MDS updates from iwMDS-iwMPN 2024 | Inflammation & immune interventions in MDS: targeting STAT, IRAK4 & more

Dr. Andrew Wei:

Hello, my name is Andrew Wei, and we’re currently here in Boston at the International Workshop for MDS. And we have with me three panelists who have just given their discussions today on the topic of inflammation and immunity in MDS. I have Dr. Shastri, Dr. Platzbecker, and Dr. Zeidan. Good afternoon to all. Dr. Shastri, you discussed the topic of STAT in MDS. Did you want to tell us briefly why do you think STAT proteins are important and what role do they have therapeutically for MDS?

Dr. Aditi Shastri:

Thank you, Dr. Wei. So STAT3 is something that I research in my lab. And STAT3 is part of the STAT transcription factors, which is downstream of JAK. They do have an important role in immunity. They regulate Tregs and are downstream of JAK. And when stimulated by a cytokine like IL-6, they get activated, phosphorylated, and mediate their effects related to cell proliferation. STATs can be dysfunctional in many malignancies, including myelodysplastic syndromes and myeloproliferative neoplasms. And something that we found and published before is really that STAT3 is upregulated in hematopoietic stem and progenitor cells from MDS patients. However, being a transcription factor, STAT, as you can imagine, is very difficult to target. But we do have state-of-the-art and better drugs now. So we are looking at preclinical data in the preclinical and clinical context with antisense inhibitors of STAT3 as well as STAT3 degraders. So we hope that by downregulating STAT3 function, we can normalize these aberrant hematopoietic stem and progenitor population and explore some combinatorial approaches as well.

Dr. Andrew Wei:

And if we did a study on STAT, targeting STAT, would we target all MDS, or how we narrow down or identify patients that are potentially STAT targetable? And what type of therapy do you think would be the most logical: molecules or degraders?

Dr. Aditi Shastri:

So we actually do have a study that we just activated at Montefiore Einstein Comprehensive Cancer Center, where I work, and will shortly be activated at MD Anderson Cancer Center, where we have patients that are relapsed/refractory to prior therapies, including hypomethylating agents or venetoclax. And we have some preclinical data to show that they actually have heightened expression of STAT3. So these would be specific subsets we would like to target. Mutationally, we are agnostic at this point. So we will be giving an antisense inhibitor of STAT3 called danvatirsen as monotherapy, followed by combination with venetoclax.

Dr. Andrew Wei:

Uwe, you gave a discussion today on the role of IRAK targeting in MDS. Do you want to tell us why is inflammation important in MDS and where are we up to with respect to therapeutically targeting this class of proteins?

Dr. Uwe Platzbecker:

Yeah, what I did in the session is to highlight the backbone of knowledge we have with regards to IRAK4 as a potential target in MDS. In general, IRAK4 is actually a kinase which is involved in autoimmunity. It’s part of the innate immunity, and if over-activated, it’s linked to also part of rheumatologic diseases. What we know in MDS, but not only in high-risk, lower-risk, but also in AML, is that the IRAK4 pathway is activated, and it’s basically co-opted by the myeloid progenitor cells. And by doing so, this drives disease proliferation, leukemogenesis, and therefore IRAK4 inhibitors have been developed in order to modify the disease course and potentially also improve ineffective hematopoiesis. There’s one IRAK4 inhibitor currently in clinical practice in clinical trials, which has been shown to have some single agent activity in predominantly FLT3, but also spliceosome mutated patients. The study is still ongoing. Also, combination trials with AZA+VEN have been started.

What I also highlighted is that these kind of drugs may be an interesting approach in lower-risk MDS patients where inflammation has been shown to drive early disease propagation and such a trial is currently actually run in Germany under our sponsorship. On the other side, there is some recent data that IRAK4 inhibition may not be the end of the story because also IRAK1 can actually compensate for IRAK4 inhibition. So therefore, dual inhibition of IRAK1 and IRAK4 may be the way to go. And again, there are some studies planned to do so. So I think in general it’s an interesting target with also some drugs being currently piloted in clinical trials, whether this is a standalone therapy, whether this is something for high-risk or low-risk for very early low-risk patients as we discussed also with the audience has still to be shown, but so far I would say at least the safety of the drug seems to be reasonable and may also pave the way for long-term treatment.

Dr. Andrew Wei:

With respect to IRAK inhibitors, what would be the most logical endpoint? Would it be symptom alleviation or disease modification?

Dr. Uwe Platzbecker:

So far from the, let’s say, from the preclinical work, it would be rather disease modification because the drugs preferentially target the leukemic stem and progenitor cells because this pathway is, of course, important as an innate immunity pathway, but it’s high-checked if you want by the cancer cells or by the leukemic cells. And therefore, disease modification should be the key endpoint for the trial so far. But it also alleviates symptoms like we sometimes see fatigue, sometimes also unrelated to transfusion support needs to be shown, but so far it’s more disease modification.

Dr. Andrew Wei:

Side effects with this class of drug?

Dr. Uwe Platzbecker:

Rhabdomyolysis has been observed, dose-dependent, at least with the drug I was talking about today. This was dose-dependent, with lower doses, it was not observed. So otherwise, apart from some heme toxicity, maybe based on the FLT3 inhibition, which is also something which has been observed. I think the drug seems to be, it’s an oral drug, seems to be very well tolerated.

Dr. Andrew Wei:

Thanks. So, Dr. Zeidan, you presented today, as a follow-up from EHA, the results of the pivotal Phase III STIMULUS-2 trial. Did you want to tell us the results of that trial, and where do you think we go now?

Dr. Amer Zeidan:

Yeah, so the STIMULUS-MDS2 trial was a randomized Phase III trial that looked at an agent called sabatolimab, which is an anti-TIM-3 antibody. So this is based on the idea that TIM-3 is an immunoregulatory receptor that’s not only expressed on T-cells but also on the leukemia stem cells and the blasts, and targeting it with this antibody has suggested promising early clinical results in Phase I studies. And based on that, the drug was taken to a randomized Phase III trial in combination with azacitidine, which is a standard of care in patients with high-risk MDS. So this trial was a double-blind, very large more than 500 patients international study, more than 30 countries. And the primary endpoint of the study was overall survival. So the drug basically was well tolerated, however, the primary endpoint was not met. However, there was a statistical trend for improvement in the overall survival, which was 22 months with the combination versus 18 and a half months with the azacitidine and a placebo, which was an advantage of three and a half months with a p-value of 0.08, which did not cross statistical significance.

When we looked at different subgroup analysis, we could not clearly identify a clinical laboratory or a molecular subgroup that benefited. We particularly made close attention to TP53 because of data suggesting that those patients could benefit more from immune checkpoint inhibition. We had 25% of patients of those enrolled on the study who had this mutation, but there was no clear benefit. There was a subset of patients who had the DNMT3A mutation, which seemed to derive some benefit, but I think small numbers and we need to understand this better. But you look at the overall results of the study, many of the efficacy endpoints, including complete response, leukemia-free survival, progression-free survival, all of them have trended in favor of the combination, which leaves the question open is: are you seeing some general mild positive effect that the sample size is not big enough to detect, or is there a subgroup that’s deriving most of the benefit but we still cannot clearly identify it? And I think this is where the current challenges with this particular antibody.

Dr. Andrew Wei:

So in the last few years, we’ve now seen unfortunately the demise of eprenetapopt, sabatolimab, magrolimab, and also pevonedistat. So I will go through each of you in turn. What have we learnt? What learnings can we take from these experiences, and what should we do with respect to MDS study design in the future to increase our chances of success? Aditi?

Dr. Aditi Shastri:

Yes. So we have had quite a learning curve in the last few years, especially with MDS and a lot of drugs that failed in the later phase clinical trials. I think one of the pivotal lessons we might have learned is that one of the best drugs we have for treating MDS are hypomethylating agents. And anything that sort of gets in the way or stymies the effect of the AZA, azacitidine, or Dacogen is something that is not going to be an optimal combination with it. So I think that this is something we have to, we cannot leave these combinations to chance. I mean, I think they really need to be rationally well evaluated to be good combinations with hypomethylating agents before we take them into clinic as combinations. There has to be sound, I think, scientific rationale for combining them. And the other thing is, I think that as we go ahead, we should just keep this in mind as we go along with clinical trials that if we have to improve upon the HMAs, then biologic rationale is key.

Dr. Andrew Wei:

Uwe?

Dr. Uwe Platzbecker:

I think I will only add a minor point, which is given the heterogeneity of the disease, which I think is maybe more heterogeneous than we thought before. We need, of course, these biological rationales to combine agents with HMAs, but we also need a decent Phase II sample size in order to make a conclusion and which also drives subsequent Phase III trial. I think a small Phase/Phase II with a CR rate of 38 or 40% in a ECOG zero. The Phase I/Phase II population is younger, is maybe not a reflection of the 75-year-old high-risk MDS patient should not drive us into large Phase III trials where sometimes you may even put patients at risk with combination trials, you may alleviate, even like with entinostat in the old days, it actually alleviated the activity of azacitidine. So I think this would be something I would like to envision for future Phase III trials in high-risk MDS patients.

Dr. Andrew Wei:

So when you say a decent sized Phase II, is that for the purpose of deciding how big the Phase III should be or to actually narrow it down to a subset?

Dr. Uwe Platzbecker:

To have the capacity to narrow it down to a subset or if there is no subset in, let’s say, a sample size of 80, let’s say 80 to 100 patients, Phase II, then to come to the final conclusion that this is an all-comers drug which will also show a benefit in a 500-patient clinical trial. I think that would be my advice.

Dr. Andrew Wei:

I mean, what’s your understanding with respect to what Uwe has just said? We tend to do a Phase II, and then we do a Phase III, as has been the case with STIMULUS. Why don’t we do adaptive studies where we do a Phase II, result, analyze, and then use that to launch into a Phase III if it’s justified? Why do we have to, in my mind, be so inefficient by running these separate studies? Can we not design smarter studies?

Dr. Amer Zeidan:

I think we can. I do think there are competing forces when you think about this issue. I think from a sponsor perspective, sometimes there is a concern about how quickly they can get to a registration in Phase III because this is the only way we can get the drug to the market. So I think there has been a tendency, sometimes maybe some over-enthusiasm for some of the drugs. I would add that in my mind, what happens is many Phase I/Phase II trials is that you have a 30-40 patient study. The CR, as Uwe mentioned, could be 40-50%. A lot of enthusiasm for that particular study just by selection because the studies that have 20% CR, nobody looks at them, and you have like a 30-40 patient study with a 50%. Everybody’s so excited. Sometimes the Phase III trial is initiated even before the same study continues to accrue.

And you see this what I call regression to the mean. By the time you are 80, 90, 100 patients in the study, you are seeing the CR going down and down. So it’s closer to 30 to 40%. So I would argue that we have not yet seen a single randomized Phase III trial that failed because the CR was disconnected from the OS. In every single failed Phase III trial, the CR was not actually statistically or meaningfully better with the combination compared to the monotherapy. And I think the issue was that we probably moved drugs too quickly, it probably should not have been moved to the Phase III. But I would add also with some drugs like the magrolimab, I don’t think that toxicity was well appreciated early on, largely because the patients who are managed in big specialized centers with a lot of blood bank supports, etc.

And once you go to Phase III trial with many size, hundreds of size less experienced investigators, less resourced hospitals, some of those toxicity issues could make a major impact on the trial. And I think having seen the ENHANCE presentation, you could see very early separation in the curves of this trial, which probably suggest a toxicity on top of not having efficacy. So I think all of these are important. Last one is, I think for me, the bottom line is that a single-arm Phase II study should have at least an 80-patient sample size before you go to a randomized Phase III trial. But my own feeling is that you probably should have a randomized Phase II before the randomized Phase III to get a better sense not only of the efficacy and the safety but, as you were saying, to better understand the biology and probably the biomarker impact with a good control arm.

Dr. Andrew Wei:

So today we’ve really covered some very challenging topics, from new targets to some of the difficulties in doing randomized studies and getting registration in MDS. So thank you for joining us, and we hope you’ve enjoyed this session.