Can it get us into a single shot?
We have to dose it.
Can it get us into a single shot? You're talking about monotherapy, or are you talking about the combos?
No, I'm talking about the combos as well.
Combos can be a single shot?
I think, well, we certainly are using them as a co-formulation in a single shot. The exact dose of the co-formulation will depend on whether the low or the high doses work the best.
Makes sense.
I think this is a disease that efficacy is king. This is a very costly disease to the patients, of course. Flares here can lead to hospitalizations and surgeries. It is also costly to health systems as well for the same reason. Efficacy really is the goal here. We are going to maximize efficacy first and then kind of figure out what the best product profile is for you.
Okay, I'm going to go, so we're still on the UC trial. We're on the monotherapy. I'll talk about combos in a second, and we'll go to the other trial in a second. On the Part A, when is the readout again?
Yeah, so we have all three of the Part A readouts next year. So, you know, alpha-4 TL1A, and IL-23. They started, you know, a bit, you know, six months between those three of them. And so it's possible that we could read them out individually. It's also, as you know, enrollment in a trial is not linear. You kind of see acceleration over time as you see sites. And so we'll actually maybe be able to see kind of a compression of the readouts relative to the.
Is this like a mid-year-ish thing, or is this a?
We haven't said anything beyond next year currently for these. I think as we get into 2026, I think we'll be a bit more specific about exactly when each of the readouts will come.
The endpoint, I realize it's a UC trial. When I went to the Clinical Trials link, I initially was expecting a Mayo Score, but obviously it's RHI. Could you just remind us the endpoint used and also their secondaries?
Yes. In an open label context, the Mayo Score includes some subjective measurements of the disease. Patients who know that they're on a drug and physicians who know that their patients are on a drug may have some bias on a few of the endpoints here. What we thought is most critical to look at as a primary endpoint is an objective, centrally read endpoint, which is what we're using, the RHI, which is a histologic score. I think what's important about that is it's centrally read, and these readers don't know whether the patient is pre or post-dose. They actually don't know which mechanism they were getting either. We actually are seeing, you know, alpha-4 versus TL1A versus IL-23. I think it's actually more controlled than most open label contexts would be.
Of course, we'll be measuring the other endpoints that people care about too. Endoscopic improvement is another clear objective endpoint that I think matters a lot, and clinical remission and other endpoints that you would expect in a blinded placebo-controlled context.
One thing I saw in the FDA guidance document for these endpoints was on RHI, FDA was concerned that reader to reader, there might be a lot of variability or center to center. I think you mentioned centrally read.
Yeah.
Wherever the central reading is being done, are we sure if they have multiple people, they just read them differently?
Yeah, I mean, I think we are using, I think it's fair to say, one of the best central reading organizations that has done many studies in IBD and has done RHI measures for many of the largest studies that have been run with very experienced readers. I think it's, you know, I think it's the best you can do in terms of control for an endpoint like that.
Okay, great. I mean, just so we understand it, what is the RHI bogey that you have in mind?
When we look at kind of the RHI, it has been used in a handful of studies already. I think we actually have reasonable benchmarks, including on the mechanisms that we're testing. RHI was used in the VARSITY study. That was, you know, the alpha-4 beta-7 against Humira. The range there for adalimumab was about a change from baseline of about five. For vedolizumab, it was a little over seven. Both of those molecules certainly work in terms of efficacy. That was the range there. We also had obefazimod as well as another example, and that change from baseline was about six on RHI. If we're talking about, you know, what's the range of changes from baseline in RHI, those timelines were a little bit different for each of those endpoints, but I think that's a realistic range of where.
Five plus.
Five to seven is where we've seen drugs that have that change on RHI lead to a positive phase III time.
It's on a scale of?
No, it's a change from baseline, right? You could be, you know, you can see, depends how severe the patients are at baseline.
Has a double digit ever happened?
Not in any, you know, mean change from baseline, certainly never over that level.
Not until Skyline.
Yeah.
Skyline combo.
Yeah, no, I think, I mean, I guess our goal for each of these monotherapies is to be at least as good as the first generation products. I think that's what we really hope to see from these things. You know, we're dosing them, I think, to potentially see incremental improvements in efficacy for each of these. I don't think we're going to be doubling the efficacy of any of these molecules or any of these mechanisms by increasing the target coverage. I do think it's possible that you're going to see that level of meaningful clinical benefit when you start to combine mechanisms together. That's really where I expect to see a meaningful efficacy improvement.
Got it. Patient selection, I recall a thing is like a ratio of 50% naive, 50% naïve, 50% experienced. Could you just walk us through how many, so how many patients is that? Twenty patients per arm on the monotherapy? So it's 10/10?
Yeah, so it's more than that, right? We said it's about 100 for this Part A component in total across these three arms. You can, you know, do the math, a little over 30 per arm is the target. There's always some wiggle room in enrollment. You know, it can vary a little bit from there. We're targeting 50/50, but we're only setting kind of strict requirements for that 50/50 in the Part B component of the study. In Part B, it will be no less than 40, no more than 60, right? It'll be in that defined range. In Part A, we're a little bit more open in terms of who we're going to enroll. We don't set us, we're still having that target, but it could vary a little more.
If we were to design this trial from scratch, there's an argument that could be made is you could have gone straight into the Part B. You didn't have to have a Part A. Is that a regulatory requirement?
It's actually really a concurrency challenge for us, right? It's basically if we had started Part B right away, we'd only be enrolling alpha-4 beta-7 initially, and you'd be six months later when you would be able to start enrolling the IL-23 arm. There's some, you know, some challenges there if you're not actually randomizing between the arms and you're with a different population. This is really, we think, a nice way for us to basically use this first six months of enrollment. Like the IL-23, we've just filed on for regulatory approvals. We'll be dosing with it imminently. We've been dosing the alpha-4 since June, right? This is a way for us to actually, you know, use this time productively, generate this preliminary safety and efficacy data for each of our molecules.
When we get to Part B, we're actually randomizing between the arms, really a nice, well-controlled study, whereas otherwise we'd have wasted the last six months here just waiting for the IL-23.
Excellent. One last one on this, and I want to head to combos. The TL1A data we saw earlier, the phase I data, I recall seeing the free TL1A knockdown, but on a median basis, not on a mean basis. Should I be reading into that?
I think in general, on all three of these mechanisms, I don't really think we're pioneering on any of them, right? All three of these mechanisms, there's something that's already been described here. We're really using the same way that others have measured PD. For alpha-4 beta-7, we use a very similar receptor occupancy assay as vedolizumab. For the TL1A, we used two different PD assays, and we both measured them and then reported them in very similar ways as the first generation. You know, Teva used this free TL1A assay that you just described, reported as, you know, median knockdown in patients. Then both Prometheus and, you know, Pfizer, Roivant at the time measured total TL1A levels and then reported those. We just really said, look, all three of those worked, right?
They used kind of the same PD measures and then reported them in similar ways and used those to select doses for phase II all successfully. I think we just said, let's do it the same way. That's really the short answer.
Fair enough. Heading into combos, I guess the first question that comes up for everyone is, do you, based on what you see internally and the feedback you hear externally, do you see a prominent role for combos going forward, even outside Spyre? I'm kind of alluding to some of the learnings we can get from Duet to inform additional combo studies.
Yeah, I mean, I think we're all excited to see the DUET outcome here in the upcoming months. I think in general, there's a huge amount of evidence supporting that combinations are going to provide some level of additive efficacy, not just from J&J's VEGA study and now the DUET study. They also ran an AFFINITY study in psoriatic arthritis that I think is highly compelling as well. We see all sorts of real-world evidence, animal studies, human genetics, all sorts of things that support the idea that in this heterogeneous disease, blocking multiple pathways is going to lead to improved efficacy. I think the level of robust evidence that combos are going to provide incremental efficacy is there. We're all interested in seeing the DUET outcome because I frankly think it's one of the hardest situations to see a major benefit in.
You're basically testing in a population that almost everyone will have failed either infliximab or adalimumab, one of the, you know, class-leading TNF inhibitors. They're going to be getting a combo that includes golimumab, which is not one of the leading TNFs. I would also say, you know, the TNF class is not one of the leading classes in IBD as well. I think if you see even exciting, you know, interesting signals from DUET, I think it bodes very well for what our combos look like with, I think, better mechanisms and then a population where we're not testing our combos in a population that have all failed one of our mechanisms. These are going to be, you know, 50/50, naïve versus refractory. Our refractory population, we're going to, you know, cap the number of people that have failed our specific components.
I think it's, you know, if we see something even interesting from DUET, certainly if J&J likes what they see, I would be very bullish on what we're doing.
Got it. I recall we had a fair amount of back and forth that day when some of the initial data from TL1A p40 bispecific got posted by Pfizer. I'm curious how you think about that. Were you ever considering doing bispecifics on your own prior to that? Was that even surprising to you for all the inconsistency we saw?
Yeah. We certainly debated it quite a bit internally before we, you know, as we were, you know, starting Spyre, should we make these as bispecifics? We threw it out pretty quickly for alpha-4 beta-7. Two of our components include alpha-4 beta, two of our three combos include alpha-4 beta-7. This is a membrane-bound target that exists in the periphery, whereas our other targets are, you know, cytokines in the inflamed tissue. You know, there is no world in which a single bispecific is binding both of them. If it did, you actually may be worried about the outcome. You would basically be bringing an inflammatory cytokine to an immune cell expressing alpha-4 beta-7. We thought that was either, you know, wasting half your antibody or actually maybe even worse.
It leaves kind of the only other potential combination, which is the TL1A IL-23, as one that you could make as a bispecific. I actually think there are many of them that have been made. I think we're aware of a lot of folks that have made that. I frankly think that the TL1A bispecific class, it wouldn't surprise me if in a few years it ends up looking like what the TNF bispecific class looked like, which is, you know, more than a dozen TNF bispecifics were made. They all got halted in phase I or phase II for one of two reasons. One, they were highly immunogenic. Two, they sometimes had far less target engagement than was expected when they were in the bispecific construct. So far we've seen two TL1A bispecific data sets. Amgen made one, right?
They made a TNF TL1A bispecific, hyperimmunogenic, lost exposure, lost target engagement with that molecule. As you mentioned, we saw the Roche TL1A p40 bispecific, also very high rates of ADAs and neutralizing ADAs as well. Looks like there's a PK impact of that immunogenicity as well. Look, I mean, there's a bunch of them that are made. Maybe one will kind of get through and not have the issues that you would expect with this design, but I think the co-formulation approach is a much lower risk. I also think we're meaningfully ahead with it as well. I think we're ahead, and I think the product profile is likely to be superior.
Got it. If I may, maybe for the purpose of time, I guess I should move on from that one then. Maybe in the last second or so, let's talk about your, let's talk about your other trial where we haven't spent as much time, SKYWAY. Can you talk about what's your timing to market on the arthritis side and the psoriasis side?
Yeah, so we're expecting all three, you know, these are placebo-controlled. In the RA side, a dose-ranging phase II study, we're expecting all three of those readouts next year. This is rheumatoid arthritis, psoriatic arthritis, and axial spondyloarthritis, you know, cumulatively a $30 billion market. Some of the same unmet need issues that we see in IBD, where you have, you know, really incomplete responses to each of the individual mechanisms out there. In RA, almost every mechanism has a black box warning for serious things that TL1A does not have. Either a quarterly or twice-annual dosing profile would be the, you know, best-in-class convenience in those space too. We think, you know, the commercial opportunity is meaningful there. You know, we're going to have these phase II data sets that are placebo-controlled and dose-ranging.
If they're really exciting, we can go to phase III on the back of that. I think it's nicely validating to see Merck and Roche follow us into these indications with, you know, a molecule that doesn't have the same properties as ours and they're a few months behind us. I think those certainly are an underappreciated part of the Spyre value proposition, I think.
Excellent. Any questions from the audience before we move on? Maybe, I guess, if I may, just a last one. Since I realize there's so many iterations possible, I would have thought a Crohn's study could have made sense.
Yeah.
I also would have thought, you never know, but the combo that you're not trying, which is sort of taking your TL1A and combining it with the alpha-4 beta-7, you never know what you can find with something like that.
We're testing that too. A short version is we are planning to run Crohn's as well. We think the likelihood is that, you know, I think all but one agent that has worked in both UC and Crohn's, right? There's high correlation between what works in one or the other. I think the winning program that comes out of the SKYLINE study, I think we certainly are going to be interested in advancing that in Crohn's as well. Kind of what study design and how we get that to market, you know, whether that's a phase II, III, like icotrokinra J&J is doing with icotrokinra, or straight to phase III if we've answered enough with SKYLINE, I think we'll see. I do think alpha-4 TL1A is a really interesting combo too.
I mean, I think we see us and many of the leading sponsors in this space view alpha-4.
I think you meant bispecifically, that would be very different.
Oh, I don't, I mean, I think alpha-4 beta-7 bispecifics are a tough proposition just given where that target is relative to the others. I think that that is not going to be the most efficient way of getting coverage of those two targets together. I just think alpha-4 beta-7 is highly likely to be a good combo component because it is a gut-selective safe mechanism, which is one of the leading products in IBD. I think, you know, we're running a study with it. Lilly is, AbbVie is, Takeda is. I think many of the leading players in this space are using this target. I think we have the best version of an alpha-4 beta-7 with a long-acting antibody.
Makes sense. I know we're at time, so I want to be respectful. Thank you so much. Good luck into next year.
Thanks.
Yeah.