iwAL 2023 | Clonal hematopoiesis, early detection & potential therapeutics

I’m Ravi Majeti from Stanford University, and I’m here with my colleagues from the last session of the iwAL workshop. I’ll let them introduce themselves.

Courtney DiNardo, MD Anderson, Koichi Takahashi, also from MD Anderson, Uma Borate from the Ohio State University.

Our session was really focused on different aspects of clonal hematopoiesis and early detection and potential therapeutics in this space.

I’ll just give a summary of my presentation, really focused on different types of experimental models that we’ve been using, where we’ve taken the ability to use CRISPR engineering in normal human hematopoietic stem and progenitor cells to engineer the mutations of pre-leukemia and clonal hematopoiesis, and investigate their functions and potential therapeutic vulnerabilities in experimental systems, both in vitro and in vivo. And we’ve specifically been looking at IDH1 mutant cases and trying to understand the potential impact of mutant IDH1 specific inhibitor ivosidenib versus some novel therapeutic approaches in our case, oxidative phosphorylation inhibitors of complex I that seem to have a preferential effect in actually eradicating stem cells that harbor IDH1 mutations, unlike ivosidenib, which seems to promote the differentiation of those cells. And I discussed some of our ongoing work in using additional methods of sequential modeling of mutations in normal human stem and progenitor cells, and looking at both cell non-autonomous and cell autonomous effects at promoting the progression and expansion of those clones.

Those are areas of active interest to our group that are a little bit relevant to thinking about the downstream translational therapies for clonal hematopoiesis. So maybe I’ll turn to Courtney and you can give a summary of your presentation. So My presentation focused on hereditary leukemia predispositions. And what I did was kind of talk about the fact that really about five to 10% of leukemias are anticipated to have some sort of underlying predisposition component, which is really kind of a recently identified finding. We, you know, we’ve heard all about, inherited solid tumors and, and it just is, is a more recent realization. And so I talked about that and I talked about just the challenges with trying to identify patients with hereditary leukemias because typically, you know, people get genetic testing from a blood sample or, you know, a saliva sample.

And in patients with leukemia, that’s actually their tumor. So you can’t discriminate between what’s a kind of a cancer associated somatic mutation and what’s an actual inherited germline mutation. So, you know, you have to be a bit more smart and unfortunately, a bit more complicated to do genetic testing  for germline mutations in patients with leukemias. And I talked about a couple different examples. And in particular, DDX41, which is one of the probably the most common inherited leukemia predisposition about 1% or so of all MDS or AML patients will have a DDX41 mutation and almost all of them are germline. And so just tr trying to kind of raise awareness and make people realize that this is actually more prevalent than I think a lot of people appreciate.

Yeah. Fantastic. Kochi, please.

My talk was focusing on the risk stratification of clonal hematopoiesis or CHIP. I reviewed the recent studies investigating the genotype specific risk of the CHIP and, both from the case control studies and also the observational studies. And I summarized that these studies have led to the identification of certain CHIP mutations which confers highest risks, or leukemia potential, those include p53 mutations IDH1 and 2 mutations, and RUNX1 mutation as well as, splicing factor gene mutation, and also JAK2 mutations. And I also went over some of the web portal tools that enables the risk prediction for certain individuals with having CHIP. and I think the uses of these web portals will enable us to identify the CHIP patients who are at highest risk of potential for the leukemia transformation, which may lead to the targeted interventions for those. Yes.

Speaking of interventions, Please.

So, thank you. So leading off Dr Takahashi’s presentation, I discussed the challenges and the opportunities of potentially intervening in these high risk patients. As Courtney mentioned, we have, you know, we sort of haven’t really kept up with our solid tumor colleagues in terms of their strategies for early detection, early intervention, and potentially cures. It’s more complicated in blood cancer, especially leukemias to one, identify these mutations early, but even more importantly, figuring out which one of these patients are particularly high-risk that they are going to develop a full-blown blood cancer in the future. And so, you know, based off some of the talk and the studies that Dr Takahashi summarized I think we now have the opportunity to identify these high-risk individuals and then potentially design studies to intervene in this patient population so they don’t develop a blood cancer.

So I specifically discussed a study that I am leading where we target a very, I think broad but important player of inflammation called IL-1 beta And we’re doing this by using an antibody called canakinumab in half the patients in this trial. And the idea is to take the first step towards learning whether we can do this successfully, and then also understand through this trial how these patients evolve when they do or do not get the therapy. But then last but not the least, also understand what success looks like or, you know, how not being successful looks like. What are we really measuring?

Are we measuring the possibility that these patients do not progress to MDS or AML? Are we measuring a reduction in these high risk mutations? Are we looking at, we know what, what this does to inflammation in their body and, and how this impacts their disease, and so on and so forth. So I think this is a great opportunity, and I think a really expanding, exciting new area in our field. Yeah.

One thing that brings together a couple of topics, I mean, this is a question for Courtney, is that in the patients who do have germline predisposition syndromes, when they potentially develop a CHIP mutation on top of the predisposition syndrome, how that scenario plays out because now you have a somatic variant that may be interacting with the germline variant?

I know it’s early days to think about this from a data point of view, but love your perspective.

Yeah. And I, it’s unfortunately a complicated answer, but we know probably the most about germline RUNX1 as an example. It’s been, was one of the first identified, and, only 30 to 40% of people who have an inherited RUNX1 mutation will go on to develop a myeloid malignancy, an MDS or AML, so you don’t wanna transplant everyone. Right. That would be, you know, giving a lot of people a really toxic therapy that they don’t actually need. So, you know, one of the things that we’re looking at is, you know, does the acquisition of these somatic mutations, this clonal evolution, identify those patients who are more likely to develop that myeloid malignancy?

So that’s definitely, it makes sense, right? And that’s probably what the story will be, but it’s gonna be different gene by gene, for instance, DDX41, which I mentioned, you don’t really get other mutations other than typically a second hotspot mutation on the DDX, on the other allele. So, you know, it’s not so much a clonal acquisition of other things. It’s whether or not you get that second hit for that particular predisposition.

Great. And, Kochi, one of the things that interests me is, of course, we can start risk stratifying patients with CH but what’s the magnitude of risk? Like how much worse off are the highest risk patients in their potential for progression versus the lowest risk patients?

Yeah. That’s an absolutely great question. I think we need to also separate the relative increased risk versus absolute, risk. And I think if we look at the relatively increased risk, it looks very impressive. Like these high-risk population have like a 90 times higher risk compared to the low-risk patients who do not have a CHIP. But when you look at the absolute risk for the transformation, yes, it looks like even in the high-risk population, maybe the 10-year cumulative incidence risk is anywhere around 40% ish. But still it’s an impressive risk. And on top of it, if you also have, I think a CCUS, you know not only have a CHIP, but also with the cytopenia, I think the previous study also found that the risk of transformation is very, very high, like 80% or close to like even a hundred percent. So I think if you truly identify a high-risk population combining the mutation data plus the clinical variables, I think you can identify patients who really have a high-risk potential to transformation.

Yeah. Wow. I didn’t realize that the absolute risk could be that high 40% to me seems pretty high.

Right? Right. I mean, those are really at a selected high-risk population, but yes. Yeah. Yeah.

And Uma, I think with the early days of clinical trials, I think what really sticks out to me is the endpoints issue. Like you were just alluding to, what are the endpoints? What are we really monitoring? What means clinical benefit to the patient? And as we think about therapeutic development, what means clinical benefit in a regulatory setting, in terms of a new drug or a new indication or an FDA approval? just your thoughts on that space.

And I think this is where the very hard conversations need to happen because, you know, data that all of you presented was if you do have that IDH-mutated high risk patient with cytopenias, with the clone, you treat them with an IDH inhibitor, their cytopenias get better, you know, the clone sort of gets really low where you can barely detect it, but it doesn’t go away. And so is the patient going to stay on therapy, you know, for three years, for five years? And obviously it’s great if they don’t get MDS or AML, but at the same time now they’re taking a treatment which is a pill which is used for cancer treatment for a really long time. So to me, the first question obviously is, is the risk high enough to justify treatment? Second, if you justify the treatment, the treatment really needs to be something that the risk benefit is tolerable to the patient. I mean, we have early stage breast cancer patients that could benefit from, you know, hormone blockers, but a lot of women don’t want to take them, and the side effects are terrible. They get hot flashes. And so it’s just a hard thing to discuss and offer the patient.

And then the third thing is, what does success look like? We believe that improvement of counts means that their leukemia risk is going to go down. Well, great, it’s easy to measure, but if that’s not the right thing to look at, and the right thing to look at is do bone marrow biopsies every few months and make sure they’re not progressing. And that’s our sort of golden endpoint. And I don’t think we know. And so I think that’s where the opportunity is in this field to really hone down and figure out sort of the right way to measure success.

Yeah. Maybe I’ll just piggyback on that and saying, in terms of our experimental systems, we’re really trying to understand if therapeutic manipulations in the engineered mice models, whether that actually will impact the clone size. Ultimately, I think if you eradicate the clone, we would think that would have a clinical benefit, but we don’t know if that’s the only way to have clinical benefit. So we are really trying to understand the clone size and which cellular compartment of the clone is being targeted. So a little bit more basic question, but I hope it will inform ultimately clinical strategies.

Maybe the last question I’ll ask, and you guys are all very clinically active, is, you know, you’re at amazing academic medical centers, you get referrals from your oncology colleagues or people are getting sequenced from different reasons, but the real world is out and about, not inside MD Anderson and Ohio State. And in terms of either identifying these patients or making sure that patients have access to these diagnostic tools and methodologies, maybe are you doing outreach to your referring physician groups? Or how would you recommend community oncology and hematology groups interface with this kind of cutting-edge area of our science?

So as a non-MD Anderson person I will say what you said is so critical because our sort of catchment area is a very large rural population within the state of Ohio, West Virginia, Appalachia, and so on. And it, we have a very scattered group of community oncologists that see every single cancer, you know, and this is sort of one of the rare things they see. So we do try to do a lot of outreach. We speak to sort of their tumor boards, grand rounds conferences like these, I think sort of an important purpose because if people do want to kind of look up what’s going on with these patients, hopefully they’ll find this video that we’re making right now. But I think what you just said is so critical because until the patient gets identified and sent to us, none of these things that we’re talking about can happen, right? So I really think that is a big challenge moving forward.

And maybe you can talk about your experiences at MD Anderson.

Yeah, I mean we don’t get a lot of referrals from the outside community, and we maybe should probably do a little bit more effort outreaching to the community practice. Our main referral source is actually a, a solid tumor clinic within the MD Anderson. And I think we need to also separately look into the risk of cancer patients having CHIP versus normal individual or community person having CHIP, and they do have a different implications. But I think in this world where like you can easily get, you know, order 23 and me and perform the genome sequencing, at home, I think we are, we need to be prepared to get all these, self-referrals. Or any of these things who were identified to have a CHIP, at the home kits or like a home sequencing kit. So I think that’s a very critical part. Yeah. Maybe just your experience.

I mean, we have, we certainly have a strong referral pattern for leukemia, right? For CHIP, you know, not as much, but I think more and more as, you know, as you all were saying, it’s a lot of things are becoming patient directed and family members directed where, you know, they’re looking up information about their health and they’re identifying kind of, these opportunities are available at X, Y, Z. And so kind of more and more I’m getting kind of direct  self-referrals or emails or phone calls to say, you know, Hey, I have this, you know, how do I come see you? So I think, there’s just so many different levels of, of outreach and of awareness. And I think, you know, what we’ve been doing and we’ll continue to do is try to reach the community oncologists who don’t see so much hem malignancy. And so, you know, for them trying to manage everything is just such an impossible task. I don’t know how they do it. But then also on top, another level of going directly to patients and patients’ family members to have that information available so they know how to seek additional information.

Yeah, absolutely. Well, like when I think back to the initial genomic studies, yeah, maybe almost 10 years ago now, thinking about identifying CHIP, we called it indeterminate potential. Yeah. Frankly, as I thought about it, I was like, wow, this really is a don’t ask, don’t tell kind of a scenario. but I do see the future is really all about us figuring out how to stratify the patients, how to think about their inherited predispositions and where we go with therapeutic interventions.

So, right. And I think the great part about these large observational studies that you talked about is they identified easy clinical parameters that anybody can see looking at a CBC. If you have low counts, if you have a high MCV, the size of your red cells is large, you have a high RDW, just a lot of different sizes. This is something anybody can do. And if they identify that and they see this picture, I think that is a simple educational message to send to people saying, these are the patients you should be worried about and get them to us sooner rather than later, and we can do the rest of the testing and maybe some counseling and risk directed therapy. So I think that’s the value of these studies that have just come out. Right. and I think that that’s gonna really distinguish from, as you said, the don’t ask, don’t tell worried, well, people versus some people that we need to be more active about. Right.

Great. Well, thank you all. It’s been a really exciting workshop.

Thank you.