Spyre Therapeutics, Inc. (SYRE) Earnings Call Transcript & Summary

Good morning, ladies and gentlemen, and welcome to the Spyre Therapeutics SPY002 Phase I interim data call. [Operator Instructions] This call is being recorded on Tuesday, June 17, 2025. I would now like to turn the conference over to Eric McIntyre, Vice President of Investor Relations and Finance.

Thank you, Joelle, and thank you all for joining. We're thrilled to be speaking with you this morning. As you saw today, we issued a press release outlining positive interim results for ongoing Phase I healthy volunteer trials of our SPY002 Anti-TL1A molecules as well as some exciting updates to our Phase II development plans. The press release and the slides that we'll be using for today's call are available on the Investors section of our website. I would remind everyone that during this call, we will be making forward-looking statements related to our current expectations and plans for the company and our clinical programs. These statements represent our views as of this call and are subject to risks and uncertainties such as the financial -- the final data, excuse me, from the Phase I clinical trial not being consistent with the interim data presented today, regulatory feedback regarding the company's planned trials and other risks covered in our SEC filings. These forward-looking statements should not be relied upon as representing our views as of any date in the future. So please review carefully. On the call with me today are Dr. Cameron Turtle, our Chief Executive Officer; Dr. Josh Friedman, Senior Vice President of Clinical Development; and Dr. Deanna Nguyen, Senior Vice President of Clinical Development. Following this presentation, we'll have a Q&A session during which Scott Burrows, our Chief Financial Officer; and Dr. Sheldon Sloan, our Chief Medical Officer, will also join. With that, let me turn the call over to Cameron.

And thanks, everyone, for joining the call this morning. We're excited today to announce interim Phase I data from our 2 next-generation Anti-TL1A antibodies, which we believe support their best-in-class potential with favorable safety profiles, complete target engagement at low doses, lack of impactful immunogenicity and PK that enables quarterly or twice annual dosing in maintenance. With these excellent interim data, we plan to advance both molecules into groundbreaking Phase II trials, leveraging capitally efficient platform and basket designs. In our SKYLINE-UC study, we will be exploring the safety and efficacy of SPY002 alongside our 2 additional long-acting monotherapies as well as 3 investigational combination therapies in ulcerative colitis. In our SKYWAY-RD study, we will test SPY072 in a basket of 3 rheumatologic diseases. Together, we believe these trials have the potential to identify multiple products with first-in-class or best-in-class profiles and provide improved options for patients and physicians in multiple disease areas with the potential for significant shareholder value creation. As a reminder, Spyre was launched with the goal of addressing the substantial unmet needs in inflammatory bowel disease by engineering next-generation versions of the biologic therapies in this disease area with leading safety and efficacy. And combining these optimized agents to create combination therapies with the potential for unparalleled efficacy and convenience. In addition, we see substantial opportunity to build upon the best in indication efficacy of Anti-TL1A in IBD by exploring our next-generation molecule in rheumatologic diseases with comparably robust scientific evidence and meaningful commercial prospects. As you'll recall, our discovery campaign identified 2 Anti-TL1A antibodies with excellent preclinical properties. These molecules by novel distinct epitopes on the TL1A trimer with in vitro potency matching or exceeding all first-generation molecules. In addition, they demonstrated half-lives in nonhuman primates that were at least 2 to 3x that of first-generation molecules. Given the excellent preclinical profiles of both molecules, we decided to advance both into parallel Phase I studies to improve our probability of identifying a best-in-class clinical profile. Fortunately, we found 2 such profiles with best-in-class potential in our Phase I studies. As we'll show in our presentation today, both molecules achieved all aspects of our target profile, favorable safety, complete target engagement, lack of impactful immunogenicity and half-lives enabling quarterly or twice annual dosing. With these results, we now turn the page to our Phase II development program, which is designed to efficiently identify best in indication product profiles across some of the largest hard-to-treat inflammatory and I&I diseases. In our SKYLINE-UC study, which began screening patients recently, we will test 3 long-acting monotherapies and 3 investigational combination therapies to identify which provides optimal safety and efficacy. Earlier this year, we disclosed our intent to study Anti-TL1A in rheumatoid arthritis. Today, given our conviction in owning potential best-in-class Anti-TL1A antibodies, we're pleased to announce a broadening of our efforts in rheumatology. In the SKYWAY-RD study, we will explore SPY072 in 3 rheumatologic diseases in which TL1A's role is strongly validated and a meaningful commercial opportunity exists for a novel mechanism with an indication leading product profile. To maximize the long-term value of our Anti-TL1A franchise with both strategic and commercial optionality, we plan to advance one of our molecules in our UC study and the other in our rheumatology study. For clarity, the molecule advancing in UC will retain the SPY002 nomenclature while we will utilize SPY072 for the molecule advancing in rheumatology. These 2 groundbreaking studies utilize innovative platform and basket designs to generate high-quality proof-of-concept data in a capitally efficient manner. In both studies, shared sites and investigators are utilized across multiple interventions or indications to maximize efficiency, providing an estimated 35% to 40% cost savings relative to running separate studies for each combination or indication. These efficient designs allow us to provide all proof-of-concept readouts from the trials with our existing cash balance and no change to our runway guidance. Specifically, these trials are expected to provide 6 Phase II proof-of-concept readouts next year; 3 monotherapy readout in ulcerative colitis plus 3 rheumatology indications for TL1A. In 2027, we plan to follow up with the full readout of placebo-controlled monotherapy and combination therapy results from the SKYLINE-UC study. Together, we believe this series of Phase II readouts provides one of the most compelling catalyst maps in our space over the next 2 years. With 9 readouts in markets totaling over $60 billion of annual revenue, we believe this development plan provides multiple paths to substantial value creation. Specifically, we believe each of our readouts as relevant precedent in terms of value that has been created for shareholders upon positive proof-of-concept results. First, our 3 planned open-label monotherapy readouts in UC each evaluating derisked targets with the potential for substantially improved convenience were precedented by Morphic's Phase IIa readout that led to its $3 billion acquisition last year. Second, our next-generation TL1A readouts across IBD and rheumatology have the potential to identify a pipeline and a product opportunity with a molecule with meaningfully improved properties compared to first-generation molecules that were the focus of multiple multibillion-dollar acquisitions and partnerships. Lastly, our investigational combination therapies in IBD have an almost unprecedented opportunity to deliver product profiles without competitive equals in development. If these are successful, advances in distinct therapeutic areas such as cystic fibrosis could be seen as precedent where combination therapies dramatically change the treatment paradigm and generate substantial value for investors. I'll now turn the call over to Dr. Josh Friedman to walk through the Phase I results for SPY002 and SPY072.

Thanks, Cameron. Over the next few minutes, I will walk us through the interim safety, PK and PD data from our first-in-human studies of SPY002 and SPY072. As you'll see, these interim data indicate that both have the potential to be best-in-class TL1A inhibitors. Let's begin with the trial design, which is identical for the 2 molecules with key endpoints of safety, PK, PD and ADA. The trials comprise 5 placebo-controlled single-dose cohorts of 8 healthy volunteers with doses escalating from 100 milligrams to 300 milligrams subcu and then from 300 milligrams IV to 1,500 milligrams IV. The approximate follow-up period available for this interim analysis range from 4 to 20 weeks. Dosing in the SPY002 study was ahead of SPY072, resulting in slightly shorter follow-up in all cohorts. As shown in this blinded analysis, both molecules were well tolerated with low rates of adverse events and no serious adverse events. There were 2 participants with treatment-related AEs, headache and migraine, respectively, in the SPY002 study and 3 participants with treatment-related AEs of gastroenteritis, rash and abdominal pain in one participant and chest tightness in 2 participants. The chest tightness was grade one, transient and resolved without intervention in both participants. There were no events with the severity above Grade 2 in either study. The next key endpoint is PK. Drug exposure curves for both molecules are shown here and demonstrate dose proportional exposures across the range of doses. Both molecules exhibit substantial half-life extension with an estimated half-life of approximately 75 days for SPY002. For SPY072, we will analyze additional time points to provide a preliminary half-life estimate. Based on these interim data, we have high confidence that these PK profiles support quarterly or twice annual dosing. There is no apparent effect of ADA on the PK of either antibody. For direct comparison between our investigational molecules, we've overlaid the 2 sets of curves with IV cohorts on the left and SC cohorts on the rig. SPY002 is shown in gray and SPY072 is shown in blue. As you can see, the 2 molecules exhibit comparable PK by both routes of administration across dose levels. The only separation observed between the molecules was at later time points in our 100-milligram cohort, where concentrations of the molecules are below the clinically relevant range we plan to investigate in our Phase II studies. Of course, a drug is effective only in so far as if it hit its target. We developed 2 PD assays for the trials. The first measures the amount of free TL1A available to bind to its receptor. We found that median free TL1A levels are reduced below the lower limit of detection in all cohorts through up to 20 weeks of follow-up at the lowest dose tested. The second PD assay measures total TL1A, which is expected to accumulate because the antibody target complex is cleared more slowly than the free target. The results match this expectation with over 100-fold increases in total TL1A at the highest doses. This is sustained throughout the observation period with even the lowest dose cohort demonstrating substantial increases out to week 20. Overall, the PD results indicate that both molecules are able to suppress active TL1A for an extended period after a single dose. In assessing the potential impact of half-life extension for our molecules, it's helpful to compare with first-generation Anti-TL1A antibodies, which have demonstrated excellent efficacy in Phase II IBD trials. The clinical remission rates generated by Afimkibart in its Phase II UC trial during the induction of maintenance periods are shown on the left. Using published PK properties of Afimkibart and our Phase I results, we simulated exposures of SPY002 with quarterly and twice annual maintenance dosing compared to the steady state levels of the Afimkibart. As you can see in these simulations, we are able to exceed trough concentrations of the 450 milligram Afimkibart dose with quarterly maintenance dosing, and we're able to exceed trough concentrations of the 150 milligram Afimkibart dose with twice annual maintenance dosing. Given the comparable potency of our molecules to a Afimkibart in vitro and with comparable or greater PD effects in vivo, we believe that matching or exceeding Afimkibart's exposures in patients is likely to lead to similar or greater efficacy. Let's recap the interim Phase I results. SPY002 and SPY072 were well tolerated with all adverse events being mild and typical of a Phase I study. Both antibodies exhibit a prolonged half-life with SPY002 exhibiting a half-life of approximately 75 days more than threefold that of first-generation Anti-TL1As. Single doses of SPY002 or SPY072 resulted in complete suppression of TL1A through up to 20 weeks of follow-up, supporting the potential for quarterly or twice annual dosing. There was no apparent impact of ADAs on PK or PD. Given these favorable interim results, we plan to advance SPY002 into our SKYLINE-UC study and SPY072 into our SKYWAY-RD study. I'm pleased to hand things over to Dr. Deanna Nguyen to speak about our recently initiated SKYLINE study.

Thanks, Josh. I'm very excited to announce the recent launch of our SKYLINE-UC platform study, which is already ongoing in multiple geographies and patients are currently in screening for our SPY001 cohort. With today's SPY002 Phase 1 announcement, we will be submitting the SPY002 data package to regulatory authorities soon as the next investigational intervention for this study, adding to the actively enrolling SPY001 cohort. The study design for the SKYLINE-UC trial is shown here on Slide 22. As we have previously described, this is a platform study with an induction, maintenance, treatment design, investigating each of our 3 monotherapies and 3 pairwise combinations. This study will be conducted in patients with moderately to severely active UC in 2 parts. Part A is an open-label evaluation of our monotherapies each being added to the study after its interim Phase I results. The objective of Part A is to demonstrate early proof of concept for our 3 monotherapies based on a primary end point of a change from baseline in the Robarts Histopathology Index or RHI, week 12. RHI is an objective endpoint that will be assessed by central readers who're blinded to whether the images were taken before or after treatment initiation. We will also be assessing important UC endpoints such as clinical remission and endoscopic improvement. Part B is placebo-controlled and evaluates our combinations and investigational monotherapies, seamlessly enrolling after Part A finishes enrollment. Part B has multiple objectives including demonstration of proof of concept, contribution of components for combinations as well as exploration of dose ranging for each of our monotherapies. The primary endpoint here will be clinical remission at week 12. The study design was developed with the goal of identifying the optimal product or products to advance to pivotal studies. Given the shared placebo and comparator arm, the SKYLINE-UC study is designed to be efficient, requiring 40% fewer patients and hence, less cost compared to conducting individual studies for all 6 programs. Another advantage of this design is that it's patient-centric and with an open label Part A and a low placebo allocation rate in Part B. The study is also attractive to sites with simple unified dosing between placebo, monotherapy and combination arms, reducing operational complexity and allowing for a blinded study. So overall, we feel this is a landmark study, and investigators share this enthusiasm given the wrong -- the strong demand we have seen for study participation from prospective site. In fact, we've already selected roughly 300 sites globally. Our development approach is uniquely enabled by our portfolio of engineered antibodies with half-life extension. The single Phase II study provides 6 shots on goal to deliver a best in indication product profile with superior efficacy and best in indication quarterly or twice annual dosing that could transform the treatment paradigm in IBD. We believe that our approach of co-formulation of monoclonal antibodies provides the highest profitability of achieving an optimal combination product profile. Critically, we believe that co-administration or co-formulation is unlikely to worsen immunogenicity to our antibodies, and they even improve it. And to this point, the aforementioned data study in which the antibodies were co-administered, it was reported that combination treatment led to a lower incidence of ADAs compared to the anti-TNF golimumab alone. This finding is in contrast to bispecific format which often leads to increased immunogenicity compared to the monoclonal components of the combination. On this Slide 25, we show a summary of several bispecifics against TNF or TL1A, which have exhibited high rates of immunogenicity and the majority of these have since been discontinued. For example, the 21A TNF bispecific form large immune complexes, leading 0to high rates of ADA with significant reductions in drug exposures in its Phase I study. To close, Slide 26 provides the readout timing for the SKYLINE-UC trial. As mentioned, patients are already in screening, and we expect dosing to occur in the coming weeks. In 2026, we plan to share open-label results from Part A of the study, followed by the placebo-controlled readout of our investigational monotherapies and combinations in 2027. I will now turn it back over to Dr. Josh Friedman to discuss our other exciting study, the SKYWAY-RD trial.

Thank you, Deanna. I'm thrilled to share our Phase II development plans in rheumatology, evaluating SPY072 in the SKYWAY-RD basket trial. Despite advances in treatment, patients with rheumatologic diseases, such as RA, axSpA and PsA need new, more effective and durable therapies. Patients have a limited number of options and have to cycle through them as they experience primary or secondary treatment failure. A range of evidence points to TL1A as a key target to reduce joint inflammation and injury in these diseases. This includes genetic links, human translational measures and animal model confirmation. The SKYWAY-RD trial will utilize a basket design to efficiently test the efficacy and safety of TL1A inhibition by SPY072 and in RA, axSpA and PsA, all under a single protocol. SKYWAY-RD comprises 3 sub-studies as illustrated here. The RA substudy will include participants with active disease and inadequate responses to conventional synthetic DMARDs, biologic DMARDs or targeted synthetic DMARDs. The active treatment obviously include high and low doses of SPY072 to assess dose response and exposure response over a range of exposures. The primary endpoint is the change in DAS28-CRP from baseline to week 12. The PsA and axSpA substudies have analogous designs with eligibility based on active disease and an adequate response to conventional or advanced therapies and key endpoints of ACR20 and change in axSpA disease activity score, respectively, at week 16. In designing SKYWAY-RD, we recognized an opportunity to efficiently establish proof of concept across 3 diseases in a single study. We estimate potential savings of 35% relative to 3 separate studies. Enrollment projections support the ability of approximately 50 sites to enroll all 3 indications, compared to contracting with 100 sites as separate studies. This illustrates the synergy gained by partnering with rheumatologists to see patients with each of these 3 diseases and utilizing a shared study infrastructure. Let's take a look at some of the evidence that motivates us to test TL1A inhibition in these rheumatologic diseases. In each disease, TL1A gene and/or protein expression is elevated as shown here in a few of the published examples. This places TL1A at the scene of the crime and the biology of TL1A supports it being a culprit in disease through activation of T cells driving joint inflammation. Furthermore, fibroblast-like synoviocytes are stimulated by TL1A, suggesting that TL1A may promote both inflammation and joint injury in these diseases. In addition, these rheumatologic diseases share inflammatory drivers with IBD, such as Th17 and TNF pathways. So the proof of concept for TL1A in IBD provides further support for these rheumatologic diseases. To add to this evidence, we tested TL1A inhibition in the rat collagen-induced arthritis model, using either a semi preventative dosing schedule or a therapeutic schedule that is after the onset of symptoms. As shown here, as assessed by arthritis score, TL1A inhibition exceeded the efficacy of anti-TNF in the first model and is comparable to the anti-TNF in the second. Selectively, these preclinical results tell us that SPY072 has the potential to be a first-in-class mechanism for the treatment of these rheumatologic conditions with a significant improvement in convenience and potentially efficacy relative to today's standard of care. If successful, we will be excited to offer patients a new MOA with infrequent dosing that allows them freedom from the daily, weekly or monthly reminders of their chronic condition. SKYWAY-RD trial will provide multiple proof-of-concept readouts in addition to those we reviewed for SKYLINE-UC. We're on track to initiate this study in Q3 of this year and expect to read out top line results for all 3 indications in 2026. I'll now turn it back over to Cameron for a closing summary and Q&A.

Thanks, Josh, for outlining the innovative SKYWAY-RD trial and what we believe is a unique approach to efficiently discover the most attractive indications for our next-generation TL1A antibody. We believe this trial has the potential to unveil a broad set of first-in-class applications for this promising molecule. In closing, I'd like to remind our listeners what to expect from us over the next couple of years. Later this year, we will provide an interim Phase I assessment of SPY003, our next-generation IL-23 antibody. If successful, SPY003 will then be added as the final component of our SKYLINE-UC study unlocking all 3 of our long-acting combinations. Following this final Phase I readout, we will turn our attention towards a series of catalysts that could identify indication leading product profiles across large inflammatory indications. In 2026, we anticipate open-label readouts for SPY001, 002 and 003 in ulcerative colitis and placebo-controlled readouts for SPY072 in RA, PsA and axSpA. In 2027, we expect to follow up with placebo-controlled monotherapy and combination results in UC. With our cash runway extending into the second half of 2028, we are well funded to execute both of these innovative, efficient studies and deliver impactful results for the patients, physicians and shareholders who serve. Thank you for your attention this morning, and I'll now turn the call back to the operator to begin the Q&A.

[Operator Instructions] Your first question comes from Alex Thompson with Stifel.

Congrats on the data. I guess 2 from us. I think the first one would be, what's your confidence that 002 is the right antibody here for UC and Crohn's and 072 is the right antibody for rheumatologic diseases? And then on the SKYLINE platform UC study, can you talk a little bit about Part B and how you plan to actually conduct dose ranging and the rationale there?

Yes. Alex, and I'm happy to start and pulling others from the team. So the first question in terms of how we selected 002 versus 072 for the 2 distinct disease areas, is really based on the data that we've gathered today and shared, we think either molecule would be an effective therapy in both indications. And frankly, we think they have comparably excellent profiles across all the parameters that we showed and could likely use either of the molecule for both indications. What ultimately led us to choose to take 002 forward in IBD and 072 into the rheumatology study is actually related to the manufacturing of both molecules and in particular, our ability to co-formulate these molecules with our other long-acting antibodies. And in the SKYLINE study, as you know, will be co-administering these molecules. But regardless of what the winner is, we want to move forward with a coformulated version of these antibodies for our pivotal studies. And with SPY002, we're able to achieve a good viscosities, coformulations of this Anti-TL1A with both of our alpha-4 beta-7 and our IL-23 antibody, and we think that makes it the optimal product to take forward in the IBD study. Regarding the dose ranging in Part B of SKYLINE. So the intent here is we're taking forward 2 doses of each of our monotherapies, and we'll just be testing a single dose of our combination therapies in this study. And the overall rationale for this is that we know that for approval here, we expect to have to dose range both of our monotherapies in combination. But in this way, we're able to solve the monotherapy dose ranging in this study. And then we would only intend to solve combination dose ranging for whichever winner comes out of the study, and we may not need to do dose ranging for all 3. Hopefully that answers your questions.

Your next question comes from Akash Tewari with Jefferies.

So I got a few questions on my end. First, how do you plan to show contribution of component in your Phase II basket studies? Are they powered to show that in your trials? And what's the bar for combo in the biologically naive and then also the experienced patients that you expect to show there? Also, can you talk about the mix of patients you anticipate in those trials and how that might change depending on some of the data we might get from J&J throughout this year? And then finally, as of now, you're not pursuing a combination strategy in some of the derm indications. Is that based on a lack of conviction in the combo approach there? Or is there perhaps limitations on where you can take your IL-23?

Good sets of questions all around, Akash, I'll see if I can go through these individually. So first, regarding for contribution of components in the SKYLINE study, that is the explicit design of the study, which is to initially test each of our active arms against placebo. But then if the monotherapies and combinations are successful against placebo, then we will test the combination for superiority to their monotherapy components at those high doses. So that's the explicit aim of the study to achieve that. The bar for success, I think, obviously, we have kind of both the statistically significant difference is something that we would be interested in seeing and then also a clinically meaningful difference, which as you always know, is a gray area among physicians, but we typically see if we're seeing a double-digit change in clinical remission between a drug and its comparators, I think that's typically been seen as clinically meaningful in this indication. So that's what we think would be an exciting difference if we saw that in our population as well. The second question regarding the mix of patients in the study. Our target here is a mixed population. So approximately a 50-50 split of patients who are naive to advanced therapies versus those who are refractory to advanced therapies. And I think you did highlight something that readout that we will be watching closely in the upcoming months, which is J&J's DUET studies where we're now seeing the combinations against the monotherapies in all refractory population, where as we've previously known that you get a very large delta in a naive population. So it is possible that we could adjust our strategy depending on what's seen in that result. And then your last question in rheumatology and whether we would be interested in testing combinations there as well. We don't yet have a VEGA like study in these indications that demonstrates that combos can deliver additive efficacy relative to their monotherapy components. However, I think from the basic biology of these indications, we think it's quite likely that other hard-to-treat I&I diseases like the ones that we're testing in our basket study, we could imagine the same type of effect where hitting multiple pathways leads to much greater efficacy. And I think it is something that we could pursue in the long term. However, what we don't yet have in those indications is the proof of concept that TL1A works as a monotherapy, which, of course, we do have in IBD. So that is really the key difference for us in terms of our conviction for going straight after combos in IBD versus room. We want to first establish that TL1A is effective on its own, and then we may decide to pursue combinations down the line.

Your next question comes from Yatin Suneja with Guggenheim.

Congrats on the data. So 2 for me. So first one is on the immunogenicity. Could you just comment on what you're seeing there? Do you need to wait a little bit longer for some of these immunogenicity to appear? So that's one. And then you touched on the TNF or the J&J study that is coming up. Could you just also comment on the safety side because one of the components at J&J using is the TNF that tends to have a black box warning, which most likely you will not have. So could you just elaborate on that differentiation that you might have?

So maybe Josh, I'll ask you to comment first on the immunogenicity data that we've gathered to date and whether we would expect to see something if there was an issue at this point?

Yes. Happy to address that. The ADA analysis is ongoing, and that's a measure that really does take time to properly assess over prolonged exposures. But I can say that given the data that we've seen so far, we are confident that there's no apparent effect on PK or PD, which is really what matters, of course, in the end. So the -- although it's not appropriate to put a number on ADA at this point, where it really matters in terms of exposures and effects, we are confident that we have good molecules.

Thanks, Josh. And I'll just add that for other molecules, including -- and a bispecifics in this space and others, the time points that we've shared already would have been sufficient to identify issues with some of those other molecules in terms of either seeing that more rapid clearance or loss of target engagement. And so though, of course, this will continue and we'll look at longer periods for our EDA analysis, we certainly feel that our confidence is high already that we don't have a substantial issue. To your second question regarding kind of the safety of combinations in the long run, I think in general, though, we really are standing on the shoulders of J&J here who is really pioneering combinations in this space and others. I think we think it's quite unlikely that the TNF containing combo that they have is likely to be the long-term best combination in IBD or other indications. And I think it's really for 2 reasons. One is the safety that you highlighted. The safety of the TNF class is well described and the black box warning there, we would expect to translate to combinations. And of course, when you add an additional immunosuppressant on top I think that is one of the main concerns is kind of will you see additive safety concerns when you do 2 things together when one of them already has safety issues. What we're happy with our portfolio here, and we obviously selected these targets for a reason, that none of our 3 target have black box warnings or the substantial monitoring requirements that we see from other classes. I think Alpha-4-beta-7 and IL-23 have very well-characterized long-term safety profiles that are clean, whereas the TL1A class is newer, we certainly don't yet have large Phase III data sets or commercial data sets, but so far, it looks very good as well. And besides safety, I also think it's unlikely that TNF based combinations are going to be the most efficacious combinations as well. We have one direct head-to-head study in this space in the VARSITY study that shows that Alpha-4-beta-7 beat TNF head-to-head in ulcerative colitis patients. And though there aren't head-to-head studies, I think anyone comparing the TL1A results to TNF results in either ulcerative colitis or Crohn's would conclude that the TL1A appears to be superior on efficacy as well. So though we're learning a lot from J&J, and we think they're kind of moving this field forward dramatically, I think it's quite likely that our combinations have -- could be superior on both the safety and efficacy aspects.

Your next question comes from Umer Raffat with Evercore ISI.

Maybe a couple of quick ones, if I may. First, the duration of follow-up you guys used was 20 weeks on 100 milligrams, and it was lesser than that 8 to 16 weeks on the higher doses. So could you remind us what was the half-life for that 100-milligram subcu dose? Was it higher than 75 days, for example, at least that's kind of we're sort of seeming to calculate? Number two, for the dose selection for Phase II, I was quite intrigued by the graph you guys showed on Slide 19. But there's one thing that confuses me, which is Merck Prometheus data does appear to suggest a dose response, Roche data doesn't. I know you're comping it versus Roche. But as you think about picking a C trough dose as well as, obviously, the C trough with the duration. Are you really just limiting the scope to Roche? Or are you thinking about the possibility of a dose response on the Merck side too? I guess what I'm really saying is, do you need a higher dose for every 6 months? And then finally, could you just remind us sort of where the ADA rates are tracking both on a single dose as well as on a multidose basis?

Yes. Thanks, Umer, I think I got all of those. So in terms of how we calculate half-life, just when we put out a single number as we have for the 002 molecule, that is usually based on a population PK model that incorporates data from all of the cohorts. I think sometimes we can share data on the individual cohorts as well and kind of show the half lives. And I think if you calculate some of them as it sounds like you've done, you do end up with even longer half-life estimates than we've seen. However, I think what we care most about is using kind of that cumulative model that incorporates data from all of the data that we have, which we would rather not report a single number until we have that kind of within a pretty narrow band, which is why we think we have it for the 002 molecule, and we'll wait for kind of another data set or 2 for 072. On the second one, which I think is actually a very good question and one that we spent quite a bit of time on in terms of how to select our dosing and which of the first-generation molecules we want to use to pick our dose selection based on. And as you know, our molecules have pretty different properties than all 3 of the first generation ones, which makes that the most challenging in terms of selecting which one to compare against. In terms of using Afimkibart, we think that is the most relevant first-generation molecule for us to compare against because the potency of our molecules is in the same ballpark as Afimkibart even though our half-life is meaningfully different. Compared to Tulisokibart, the Merck molecule, we think we have a pretty substantial improvement in potency relative to that molecule. And so we may not even need the same amount of trough concentrations to be able to achieve the same level of target engagement compared to that one. And I think your point is well taken that in the Roche study, we see a pretty mild dose response depending on which endpoints you look at in this study, and that's why we think it's reasonable to target either slightly above the 450-milligram dose with our quarterly schedule or slightly below with the Q 6 months. We think that's reasonable based on their dose response data. We've certainly seen the dose response that was shown with Tulisokibart in the Merck study and then saw them increase the dosing frequency in the Phase III study as well. However, we think that actually could be due to the fact that you get less target engagement with the lower potency molecule there and we may not need that level of drug to be able to get the full target engagement that we think we need in the maintenance setting. So that is the reason, that is why we feel pretty comfortable with both the quarterly and twice annual dosing profiles that we have shown the slide there. And then lastly in terms of the ADA rates, I'll kind of echo what Josh said earlier, which is that, of course, we continue to measure ADAs over time, and it's not something that and you pick any set of weeks and you're done in terms of the ADA analysis, I think we're encouraged by what we see so far. And when we look at the totality of our ADA rates, they're actually very low. They're lower than anything we've seen from the first-generation molecules. But we don't think it's necessarily fair to kind of report that just yet, given that we don't have long-term exposures from all of our doses.

Cameron, if I could just add on the question of the Afimkibart. One other motivation was that, we wanted to make an interpretable and fair comparison and Afimkibart, we just happen to have publicly available pharmacologic data on Afimkibart that enables us to do the most meaningful simulations.

Good point, Josh.

Your next question comes from Faisal with Leerink Partners.

Just wanted to ask on a little more specificity here. So on the upcoming J&J DUET readout for their TNF IL-23 co-formulation in the treatment experienced IBD patients. Can I kind of push you to clarify what would you like to see to validate your strategy like on a read-through basis? And then can you also just reiterate for us why you believe you're well positioned versus that combination?

Of course, and thanks Faisal. So just as a background here for those who haven't followed the trials closely, so the DUET studies are 2 large studies looking in ulcerative colitis and Crohn's patients after the same combination that was tested in the VEGA study, which was done in naive ulcerative colitis patients. So in this case, we're now seeing the refractory population of both UC and Crohn's to add to our understanding, which we've seen from VEGA. And what we are interested in and what we believe is likely to come following the DUET studies would be a pivotal study or program that is exploring the combination in a mixed population. I think almost every approval in this space has looked at a Phase III in a population of patients who are both naive and refractory approximately a 50-50 mix in those Phase III studies. And so what we think the right bar is to when we look at DUET to see that's the result in the refractory population. We already know the results from the VEGA population. And we would expect that their Phase III looks like a blend of those 2 and their ultimate label, we expect to be a blend of the VEGA and DUET results. And I think ultimately, I think it's actually the same bar that I mentioned previously that we think would be meaningful on that Phase III and commercial label, which is if they show a double-digit difference in terms of the clinical remission rate from combinations relative to monotherapies, I think we've seen that the practice changing in this space multiple times in both the VARSITY study showing not even a 10% delta between Alpha-4-beta-7 and TNF and still led to a meaningful shift towards ENTYVIO or even more recently with the P19 inhibitors showing approximately that level of delta versus Stelara also led to a meaningful practice shift. So that is kind of, I think, a reasonable efficacy bar to look for. It's that about a 10 percentage shift in that mixed population between VEGA and DUET. And then, of course, that's kind of not thinking about the efficacy side -- or sorry, the safety side as well, which I think that we covered that point previously in the prior question. And again, just reiterating why we think our combinations are likely to outperform here. I think Alpha-4-beta-7, one of our combination components has beaten TNF head-to-head in both safety and efficacy in this space. And TL1A, though we haven't yet seen a head-to-head study comparing TL1A to TNF; again, anyone looking at the TL1A studies relative to the TNF studies would conclude that TL1A looks to be superior as well from the efficacy perspective. None of that is even including the fact that our combinations are likely to be quarterly, if not twice annual combos, whereas the J&J combo will be dosed on a monthly basis.

Your next question comes from Sam Slutsky with LifeSci Capital.

A couple of quick ones for me. Any clear differences on the monomer versus trimer binding between 002 and 072? And then what's kind of the current view on this topic? And then also, just apologies if I missed this, but will Part B of the ulcerative colitis study use every 3 or 6 months dose? And then what's the route of administration for both induction and maintenance?

Yes. Thanks, Sam, for the questions. So for number one, on the monomer versus trimer discussion, I think the -- perhaps the most relevant assay to look at is the total TL1A assay, where -- this was an assay that was shown by both Pfizer and Prometheus originally exploring kind of the binding of antibodies that bind to either the trimeric or the monomeric form of TL1A, and we see that -- and if those 2 binding modes lead to a different change in the amount of TL1A that is stabilized in solution. And we see that one of our molecules, 072 performs very similarly to Prometheus in terms of the total change. And so we think it has a binding mode more like that. Whereas 002, the molecule, it actually has kind of changes in total TL1A that are somewhere between the Pfizer and Prometheus antibody. So they have some aspect of monomer binding and trimer binding for that one. All that said, we do not think it makes a meaningful difference in terms of either target engagement or efficacy here. When we look at the Phase II data across the first-generation molecules in the induction setting, we don't see a meaningful difference depending on the epitope or binding mode of the molecules. And so ultimately, this didn't lead to a difference in our perception between the molecules and of course, likely the decision that we think they were both excellent based on the properties we've seen. In terms of the dosing profile in the SKYLINE study, we're not sharing the full dosing details just for competitive reasons on all of our molecules, but I think we do believe that the half-life that we've shown here will enable quarterly or twice annual dosing and we intend to explore that. We will dose all of them in the induction setting intravenously initially for 2 doses. And that's really driven by our third molecule, even though we don't yet have the Phase I data for that molecule, we anticipate dosing that molecule somewhat like AbbVie SKYRIZI or risankizumab, which does require quite a heavy induction load to get maximal efficacy. And so that's why to maintain the blind in this study, we're testing all of them on an IV loading basis. So if the IL-23 does not move forward, we would anticipate the potential to take forward an all subcutaneous version of either our alpha 4 or our TL1A molecules.

Your next question comes from Tyler Van Buren with TD Cowen.

This is [indiscernible] on for Tyler. I had 2 quick questions here. First, what would be required for monotherapy proof of concept? And then the second one on ADAs, could you elaborate on why the risk is higher for bispecifics?

Yes. Thanks for the question. So in terms of the monotherapy proof of concept, I think we would -- we view kind of the 2 different readouts from the SKYLINE study as sequential or incremental proof of -- that these molecules have either similar or improved efficacy compared to the first-generation molecules. So we think the open-label data that will be smaller and without a comparator, we think it actually will be helpful in terms of demonstrating that we have active versions of each of these molecules that can be dosed on a far more convenient basis than the first-generation molecules. And then in the placebo-controlled data, we think that, that will be the more definitive answer in terms of the numerical comparison compared to the first-generation molecules. And then in terms of the immunogenicity risk between monoclonals and multispecifics, there's really, I think, 2 aspects of this. The first is the ability of antibodies to form large complexes and that's possible to be done with a monoclonal, particularly against some trimeric targets, like TL1A or TNF. We can see the cross-linking of the target with antibodies that lead to immunogenicity and can lead to the more rapid clearance of antibodies. As you can probably imagine, if you have a multi-specific antibody that can bind to multiple different types of targets with either trimeric or other forms, you can imagine the range of complexes that could form is even larger, and that's been observed with multiple bispecifics. The other risk that is somewhat distinct from that immune complex risk is that some of these multi-specific constructions use linkers between the native antibody fragments. And those can be more immunogenic than the native antibody structures themselves. So those are kind of 2 different aspects that can make bispecifics more immunogenic than monoclonals or the combinations of monoclonals.

Your next question comes from Paul Choi with Goldman Sachs.

I just had a question on the -- first question on the dosing -- range for the subcutaneous dosing and you put your cap at 300. Was that primarily capped at 300, given the co-formulation and manufacturing considerations you listed earlier? Or was there any view that you could go higher in a subcu formulation for induction in the RA or rheumatoid setting, whereas IV formulations are used a little more frequently in GI disease? And then I have a follow-up question.

Yes, sure. So the short version is no. We're actually not capped at 300 milligrams. For both the monotherapy versions of our TL1As and coformulations of TL1As with other antibodies, we can get up to about 200 mg/mL which is really at the very high end of what conventional formulations are able to achieve. And we're able to do that at viscosities that would enable using standard autoinjector for Phase III and commercialization. So really strong product profiles, and we would expect to be able to use a 2 mL approximately autoinjector. So that's about 400 milligrams in a single 2 mL injection. So that would be our expectation for either the monotherapy of TL1A or co-formulations of the TL1A with our other antibodies.

Would you have any thought as to a potentially higher dose for subcu induction in the SKYLINE study?

Yes. So for the IV, as you point out, we really are not limited by kind of the amount that we want to -- are able to dose in a single injection. So with IV, we can really ensure that we get exactly the right amount of drug on board quickly. And as I think we've seen from many agents in IBD, the amount of drug that's required on board during the acute flare of the induction period is -- can be substantially higher than is required once remission is achieved and patients are in maintenance. So we do think that kind of IV loading lets us have some flexibility in terms of how much we dose. But given the potency and half-life of the molecules we have here, it's still relatively low dosing that we could transition to an all subcu version for either of our Alpha-4 or our TL1A.

Okay. Great. And then just one trial operational question, which is at least in the -- regarding the induction portion for Part B of the SKYLINE study. Is that something you can initiate roughly in parallel with the Part A monotherapy or will that be only initiated after the maintenance portion is completed for the induction portion of Part B?

That's a very good question. And really the power of the platform study here is that the intent is that we get, as Deanna mentioned, approximately 300 sites going in this study over the next few quarters here. And then those sites, we hope to have them continue be enrolling continuously really for the next 1.5 years or so to fill up this study. And so what that means is kind of seamless transition of enrolling the Part A cohorts to the Part B cohort of the study. So that we never have to stop and restart again. That's our goal overall here. And so that doesn't mean kind of waiting even for the 12-week induction data from Part A before we start enrolling Part B, we expect this to be seamless.

Your next question comes from Julian Harrison with BTIG.

Congratulations on all the recent progress. On SKYLINE-UC, I'm wondering if you could talk about the steroid protocol, what's allowed and what's not? And then in SKYWAY-RD, will you be enrolling any TNF experience or refractory patients across these 3 diseases? Curious how you think of 072's activity in those subsets?

Yes. Thanks, Julian, good questions. Maybe I'll let Deanna, you want to comment first on the steroid protocol for the SKYLINE study?

Sure. We are not disclosing the details of the study design, but what we can say is that the handling of concurrent steroids in terms of low doses and tapering is pretty similar to a typical UC study.

Okay. And then on -- Josh, do you want to comment on the SKYWAY study in terms of the patient populations that we're enrolling there and specifically the TNF refractory population?

Yes. So I'm happy to. We intend to enroll a mixed population of both tsDMARD failures who haven't been previously exposed to advanced therapies. And patients previously exposed to advanced therapies, biologic and tsDMARDs. And that second group would include anti-TNF failures.

Your next question comes from Andy Chen with Wolfe Research.

Maybe another question on immunogenicity. I know you've already answered many questions on that topic already. But -- so we understand why immunogenicity is better than bispecifics and you see no ADA impact in your monotherapy setting. But in the co-formulated setting, I'm just curious if you have additional evidence that your co-formulated antibodies won't have incremental immunogenicity on top -- like more than -- basically more than a single therapy. Do you have maybe computational or in vitro predictions saying that each of your 3 different combos is basically similar to the singles?

Thanks, Andy. It's an interesting question. I think we talked a little bit previously about the philosophical region or theoretical reason why we anticipate or kind of we've seen in the past that bispecifics have higher immunogenicity for both kind of the range of complexes that are able to form when you're binding multiple targets together, but then also kind of the linkers that we mentioned, the non-native linkers that can lead to immunogenicity. So those are the 2 theoretical reasons why bispecifics would be more immunogenic. In terms of the demonstration that ours are not, I think we, as you know, haven't dosed these yet in humans, but we actually have co-administered these antibodies in both pharmacologic assays. So testing the combination efficacy compared to monotherapy efficacy in a variety of animal models. And then we've also conducted combination tox studies, which also look at pharmacokinetics as well. And in all of those studies, we haven't seen any evidence that either we see more rapid clearance or we see kind of toxicity associated with immunogenicity as well, nor do we see kind of the formation of complexes that lead to the increased clearance of either antibody. In fact, they perform almost identically when they are administered individually or co-administered in nonhuman primates. Of course, we'll have to see that, that translates to humans as well, but because both the biology suggests that they're likely to be fine, the VEGA study actually suggests an improvement and we haven't yet seen any evidence of issues when we co-administered, we feel pretty confident that this is likely how they'll perform going forward.

Your next question comes from Colleen Kusy with Baird.

Congrats on the data. As we think about the pivotal studies in UC, is there an upper limit of how many assets you bring forward into pivotal study? And what will that study be based on? And then for the SKYWAY study, across the 3 indications, can you speak to a level of preclinical data supporting each whether you view any of these as lower risk relative to others and just confirm that you view yourself as first-in-class across those 3 sub-studies?

Colleen, I'll take the first one and then Josh, maybe I'll let you comment on the evidence for each of the 3 indications in SKYWAY. But in terms of for the pivotal choice, I think really, it's one of the most interesting questions that we'll be working on over the next couple of years, it is kind of what's the product profile to take one or multiple winners forward? I think we believe there's a reasonable probability that each of our combination has the potential to demonstrate additive efficacy, no incremental safety issues and be doseable on a quarterly or twice annual basis, which we would think -- any product that shows that profile, we would think would be -- looks like a best in indication profile for these disease areas, which are a $30 billion market in downstream. So we think there could be deficient rationale to justify investments of bringing more than one forward. But of course, we need to see the profiles and verify that before we would decide how many to take forward. And Josh, I'll let you comment on the data for the 3 indications that we're testing in SKYWAY.

Yes, it's a really interesting question. And I just touched on some of the data that's out there and then some that we've generated ourselves. And I have to say, I can't take a favorite. I think that in evaluating the genetic translational and animal model data, for me, they're all over the threshold to make it very clear that the real thing to do is to do a clinical trial and assess which is best. The preclinical model of inflammatory arthritis in a way, is generalizable across all 3 diseases that are involved in inflammatory arthritis. I don't think it distinguishes well between the two. There's certainly a world of other animal models that could be explored and we're looking into that. But I don't think I would put my chips in any one of them ahead of the others.

Your next question comes from David Nierengarten with Wedbush Securities.

I had one on clinical trial recruitment. Just -- when you think about physicians' motivations here to recruit patients on the study, is there -- and presuming you've socialized this clinical trial design with plenty of physicians, is there a worry or concern that there might be a bias towards kind of getting up patients for the combo arms and avoiding getting patients into the monotherapy arm or is there enough appetite out there for new agents to recruit as they come? And then as a part of that, is there any stratifications by a number of previous treatments in the recruiting and reporting data?

Really good question, David. I think from a recruitment perspective, I believe we try to be very thoughtful in terms of designing this study to be attractive really throughout the duration of it. It is the rationale for why Part A is open label, so that patients don't have to randomize between placebo and a drug that they may already have commercial access to against the same mechanism. And I think we've seen, given that we've only been active for a few weeks and already are seeing multiple patients in screening from sites where they have access to other things. We believe there is substantial appetite for long-acting versions of mechanisms that are pretty well proven. And I think at this point, we've now had conversations with approximately 300 investigators across the globe to get them excited about this study, and I think it's very high. And then as we transition to Part B and have to introduce the placebo, now we have 3 combination arms in that portion of the study that should drive quite a bit of enthusiasm as well. Maybe I'll pause and let -- I don't know Sheldon, do you want to comment, you've run many studies in this space over the years in terms of kind of what we're seeing relative to previous?

Yes. I think the fact is that investigators are very excited about the trial. They've all -- many of them have already been participating in platform trials, so they understand this kind of setup. So J&J, AbbVie have run similar number of arms. But I think on the other hand, on the other side of this is we have a fairly deep bench of people who are -- who have previous experience with clinical trials, multiple companies, J&J, Arena and Prometheus. So we have a very deep bench of gastroenterologists at our team, and we're out there talking to the investigators and they're telling us they are excited about the new agents, the half-life extended as well as the combination therapy. So we're very optimistic about our prospects of enrollment.

Thanks, Sheldon. And maybe I think the second question was regarding stratification in the trial. Deanna, do you want to cover that one?

Sure. So we do have several stratification factors. But given this is not a large Phase III study with large sample size in each arm, we have limited numbers of factors, but we certainly take into account the status of product exposure against therapies.

Given we are at the hour mark, I will now turn the call over to Cameron for closing remarks.

I just want to thank you all for making the time this morning. We're very excited to have shared very positive interim data on both of our TL1A molecules, and I think we are more excited to be turning the page to look at what we believe are 2 groundbreaking Phase II studies that we expect to deliver a series of exciting readouts over the next couple of years. So thank you for making the time this morning, and we look forward to catching up with more of you.

Ladies and gentlemen, this concludes your conference call for today. We thank you for participating and ask that you please disconnect your lines.