iwAL 2019 | Biological mechanisms of AML: a 2019 update

Mark Levis:
I’m Mark Levis at the International Workshop on Acute Leukemias in Barcelona, 2019. I’m here with Klaus Metzeler and Ravi Majeti. We’ve just finished the morning session on, call it biology and genetics of acute leukemia focusing on AML. Ravi, I’ll start with you. We grew up with this concept of the leukemia stem cell and we’d been chasing it for years, but now maybe it’s beginning to be turned on its head, or the things that we learned. So talk more about how that is changing now.

Ravi Majeti:
Well, I definitely think that the concepts of stem cells in acute AML in particular are very much different than the original observations. There’s more complexity, there’s complexity around the immunophenotype, and how definitive that is in terms of identifying and isolating leukemia stem cells, and there’s complexity around the disease itself, and the genomics and the functional diversity of cells and subsets within AML.

Ravi Majeti:
I think the concept of leukemia stem cells is still an important one, an important lens with which to view AML. Not all of the leukemic blasts are the same, and why that heterogeneity exists, how it relates to the biological features of the disease and most importantly, how it relates to therapy are still really important questions.

Mark Levis:
So, it sounds like the models are appropriately advancing. I think we’re going to have to get smarter with our models of studying stem cells. You described an interesting, perhaps more advanced model of this ossicle, but really trying to imitate the human microenvironment as best we can within this funny little immunodeficient mouse. But I think the data you showed I think was really eyeopening that as we sort of suspected over these last few years, boy, anybody can grow up to be a leukemia stem cell these days practically.

Ravi Majeti:
Well, the relationships of the different subsets of cells in AML is definitely more complex than what we know to occur in normal blood development. There can be a lot more developmental diversity among the cells, a lot more interchangeability in their functions. Like I always say, the arrows that are one way in normal hematopoiesis when we draw our schematics are definitely potentially more dotted lines or potentially two-way arrows when we get to leukemia. That’s not surprising when you think about how deranged they are from normal blood development.

Mark Levis:
Klaus, this intersects heavily with this whole concept of clonal hematopoiesis and really the two areas or are obviously intertwined. You presented data, pretty much implying again, what we’ve been learning over the last few years. As we age, our bone marrows, everybody’s bone marrow is slowly getting pretty bad in different ways.

Mark Levis:
How does this reflect how we’re going to… What are the treatment implications of this I guess is what you outlined. Again, I brought up maybe an allogeneic transplant donor having an old marrow. But just, do we need to stamp out every little clone that we see like a cockroach, they’re scurrying within the bone marrow. I mean how does, how is this going to change what we do?

Klaus Metzeler:
It’s really puzzling and actually for me, for someone who’s doing genetic diagnostics in AML, it even starts before you think about treating a patient. I mean when Ravi Majeti’s group and others brought up that concept of pre-leukemic stem cells couple of years ago, we all got very excited because we thought, oh, we will be able to screen for pre-leukemia just as we screen for colon cancer, and we have a clear implication for let’s say preventive therapies to catch those patients early and stop them from developing AML essentially.

Klaus Metzeler:
But now what we’re learning is that the more closely we look with very sensitive assays, basically almost everyone above a certain age seems to have clonal hematopoiesis. Then the question becomes, how do we really define the risk and identify who will progress to some sort of malignant phenotype before we can even think about treating them. That’s where I think we need to do the work first. But then of course I think we have two different implications.

Klaus Metzeler:
The first is in healthy individuals who have high-risk mutations. And some of them are actually targetable. Think about IDH inhibitors or spliceosome modulating drugs. You could think about early interventions, but then we also see that these pre-malignant pre-leukemic clones persist after chemotherapy.

Mark Levis:
So you’re using your targeted therapy to swat this stem cell, but oh wait, that’s the stem cell. And so that one doesn’t have the mutation. And so… I actually was thinking Ravi when you were talking of the scene from this Harry Potter movie where they were trying to grab this Golden Cup and when you touch it, it exploded into hundreds of cups that look just like it.

Mark Levis:
Which one was the cup? Which one is the stem cell? Oh wait, it’s this one? No, it’s this one. So, the clonal hematopoiesis, is that in the stem cell? Well, actually it could be in two different stem cells. One with this clone, one with that clone.

Klaus Metzeler:
I mean that is actually an interesting question. Maybe also To Ravi. I mean, it comes down to the question, if you have a pre-leukemic clone with a DNMT3 imitation, is that a valid therapeutic target? First of all because what you said, and maybe several other clones waiting in line and if you manage to hit that one with a specific drug, you can think about whether immunotherapeutic approaches may target several apparent clones at the same time. But if you’re very specific, you may just eradicate one clone and come up with another.

Klaus Metzeler:
Then the other question is, if that clone acquires additional mutation, for example, FLT3 transforms into leukemia clone, is the DNMT3 imitation still required to sustain the growth of that transformed clone, or can you still target that at that time point? What’s your opinion on that?

Ravi Majeti:
These are great questions. I think the concept of the pre-leukemic mutations, which are predominantly epigenetic regulators. Do they set the stage for mutations like FLT3 kinase mutations to then drive the disease, and are they still required at that point? I think the jury is still out. There is a little bit of evidence, experimental evidence from a TET-inducible IDH mutation model suggesting in those cases that the IDH mutation is still required.

Ravi Majeti:
I think we’re seeing that from the patient data as well. But in terms of targeting the pre-leukemic stem cells, I think there’s an important point that may not be obvious when you think about the different mutations and whether they are targetable, like the spliceosomal IDH. In fact it’s that, context matters and the same mutation in the stem cell may not be targetable by a targeted inhibitor as it is in a different cell.

Ravi Majeti:
I’m always drawn in many cases to our thoughts on chronic phase CML where the kinase inhibitors, imatinib and its cousins are very effective at modulating and disrupting and killing the progenitor compartment, but they’re not as effective in the stem cell compartment; same mutation, very similar cell and very different requirements.

Ravi Majeti:
You could imagine that same biology being recapitulated with other mutations. If we think about BCR-ABL as a pre-leukemic mutation, which in many ways it is, versus IDH mutations or spliceosomal mutations, they may have very different requirements.

Mark Levis:
We’re beginning to get a glimpse of the chaos going on within the marrow stem cells, clonal hematopoiesis. But I think we’re still getting, we have a much better picture than we did 5 and 10 years ago. We’re beginning to actually see how this all works. I think it’s actually pretty exciting.