Thank you, everyone, for joining the Fireside Chat today. With us today, we have Jim Dentzer, CEO of Curis. I'm Soumit Roy, Head of Healthcare Research at Jones. Really, really glad to have you here, Jim. I would say Curis has really led the field of IRAK4 over the years. Although, you know, the oral inhibitor is being developed in hematological malignancies, I have seen some of the data where it has indications of even functional and solid tumors. Not going to go there. We're going to stay with the hematological malignancies, and we would really love to understand more into it. Would love, Jim, if you can start with an overview of the IRAK4 program and the EMA and the leukemia trials that you're doing it.
Sure.
Brief overview, and then we'll go into it.
Excellent. First, thank you for having me. Really appreciate it. Curis has been the leader in IRAK4 now for several years. There are more companies coming into it as we have started to publish data, and it appears to be a terrific target. The clinical data and the preclinical data are really lining up nicely. We have been focusing in on NHL and AML. We do have five ongoing studies through ISTs in solid tumors, but you are right, it is a little premature to go down that path yet. So far, the NHL and the AML data are really quite compelling, and I think that is what is driving interest in the target.
Right. You kind of really narrowed down to a specific subset of the lymphoma side, but connect the lines where the mechanism of IRAK4 fits into the standard of care, BTK inhibitors, and how you see it really affecting both the leukemia and the lymphoma side of it.
Sure. Mechanistically, the importance of IRAK4 as a target is a little different in NHL versus AML. Obviously, the same target, but the purpose and mechanism work slightly differently. On the NHL side, all of the NHL indications, there are six of them that get used or where BTK inhibitors get used today, are all driven fundamentally by NF-kappaB. In the NF-kappaB complex, which governs apoptosis for the malignant cells, when that NF-kappaB complex gets overactive, what you really want to do in treating the patients is downregulate that activity. The way to do that for the last 10 years is with a BTK inhibitor, and those block the BCR pathway, which is one of two pathways that drive NF-kappaB. It has been very successful following pharmaceuticals, which is now AbbVie, J & J. There are four companies, about to be six, all targeting the BTK kinase, which all downregulate NF-kappaB.
There's only one company that affects the other pathway driving NF-kappaB, and that's Curis with IRAK4. The idea would be within NHL, if you want to maximize downregulation of NF-kappaB, don't hit one pathway or the other, block both. It's as simple as that. Pick your favorite BTK inhibitor in our view, block the BCR pathway, that's a good thing, add emavusertib to it, and it will make it better. It's really a combination story in NHL.
Talk to me how you decided on, you know, we are on a broader lymphoma side, and now it narrowed down to this CNS lymphoma patients only. The transition happened, you think there is maybe this is more active in PCNSL for a certain reason versus the larger NHL population?
I think it's going to be appropriate commercially for all six. I mean, let's start there. The long-term view for IRAK4 in NHL is wherever a BTK inhibitor is being used, adding this ought to make it better. Now when we look at the six indications, which is the most attractive to go after for our first label? If you assume the long-term plan is to go after all six, you pick one first. We picked primary CNS lymphoma fundamentally for three reasons. The first is we know that primary CNS lymphoma is associated with a high degree of toll-like receptor activity. That means, all things being equal, in a relatively small number of patients, just 20 patients, we should see a signal that this drug is working. We're already seeing that. That worked out well.
The second reason is from a regulatory perspective, there are no drugs approved. BDK gets used, it's in the NCCN guidelines, it is a standard of care, but there are no drugs that have ever bothered to go through the regulatory process for approval in either the U.S. or Europe. There is one BDK that's in development in the U.S., but there are none currently approved. We were assuming that if we got good data and we went to the FDA and EMA and asked for favorable treatment for an accelerated path, that we'd be successful. In hindsight, that's what we just got. We just announced that a couple of weeks ago. The last one is, of course, just commercially, it's an ultra-orphan disease, very small number of patients.
In that kind of environment, we would hope to be in a position to have a favorable discussion with payers when it comes to It's a little premature to talk about that just yet.
Right.
Those were the three reasons, and so far it's panned out exactly that way.
Talk to us a bit about how these patients are even treated, standard of care in the front line, how they're diagnosed, treated, and when you look at lymphoma CNS versus non-CNS, you are removing the non-CNS patients. What's the split like?
Yeah. They are primarily CNS patients in primary CNS lymphoma, but we are also talking very small numbers, right? We are splitting minute hairs in a very small market. Within CNS lymphoma, the patient course of treatment is really in line one, it is chemo and radiation, typically with high-dose methotrexate. It works pretty well. If the median survival for patients is roughly three and a half years, they are going to spend a year in that initial line with chemo and radiation. They will spend the next two and a half years cycling through any number of additional treatments. That will include a BTK inhibitor.
Is six months the time to progression after methotrexate?
About 12.
12.
Typically they're in 12 months in that first line, and then they'll progress, and then they'll go on ibrutinib, then they'll progress, and then they'll try anything else they can. Some will try Rituxan, some will try LEN, they'll try anything that they think might work. If they're really in good shape, they might try another round of chemo. They're going to cycle through these various treatments for the next two and a half years after line one. We're focusing, of course, in that group. We're doing both BTK naive, so second-line patients, as well as salvage line, patients who've just progressed on a BTK.
Recently you put out a new set of data with additional three patients and also, we saw the unconfirmed CR from the previous update got confirmed. There's a little lack of response in the new three patients that got added. A few years ago, we have IRAK4, the long-form mutation was one of the main focuses. Is there some reason there, some kind of genomic variability among the patients that's causing the difference in response?
No. It's a great question. The long isoform of IRAK4 is really important in the context of AML, not really NHL. The reason why it matters so much in AML is because there are two isoforms, right? The short isoform and the long isoform. Short isoform is wild type, and it's fine. The long isoform, unfortunately, the death domain is conserved, and it binds to MYD88 within the toll-like receptor pathway. It binds to MYD88 and forces constitutive activation of NF-kappaB. In that case, addressing the long isoform specifically makes a lot of sense. In the context of NHL, it's not that IRAK4 is a direct driver of disease. It's more that IRAK4 is a convenient channel to shut down the toll-like receptor pathway. We don't really see a difference. We won't expect to see long isoform patients.
It's more by shutting down, even the short isoform, by shutting down IRAK4, you shut down the toll-like receptor path, and that's what forces downregulation of NF-kappaB in non-Hodgkin's.
Anything you could discern from these 13 patients? What could be? Is it the disease burden, any kind of genomic factors, or is it time to progression was very short in some of these patients who are non-responders, so they might be refractory to begin with?
Yeah. Let me start with the general overview of there are a number of things that we're seeing that correlate with better outcomes among our patients. The first one and the most obvious one is earlier stage disease patients tend to have better outcomes, right? If in our naive population, for example, we're getting a 71% ORR, we're getting more like a 30% ORR in the BTK experienced patients. Of course, earlier is better. That's true of every cancer. It's not really a surprise. The other thing that we notice when we look at the data is tumor size at baseline. Again, this is not a real surprise, but the larger the tumor at baseline, the more likely that you're going to have a difficult disease to treat.
When we looked at a cutoff of 300 mm squared, so roughly the size of a grape, if you have a primary lesion that's the size of a grape or smaller, you have a really good chance of getting a response even in the relapsed refractory setting. If you have one that's bigger than that, you might get tumor reduction, but your odds of getting a response are much lower. We are still analyzing the genetic factors. Stay tuned. Hopefully by ASCO, we'll have some data that we can publish about correlations there. We do know that there are specific genetic mutations that are associated with BTK resistance. Depending upon what's causing that BTK resistance, we may expect to see a difference in our data.
For example, if the reason that a specific mutation causes BTK resistance is because it forces an escape path from the BCR pathway to the toll-like receptor pathway, and that happens, then we would expect those patients to do really well on our drug because we're shutting down that path. If it's some other mechanism where they're just making BTK less effective, but it doesn't have a corresponding impact to TLR signaling, then those patients, half the drug that we're giving them in the combination won't work as well.
Remind me the drug dose level. This was 200 mg b.i.d. the 13 patients. Anything to do with the CNS exposure of the drug?
Yeah. We're exploring that as well. We've done a lot of studies outside of NHL, but even within NHL outside of primary CNS lymphoma, I think we're over 225 patients now. Broadly, and of course, there's variability patient by patient and tumor type by tumor type, but broadly, 300 mg is a good dose for mono and 200 mg is a good dose for combination. Within that, though, we get responses at all three at 100 mg, 200 mg , and 300 mg. That was a source, as you may remember, a few years back, of a lot of questions from FDA.
Not acting that good.
Oh, the high-class headache of having a drug that works at multiple doses, right? We do not see, other than the general observation that we have had before, that 200 looks like it is a superior dose for combination with BTK.
I will ask this question, and we have unfortunately seen this been an issue when the drug goes into phase three and fails. MO was never selected for CNS penetration. You see it is working, it is getting there. Is it attributed to poorer or damaged blood-brain barrier in these cancer patients post-radiation, post-chemo and everything? Or do you want to go ahead and go to the second, try to select for something more brain-specific?
Yeah. So we actually do get great blood-brain barrier penetration from prior PK studies that we have done. That said, we are continuing to assess the data that we're getting in this study and to evaluate that. Some of the assays for determining what the penetration rate are tricky to interpret, right? At best. At the end of the day, maybe the best, most reliable indicator of whether or not we're getting penetration is whether or not we're getting tumor reduction. Broadly, as you know, in the 20 BTK experienced patients that we've seen, these are patients that are on a BTK inhibitor and progress, their tumors are now growing. We now have MRI data back for 13 of those patients, and nine of them are showing a reversal of disease. They're progressing on ibrutinib. We keep them on ibrutinib and just add emavusertib.
Now the tumors start to shrink. In fact, six of them, they've shrunk so much we got a response. For me, that's the most reliable indicator that we're not just penetrating the BBB, but we're engaging and we're engaging successfully. It is tricky to try and interpret the PK.
You recently announced potential path for accelerated approval, 30-40 additional patients in the PCNSL. Can we call this? When can we call this a pivotal trial?
Right now.
Right now we can call it?
Yeah, absolutely.
All right. With 30-40 additional patient enrollment in the next 12-18 months?
That's right. And it's an estimate. It's hard to find these patients. There aren't a lot. There are 3,500 patients that are diagnosed, newly diagnosed every year in the U.S. and Europe, the top five countries in Europe, largest five countries in Europe. We have roughly 30 clinical sites just over. The way we forecast is one patient per clinical site per calendar year. Now there are going to be some sites like Dana-Farber that might have more than that. And there'll be other sites that might not have any. But on average, in this ultra-orphan setting, we're feeling pretty confident that one patient per year per clinical site is enough. With 30 clinical sites, that means in 12 months, we'll have 30 incremental patients. If you want 40 patients, that's 18 months.
How many sites are you in now?
I'm sorry?
How many sites?
Just over 30.
Just over 30. You do not have to, you are almost there. Okay.
Yeah. Our sites are really well split. We're in all the key centers for primary CNS lymphoma in the U.S. and Europe, and then we have three sites in Israel as well.
What's the interaction or the reaction from the investigators when you go and show this data, hurting patient data and their interest in enrolling patients?
They get excited. I think the single biggest thing that we have going for us when we have these conversations is the consistency. The mechanism of action is clear and well established. All of the KOLs know it. The preclinical data support it. What you would expect given the mechanism is what we saw in the lab in multiple presentations. Now that we have clinical data to show as well, it's working exactly where you think it should work, and it doesn't work in places where you think it shouldn't work. That predictability gives everybody high confidence. The flip side of that is what's the hurdle when we talk to them? The single biggest hurdle is it's hard to find these patients. As I say, there are only 3,500 of them diagnosed per year.
Only 2,200 of them are going to go to second and third line. Now you're following these patients that are on a BTK inhibitor, mostly ibrutinib, but increasingly we get patients that were on tirabrutinib in their clinical trial. As they fail, then they're eligible to come on. What we try to do in our conversations with clinical sites is, of course, keep us in mind as you're tracking your patients that are on a BTK inhibitor. As they progress, they're going to be eligible for our clinical trial.
As long as they're in reasonable shape and you think that they could stay in a clinical trial for more than a couple of cycles, then leave them on their BTK inhibitor, add emavusertib to it, and let's see if we can reverse the course of disease the way we've been able to with most of the patients so far.
Switching to AML, the staggered value inflection one can say. Currently, you are in 7-day safety and then 14-day, 21-day. This is frontline setting post-relapse after Veniza. No, this is Veniza combo with Emma. Talk to me why you have to go through this process of 7-day, 14-day, 21-day, and then that pushes out your efficacy data all the way to like.
It does. Yeah. It's something that's a little frustrating now. A year from now, we're going to be so happy we did it this way. The thought process here, let me start from the beginning. That we've got a drug that in the context of AML, the mechanism's a little different. It works in monotherapy in one-third of the population. FLT3 patients, we also hit FLT3. In FLT3 patients, they can take this as monotherapy in a relapsed refractory setting. Almost all patients with AML express that long isoform of IRAK4, the one that's causing constitutive activation of the MYD88 and therefore the toll-like receptor path. All of those patients should have emavusertib added to their regimen. Their regimen today is venetoclax and azacitidine. It works pretty well, 60-65% CRs, right? We would hope to do better than that, but that's a pretty high bar.
It's a high bar.
The concern that we wanted to address with our safety run-in study was if we started treating patients on the triplet on day one, and they did not already have exposure, meaning the PIs did not already have experience with Emma in combination with Veniza. If they saw any safety issues, some of them might be tempted to say, "It's the new drug." We know that Veniza independently and certainly together, they are tricky to dose. They are toxic. They are cytopenic. We did not want to have that risk. The way to take that away is showing patients that adding Emma to Veniza is okay, that there is no DDI. The way we constructed that run-in study was take them in front line, but only those patients on Veniza who have been able to get to a CR. Now their platelets and neutrophils have fully recovered. At that point, we add Emma.
We can't, of course, get them to a deeper CR, right? They're already in CR. What we can do is show with data that adding emavusertib to venetoclax doesn't cause a problem. If it does, then, of course, we'll have a different issue. So far, we're trying three different doses to sort of escalate up. We're trying to mimic the way people dose in the real-world setting. Let's try 7 days, then 14 days, and then we'll try 21 days. Assuming that we can get through all of those hurdles, we can move to a cycle one day one dosing regimen.
This is the thing. You are active in FLT3 mutant population who relapsed on the prior FLT3 versus so you could stay there as a monotherapy and get approved in second line, third line versus going into Veniza combo in frontline, longer trial. Yeah. Also likely physicians will say, "Look, I can get two-thirds of the patient in CR, push them toward transplantation." Why 67% improving it to 75%? You have to run like 300-400 patient trials. That is much more expensive, much more longer. Yeah. Why push that angle rather than just getting it as a monotherapy and see?
It's a great question. It's hotly debated internally, I will tell you. It's hotly debated when we talk to the KOLs. It really depends on which KOL you speak with. If you speak to someone who sees a lot of FLT3 patients, in their view, the data are really compelling. This looks like it's a best in class for patients that have a FLT3 mutation. There are three that are approved, right? Midostaurin, Quizartinib, and Gilteritinib. The data for this as a monotherapy look considerably better than the other three. They're all over the relapsed refractory monotherapy study, get it approved, let them use it in the clinical setting, let it be commercially available. If you don't see a lot of FLT3 patients, what you want is something that makes the efficacy increase in your frontline setting.
They all love the idea of IRAK4 as being a novel target. They all know venetoclax doesn't hit it. They all know it's in every patient. They have seen the preclinical data, which are really compelling, that this is something that could really improve the efficacy. I know 60-65% is a terrific CR rate. They are, of course, looking for something even higher. No question, some of them will want to use it as a bridge to transplant. Absolutely.
Sequence it as something to do for these relapsed patients.
Exactly.
Because there's nothing out of.
That's right.
After Venetoclax, nothing.
That's right.
Efficacy, relapse again, now experimental.
That's right. By the way, you mentioned something in passing. We also get not just triplet, but interest in an oral-oral doublet in front line of just emavusertib and then drop the AZA component.
I thought that was a good one.
I'm sorry?
That was a good one because then you don't have the safety concern with AZA, and you can treat these patients who cannot go on 7 plus 3.
Yeah. As I said, it's an interesting wrinkle. It's a high-class question for us of which one we fund because we're sort of resource-limited. We need to be capital efficient. Depending upon which investigator you're speaking with, you get a lot of passion and enthusiasm for one particular regimen. It starts with, do you do relapsed refractory monotherapy or do you do frontline combo? Within frontline combo, you've got a lot of different views out there. Increasingly, people like Venetoclax, but they're very frustrated by it. Weirdly, nobody prescribes it the same way, which is a little bit of a nightmare as a drug developer when you try to figure out how you want to write a protocol. As I said, that is something that's of high interest to us.
I think when we step back, NHL versus AML is the biggest decision. Once you go down the AML path and you start saying, "Well, which of these can we fund?" you immediately go to, "Do we fund the frontline or do we fund the FLT3 relapsed refractory study?" Within that, how do you do it? Do you do a head-to-head in FLT3? A lot of people would like that. Or do you do, as I say, the triplet or the doublet in frontline?
Great. What do we expect? Data, mid-year, ASH, or should we wait till ASH to get the next update?
Yeah. We will definitely be at ASCO and EHA. Given that we've just given a data update, there's not a whole lot of incremental data to show other than some of the biomarker data that we're collecting and some of the genetic data that we're collecting. That could be of sort of scientific interest, I think, at ASCO and EHA. ASH is a good time for.
ASH is, yeah.
More efficacy update in PCNSL.
It gives you some room to enroll.
It does. As I say, we're enrolling a patient per calendar year per site. If you assume that by the time we have abstracts due, we'll have a lot more data we can put in. Of course, you update it right before the conference. I would expect that would be an opportunity for us to have more incremental view.
Great. Thank you again for joining us.
Oh, thank you.
You want this story to work. We are looking forward to the next update.
Thank you very much.
Pleasure.
I really appreciate it.
Hope you're enjoying the.