Harry Erba, MD, PhD: I’m going to come back to Rami, and we’re going to get back on track and remember that we do cure people with intensive chemotherapy. There’s been a lot of talk about using minimal residual disease assays, MRD assays, for guiding therapy. Walk us through how you think about consolidation. How many cycles of consolidation and transplant? And where does MRD come into decision making about those choices?

Rami Komrokji, MD: We can start with the group that we know most about: the patients who have core binding abnormalities. Those patients benefit from intensive chemotherapy, as you alluded. Nowadays, we add gemtuzumab as well for those patients.

In those patients, the role of consolidation is very well established. Typically, they should get 3 to 4 rounds of consolidation with a curative intent. For the other groups, once they get into a remission, the role of consolidation can sometimes be controversial. How many cycles are ideal? We’ve shied away from doing the high-dose cytarabine in the noncore binding cases. The exact number of rounds is not known: somewhere from 2 to 4 rounds of treatment.

It gets even more controversial in older patients. What’s the benefit of consolidation in older patients? How many cycles? Most people agree that some consolidation may be better than none: 1 to 2 cycles. That’s also where the role of assessing the disease comes: looking at the MRD status for the disease. When we defer to MRD status, it’s a very generic description we are talking about. First, to make a point, that when we are talking about the NGS [next-generation sequencing] panels that we use, those are not MRD measurements, because they have a certain threshold of 5% to 10%. When you are saying the next-generation sequencing test is back negative, that means you are below the level of detection, which is not a true MRD status.

The MRD status can be checked by molecular mutations and by flow cytometry as well. That’s an evolving field. Even I am sometimes confused about what to do with that data. If they’re MRD negative, then they are going to do better. If they go to transplant, are they MRD negative or MRD positive?

Technically, if they’re MRD negative, they are going to do better with a transplant. But can you avoid it? MRD-positive patients are going to do worse with the transplant, but they are probably going to have worse outcomes without going to the transplant. It gets confusing, and you think about this all the time for patients. What do we do? There is no doubt that transplant remains a curative strategy for a lot of patients, and there is no doubt that the MRD role is evolving. The ALL [acute lymphoblastic leukemia] field is ahead of the AML field, but what I can say is this: MRD negativity is a good thing to achieve.

Mark Levis, MD, PhD: In ALL, they have a reliable MRD assay and a drug.

Harry Erba, MD, PhD: Naval, when you get back their NGS panel, does it matter what is still positive while you’re in remission? Does that make a difference?

Naval Daver, MD: I would say, as a general, sweeping statement, not much. It doesn’t change my approach. A couple of things that Rami said are very important. For practical decision making, if I decided at baseline that this patient is intermediate or high risk and transplant was my goal, and I get the patient in morphological remission in AML—which is very different from ALL—and transplant is lined up, I would proceed to transplant whether he’s MRD positive or negative.

My transplant colleagues may push back—sometimes they do, and sometimes they don’t—but we all agree that MRD-positive transplant doesn’t do as well as MRD-negative transplant. It seems as though it does better than no transplant; it’s a “don’t let perfect be the enemy of good” situation. That’s how we’re using MRD for transplant-related decisions.

Naval Daver, MD: We learned from the Hobon analysis that certain mutations present at the time of remission—ASXL1, TET2, and DNAM3A—don’t seem to impact long-term prognosis. Does that mean that patient doesn’t need a transplant, Dan?

Dan Pollyea, MD, MS: This is tough; we struggle with this all the time because those DTA mutations that persist, according to the New England Journal of Medicine paper, don’t impact long-term outcomes like survival and don’t appear to be associated with a CHIP-like status, or a reset to a CHIP-like status. Based on the data from that group in that paper, it’s not unreasonable to not consider those mutations as MRD markers. But the question we struggle with is this: on an individualized basis, what else could be CHIP that’s not a DTA mutation? DTA fell out of that analysis because they’re so common, but in an individual patient, IDH can be a CHIP mutation as well as multiple others. How do we know what’s important and what to ignore? At the moment, we don’t.

Rami Komrokji, MD: The other area of the MRD status does come in a good risk group. If you did treatment for those patients and they have persistent disease, are you going to take them to transplant now or not? There’s probably more of a role there. If somebody has NPM1 step-positive status, core binding, those patients may benefit.

Mark Levis, MD, PhD: To your point, what do we do? The transplanters don’t want to take the patient who’s MRD positive. Their own data are showing clearly that trying to make that MRD go away by hammering the patients doesn’t help.

Harry Erba, MD, PhD: That’s a really important point. I was at the 1 of the oral sessions for allotransplants. Some of my colleagues wondered what I was doing there.

Mark Levis, MD, PhD: I was at the same session.

Harry Erba, MD, PhD: It was very important information for us because the data from Fred Hutchinson Cancer Research Center that was presented by Jacob Appelbaum addressed that. They didn’t specify to a physician whether the patient needed to be MRD positive or MRD negative prior to a transplant. They left it to the doctors, the patients, and the transplanters to decide.

What they showed was patients who were MRD positive who received intensive chemotherapy, or any chemotherapy, prior to going to transplant had a worse outcome. Many of them didn’t become MRD negative because they were already defined to be chemotherapy resistant. This was not a randomized trial, and they’re thinking about doing that, but it does suggest that maybe beating on these patients with more chemotherapy isn’t the way to go. What we need to find out may be immunotherapies or different approaches that might improve their outcomes.

Mark Levis, MD, PhD: You erase it with a targeted agent.

Dan Pollyea, MD, MS: We presented similar data at this meeting albeit in the poster session. No one named Appelbaum was on our paper, but it is very interesting that, if you’re MRD negative, you do better. If you’re MRD positive, you do worse. It doesn’t matter if you’re MRD positive going into a transplant if we try to consolidate that MRD away: you still do as poorly as an MRD-positive patient. This is not an ALL situation. We would love to be able to mop up MRD, but can we, will it matter, and how will it matter?

Harry Erba, MD, PhD: As moderator, I’m going to say something to pull this together and then go on quickly so no one could disagree. It’s really important for the people listening to this because there’s a lot of hype about MRD. Why are we spending a lot of money on the individual patient outside clinical trials—not in clinical trials but outside clinical trials—getting this information? It may subvert the best option for therapy for those patients, and you’re spending money. This is different from a guided PCR [polymerase chain reaction] analysis for FLT3 and for IDH where have therapeutic targets. This NGS panel or flow cytometry, I see it as just adding expense to the care of patients with AML without a provenbenefit yet, and I would not recommend doing that outside the context of a clinical trial where the assay is being developed appropriately.

Transcript Edited for Clarity