRD but was looking at measurable leukemia-associated mutations. And patients with those, again DNMT3A, TET2, ASXL1, had a lower rate of relapse than if you had any other still detectable mutation once you were in remission after induction. So, if you’re seeing mutations associated with leukemia, that’s bad. That’s just like those abnormal cells that Mark’s describing by flow. But if you’re seeing certain mutations associated with CHIP, it may not be that bad. It doesn’t mean they don’t relapse, it just means it’s lower, which gets back to the question of can you combine modalities? Can you do flow plus a next-generation? And, on that study, that looked like that was a bit better at figuring out maybe not the 70/30 but a little bit wider. But it’s still not as good as we would like it to be.

Harry Erba, MD, PhD: But the real question though, I mean those are all real questions, but then there’s the question for the clinician. So it comes back from the laboratory that the patient is MRD-positive. Do we know that some consolidation or taking straight to transplant actually makes a difference? There was a really interesting presentation by the Toronto group that looked at the use of MRD in the core-binding factor leukemias where it has been recommended by ELN [European LeukemiaNet] that at the end of induction, consolidation, every 3 months you do a bone marrow biopsy looking for the RNA from the 8;21 or inversion 16. And what they showed was that in the majority of those patients where they can detect the minimal residual disease relapse, where they want to take them to transplant, the patients would have morphologic relapse before they can get to transplant. And we don’t even know if we did transplant in the MRD-positive state, if it would be a better outcome.

Alexander Perl, MD: Or if they respond to salvage, they relapse, you treat them with salvage, and they go to transplant then.

Mark J. Levis, MD, PhD: It sort of upset the ELN representatives, as I recall.

Naval G. Daver, MD: The other side of the MRD, so I think we all agree, FLAG-IDA [idarubicin/fludarabine/cytarabine/G-CSF], 3+7 induction, MRD gives us some and probably good predictive value. Okay, it’s not 100%. It’s 70%, 80%. But then you have this whole new wave of drugs we’re talking about, AZA [azacitidine]/venetoclax, FLT3 inhibitors, IDH [isocitrate dehydrogenase] inhibitors, immunotherapies, where we really don’t even clearly know if there is MRD eradication. There may be some benefit, and a small subset achieve it, like 10% in IDH, and surely do better than the others. But the others actually do quite well even if they don’t become MRD-negative. And I think that’s going to be important in the community. Don’t stop IDH and FLT3 inhibitors if you see the morphological remission but the IDH or the flow is still positive.

Harry Erba, MD, PhD: We’re going to get to this later when we talk about BiTEs [bi-specific T-cell engagers] and DARTs [dual-affinity re-targeting antibodies]. That might be the best place for those as an MRD eraser in AML, the proof of principle from ALL [acute lymphoblastic leukemia].

Mark J. Levis, MD, PhD: Although, in fact, on the flip side of that, your patient is going to be on your wonderful new targeted agent inhibitor for how long? Can you use MRD to say, “No, I actually do not any longer detect your IDH2 mutation, or I no longer detect that FLT3 mutation.” Can you come off maintenance therapy? Because maintenance therapy is now being raised as a big issue going forward with these agents.

Alexander Perl, MD: And when should you measure it? Do you need to look before transplant or after transplant?

Transcript Edited for Clarity