Good day, everyone. My name is Olivia, and I will be your conference operator today. At this time, I would like to welcome you to the Kymera Therapeutics Second Quarter 2025 Results Call. [Operator Instructions] At this time, I would like to turn the call over to Bruce Jacobs, Chief Financial Officer.

Good morning. I'm Bruce Jacobs, and I'm kicking off this in place of Justine Koenigsberg, our Head of IR, who is out today. Joining me this morning are Nello Mainolfi, Founder, President and CEO; and Jared Gollob, our Chief Medical Officer. Following our prepared remarks, we'll open the call to questions from our publishing analysts. If you'd like to ask a question, please use the raise hand icon, which can be found at the bottom of your meeting window. And to help us move efficiently through the Q&A session, we ask that you're ready to unmute your line when called upon. In addition, we ask that you please limit your question to one and a relevant follow-on to be sure we have enough time to address everyone's questions this morning. Before we begin, I'd like to remind you that today's discussion will include forward-looking statements about our future expectations, plans and prospects. These statements are subject to risks and uncertainties that may cause actual results to differ materially from those projected. The description of these risks can be found in our most recent 10-Q filed with the SEC. Any forward-looking statements speak only as of today's date, and we assume no obligation to update any forward-looking statements made on today's call. With that, I'll turn the call over to Nello.

Thanks, Bruce, and thank you for joining us this morning. As I mentioned at the beginning of the year, we set our sets up for a very productive and exciting 2025, and we're delivering on that promise. The updates we've shared in the first half of the year represent a powerful validation of Kymera's innovative and disciplined approach to drug development within the biopharma industry, while paving the way for our future progress across our high-impact immunology pipeline. We're committed to leveraging the unique capabilities we have developed to unlock disease biology and deliver groundbreaking oral degrader medicines for areas not served well by existing technologies. Today's immunology treatment landscape still leaves millions of patients without adequate options, forcing difficult trade-offs between efficacy, safety, cost and convenience. Millions of patients with life-altering immune-inflammatory diseases don't have access to advanced systemic therapies, mostly injectable biologics. This is true if we look across countries with extremely diverse systems on how they prescribe, reimburse and deliver these highly effective medicines. The issue is really more fundamental than the inefficiencies of the health care ecosystems around the world. Simply put, well-tolerated oral drugs that can be as effective as this difficult to access injectable biologics have the potential to transform the treatment landscape and in doing so, impact lives of millions of patients. This is what we're set to do at Kymera. It's an exciting time for the company, and I want to take a moment to briefly recap some of the key accomplishments of the first half of 2025. Starting with our first-in-class STAT6 program. We completed the first KT-621 trial in healthy volunteers and reported positive results that exceeded even our high expectations, surpassing our target product profile. Importantly, the data further derisks our path forward and highlights the possibility of KT-621's/dupilumab…

Thanks, Nello. Looking back on the last quarter, we were excited to share the first KT-621 clinical data, which we believe greatly derisks the next stage of development. We identified clear goals for the Phase I healthy volunteer study, and the data not only hit the mark but in many instances, exceeded our expectations with compelling translation from preclinical studies to humans. The primary objective in the healthy volunteer study was to show that we can robustly degrade STAT6 in blood and skin which we define as a reduction of 90% or more at doses that are safe and well tolerated. As shown here, the results exceeded our expectations across every measure. We showed more than 95% degradation in both skin and blood at very low doses. The safety profile was undifferentiated from placebo, and we are encouraged by the biomarker profile, which we believe is at least comparable to what dupilumab showed in healthy volunteers or patients and, in some cases, is superior. That said, I'd like to take a few minutes this morning to recap the results and next steps with KT-621 which we believe has the potential to profoundly alter how Th2 diseases are treated by delivering an oral drug with a biologics-like profile. For the full data set, please reference the slides presented in early June, which are available on our website. As a reminder, we enrolled 118 volunteers in a randomized, double-blind, placebo-controlled study to assess single and multiple ascending doses of KT 621 across a range of doses from 6.25 to 800 milligrams in SAD and from 1.5 to 200 milligrams in MAD. In all SAD cohorts, including the lowest dose of 6.25 milligrams, KT-621 degraded STAT6 by 90% or more and a dose of 75 milligrams or greater achieved complete degradation with greater…

Thanks, Jared. I will quickly run through our results for the quarter. Also because this past quarter has been busy with collaboration and financing activity, I wanted to provide a brief summary of our recent news as well. As I walk through the second quarter results, please reference the tables found in today's press release and 10-Q, which was filed this morning. Revenue in the second quarter of 2025 was $11.5 million, all of which was attributable to the Sanofi collaboration. With respect to operating expenses, R&D for the quarter was $78.4 million. Of that, approximately $8 million represented non-cash stock-based compensation. The adjusted cash R&D spend of $70.4 million, which excludes that stock-based comp reflects a 3% decrease from the comparable amount in the first quarter of 2025. On the G&A side, our spending for the quarter was $17.6 million, of which $7.4 million was non-cash stock-based comp. The adjusted cash G&A spend of $10.2 million, again, excluding that stock-based compensation reflects a 6% increase from the comparable amount in the prior quarter. And overall, adjusted operating expenses in total were down slightly from the prior sequential quarter. We ended June with a cash balance of $963 million. Our quarter end cash balance included the base proceeds from our $250 million follow-on offering that closed at the end of June. The June total does not include either the additional proceeds from the underwriters' over-allotment option, which was fully exercised in July or the first payment that we received from Gilead as part of our recently signed CDK2 partnership, both of which were received in July. As a result, we ended the month of July with a cash balance of approximately $1 billion, providing a cash runway into the second half of 2028. Just a quick reminder that our runway…

[Operator Instructions] First question is from Michael Schmidt from Guggenheim.

It's Paul on for Michael. I had one on the dose levels that you're exploring for 621. Maybe first, could you provide some color on that decision to add the second dose in the Phase Ib study, I think it's probably safe to assume that both the doses fall within the broad range that achieved complete STAT6 degradation, but just wondering how you're thinking about exploring both the high and the low dose, which is perhaps 2 doses in the higher range?

Yes. Thanks, Michael. We want to make sure we're -- you can hear us. So as we said today, so the doses -- both doses are within the range that we explored in the Phase I healthy volunteer study. And as we've also said, as you're aware, we had initially decided to explore 1 dose, thinking that roughly 20 patients will be enough to give us the data to speak to what is the profile of that 1 dose. And then obviously, as we were moving along with the enrollment and given how quickly it was going and given that we were able to assess the performance in the 1 dose, we decided to explore an additional dose so that we'll get even robust translation from healthy volunteer to patients of STAT6 degradation. I think it's important to keep in mind that in the healthy volunteer data, we had multiple doses. I would say, almost all those doses besides one met our target product profile. And so we wanted to confirm that the really, really robust profile could be translated into patients with the same level of fidelity. And so I think we're happy that we did that. Obviously, I'm not going to speak too high, low, medium, et cetera. Rest assured that the main goal was really to refine the Phase IIb dose selection. And so all that happened so quickly that now between the healthy volunteer data and whatever data we have access from this study, we're able to firm up and select the Phase III -- the Phase IIb doses even in the absence of completing the Phase Ib study.

Great. And if I kind of a quick follow-up on that point. And mostly on just what back into the dose selection for the Phase II studies? Was it predominantly the healthy volunteer data you presented? Was there anything emerging from the Japanese studies or GLP tox? Can you say if there's a different range of doses being explored between the AD and the asthma studies.

Yes. So great question actually. So as you saw, it was a very, very busy Q2. I don't think we've had a busier Q2 in the history of the company, given everything that we've accomplished. So I will say that if you look back at the healthy volunteer data, there was a dose selection based on this data. Everything else that we've done confirm was able to confirm our initial instinct. We didn't learn, to be honest, anything new that made us change the initial instinct, let's say, on dose selection, but it was highly encouraging that everything that we've seen in healthy volunteer was supported by, obviously, the four-month tox, which we said was completely clean. The Japanese study, which was very much in line with the U.S. study and the early -- let's call it, early data for the Phase Ib.

The next question is from Derek Archila at Wells Fargo.

Congrats on the progress here. So just 1 and a follow-up. So basically, I just want to understand maybe following up on this line of questioning just in terms of what you would expect to see at these additional doses that you're looking at in the Phase IIb, ultimately, like we saw very good degradation and pretty quick. So I guess how do you think some of the doses will differentiate. And then just a follow-up to that, what do you actually expect to see with the follow-on STAT6 that you're developing? And what sort of optionality are you really looking for with that molecule.

Great question, Derek. So the first one, I just want to confirm we're talking about the dose for the Phase IIb. So I think the important thing for a drive, obviously, is to find a dose that has the best risk-reward profile. And so I think what we want to ask in a 16 -- let's say, for AD, a 16-week study, is what is the maximal or we believe close to maximal at that point, level of clinical activity that we will see and what is the safety profile at different levels of degradation. Obviously, we will explore maximum degradation, which we call complete, which, again, is where really we see in most subjects STAT6 level be below the lower limit of quantitation. So we want to ask that question, what is the clinical profile of maximal degradation? And then obviously, we want to ask the question at a couple of lower doses just to again, at the end of the study, being sure that we're taking into Phase III, the profile that we believe has the best risk reward. So it's obviously a necessary step that we need to take as a company to fulfill regulatory requirements to do dose ranging studies before selecting a Phase III dose. I think we have bets in the company where the Phase III dose will be already, but we got to run the studies and make sure that we do all the right steps to derisk the program. With regard to the follow-on molecules. So that's something many of you that has followed us for years know that every program, we always have a next-generation molecule as you saw for IRAK4, Sanofi decided to focus the efforts on the follow-on, and we call it the next-generation molecule KT-485. For STAT6, to…

Question is from Andrew Newkirk at Goldman.

Two for me as well. Maybe the first, recognizing the primary objective of the Phase Ib data, is to show a dupi-like profile here on biomarkers, but I was hoping you might be willing to frame your expectations on what you'd like to see on the clinical efficacy measures, particularly EASI 75 as well as NRS. And then secondly, just noting the completion of the GLP tox studies that you mentioned. And obviously, your Phase I healthy volunteer also looked really clean, but if you could just speak to the potential safety risk of degrading STAT6 completely, what type of signals are you most looking for in the Phase Ib to really feel comfortable here with the safety profile as you move forward?

Thanks. Great question. Jared, I thought maybe you could take at least the first one, if not both.

Yes. In terms of on the clinical expectations, I think we emphasize always that the primary objective here is to show a robust STAT6 degradation in the blood and an active AD skin lesions and to show that, that results in a dupi-like biomarker, in fact, both in blood and skin were -- and skin, we're looking at the Th2 transcriptome and we don't want to see a dupi-like effect there. We sort of have set of expectations around biomarkers, I think TARC is the most important blood biomarker probably in AD, where dupi studies have shown even at 28 days, about a 70-plus percent reduction in dupilumab. So that's a general [ war plug ] that we would expect to see in patients who like in those dupi-AD studies had greatly elevated TARC levels at baseline. We'll be looking at other biomarkers in the blood as well as these various transcriptional biomarkers in the skin. In terms of clinical endpoints, again, we've always emphasized that in the absence of a placebo control, these are more exploratory. However, we think we do have an opportunity to look at end points like EASI and pruritus, NRS and IGA because we know from dupi that you can see impact on those biomarkers as early as 28 days. And we're not really giving specific numbers where that bar would be set. I think the published data are out there with dupi and one can look at those published data at 28 days and get a sense for what we mean by sort of being in the ballpark with regard to those clinical endpoints. In terms of your second question around safety risk, as you noted, we've been very pleased with what we've seen in our GLP tox studies. We've now completed our four-month tox studies as now indicated, and we've seen no safety signals whatsoever. That's very in line with our four-week GLP tox in our prior non-GLP tox studies. We are very encouraged by the fact that our safety profile was undifferentiated from placebo in healthy volunteers with 2 weeks of dosing. So that's very encouraging. And now we'll be looking at safety with 4 weeks of dosing, of course, in the Phase Ib. I think overall, this is in line with our expectations based on our mechanism of action and based on the fact that it appears that STAT6 is highly selective for the IL-4/IL-13 pathways and human genetics. I have pointed not just to the phenotype of abnormalities of STAT6, but also to the safety of knocking down STAT6 have mouse knockout studies. And so this is all in line with what we expected for a transcription factor that is very specific for IL-4 and IL-13, and for a drug like ours, and it's highly selective just for STAT6.

I would only -- thanks, Jared. I wouldn't say anything differently. I will only add one thing, just to be clear, again, as Jared said, on the clinical endpoints, it's difficult to compare -- it's also difficult to compare placebo-control randomized study, like the industries followed these arguments over comparing placebo-controlled studies. So it's even more difficult to compare noncontrolled study. But I just want to say our expectations are that we will have a very active drug. I don't want to hide behind the impossibility to compare. We expect that this mechanism is going to be on par with what dupilumab has shown and that's the bar for us without talking about numbers.

The next question is from Faisal Khurshid with Leerink.

Just wanted to ask on the doses for the Phase Ib and the Phase IIb. Are you able to confirm if the dose that you added to the Phase Ib is higher or lower than the dose that you originally went in with. And could you also confirm if like either or both of these doses are part of like the 3 that you've selected for Phase IIb?

So I don't want to get into the higher or lower, because I think whatever I say, it's going to be viewed one way or the other. What I can say is that both doses have been tested in the healthy volunteer studies. I don't want to talk about what our doses are for the IIb because I think we might choose to keep that, as I've said in other venues, to keep that close to the vest for as long as we can, only for competitive reasons. All I can say that we have several doses in the healthy volunteers that performed really well. And so really, the main driver here are these doses going to perform as well in patients given that -- I actually don't remember the number, Bruce will know better, but we're spending tens, if not 100 plus millions of dollars in these 2 studies. And we're not going to optimize over -- for these studies on making or thinking that we selected the right doses. These are consequential decisions. And so given that we had the time to do it, we said, let's make sure. So that's really what's behind these. And I think once we'll share the data we can add a bit more color to work in first.

Got it. Makes sense. And then could you confirm if it's still 20 patients for the Phase Ib? And then also like between the 2 doses, would you like to or do you have to see a dose response between those 2 doses?

Great question. So I think what we said that the goal of the Ib was approximately 20 patients, and that's still the case. I don't want to get into the dose response. I think we will talk about it once we share the data.

The next question is from Alex Thomson at Stifel.

I guess another question on the next-gen STAT6. How different is the scaffold binding to STAT6 than 621? Is that a key part of this decision making? And when might you consider potentially splitting indications here? Is that a near-term decision? Or are you going to wait quite a while before that comes down?

Yes. So what I can say -- that's a great question, Alex. So we have several scaffolds, let's call it, across actually all [ binding molecules ], whether it's E3 ligase or it's STAT6. We have plenty of chemistry, some patterns are published from us as many have seen, there is plenty they haven't yet. So we have a plethora of chemistry in this program that covers everything that you can imagine. So maybe I'll leave at that. On the indication splitting, it's a bit obviously challenging to think about that particular end game given kind of the evolving landscape right now in terms of pricing and reimbursement and global versus U.S. So I think we want to keep maximum optionality and that kind of the goal behind everything that we're doing. But it's difficult for us right now to at least disclose what's the latest thinking on that. But as we get closer to Phase III, which actually with the recent raise, hopefully was clear from our remarks earlier now with the money we have in hand, we can actually initiate multiple Phase III studies. So I think as we get closer to those, we'll be able to disclose more about what our indication sequence and strategy will be.

The next question is from Tazeen Ahmad at Bank of America.

Going back to the data that you're going to have by year-end, I just wanted to ask, you talked about your expectations for what data you're going to show. Should we assume that you're also going to be able to show some level of itch data? I ask because some doctors have indicated that in addition to, let's say, EASI scores, that is something that they feel is important when they're going to make a decision in a real-world setting about what potential options they might choose.

Yes. Tazeen, it's a great question. And as Jared said, yes, we will show EASI, pruritus, NRS. And so itch is going to be an important factor. As you know, itch is -- probably has the biggest impact on quality of life of these patients and so it's something that we're watching very closely. So we will share that data as well.

And will that be for all the patients that you're going to show or it'll be a subset?

Well, I mean if we have collected the data, yes, so we'll share it. So yes, it should be all patients.

Next question is from Kelly Shi at Jefferies.

Congrats on the quarter. So one question on STAT6. Conjunctivitis is believed to be an on-target AE for DUPIXENT. So do you expect to see a similar level of conjunctivitis in KT-621 Phase Ib trial like in 1 or 2 patients, and also could a daily oral drug differentiator in safety profile versus injectables due to a potentially more flat PK curve?

Maybe I'll start, and then I'll pass it to Jared, that can speak more to the medical part. I mean our view with Kymera is that STAT6 and hopefully, it's not just here at Kymera, STAT6 is the selected transcription factor of IL-4 and IL-13, and we've shown pre-clinically, early-clinically and hopefully, will show in Q4 that if you block STAT6, you can phenocopy dupilumab. So if conjunctivitis, which is actually mostly if not only seen in atopic dermatitis patients, so it's really a feature of the disease and these IL-4 and IL-13 biologics. So again, if conjunctivitis is an on-mechanism event -- adverse event for IL-4 and IL-13 biology, then we expect to see. If it's to do with the receptor or the cytokines, then we wouldn't see it. So it's hard for us to know. Maybe, Jared, you can speak to also, is it seen after only 4 weeks. I don't know...

Yes. I mean, mechanistically, it's not really know why some patients, especially AD patients do develop conjunctivitis. If you look at the dupilumab studies, when you do see conjunctivitis, oftentimes, you'll see it within the first, say, 4 to 8 weeks or so of treatment. And then over time, it actually tends to diminish. It is an adverse event that one does see with dupi, it's not a dose-limiting adverse event and most of the case of dupi are sort of in the mild-to-moderate range. I think importantly, we haven't seen it preclinically in our tox studies, we haven't seen in healthy volunteers, and we really wouldn't expect to see it there in healthy volunteers since this appears to be something unique to AD patients. But as Nello said, we don't have any reason to believe that we'd see either less or more AD patients compared to what dupilumab has seen.

Yes, we're watching it because it's an interesting -- obviously, feature of many of these drugs, right? It's not only dupilumab, all the IL-13 drugs have it. So we're watching that very closely and see if we see it in our four-week studies. And we'll -- obviously, we'll share all the data in 4Q. And then you talked about the safety difference between once oral daily and biologics. I mean from what we have both understood and what we've empirically derived in our preclinical studies, dupilumab has a very, very robust pathway blockade. I would compare the dupilumab pathway blockade pretty much in line with the level of pathway blockade we see from our 50-100 mg dose, 200 -- the complete degradation type of pathway blockade. I would put it on that kind of level of pathway blockade. So if that's the case, then I don't see why pathway blockade coming from STAT6 degradation should look different from pathway blockade from an IL-4 receptor -- IL-4 blockade. So anyway, I think that's another feature and another part of the analysis that we will do again. I'll repeat in our preclinical study, in our adding the four-month tox, K-621 has been exceptionally well tolerated. So we'll continue to, again, watch everything that happens in the clinic.

Question is from Judah Frommer at Morgan Stanley.

Just one for us. I guess can you comment a little bit further on enrollment progress and the success you're having there and maybe what feedback is from investigators. Is the oral administration of the drug resonating. I'm curious if you think you'd have similar success if there were a placebo arm in the trial?

Yes, it's a great question. So the challenge of a 28-day study, remember, is that the patients are not going to be on the drug beyond day 28. We don't have an extension arm to the study. So it's -- I would say before we started the study, we were nervous because there are a huge amount of incentives for patients to come on to the study besides knowing they'll be on an active drug. And that's part of the reason why we decided not to have placebo, we thought it would have had an impact on our enrollment rate. As we expect -- part of us or our expectation was that patients do want an oral drug. And so I think we are seeing that in our study and this has allowed us to meet our enrollment goal, I would say, even exceed our enrollment goal for sure. And maybe that's where I'm going to leave it. I think once we start seeing more differentiation, it's probably going to be in the Phase IIb study where now you're offering 16-week, potentially [ OLE ], so that would be interesting to see our enrollment versus biologics and whether it's telling us also something about what patients are also looking for in the market.

The next question is from Kripa Devarakonda at Truist.

Congrats on the progress through the quarter. So I'll ask one non-STAT6 question. So congrats on your CDK partnership with Gilead. I know this diversifies your pipeline into oncology, where you've been focused a little bit more on I&I in the recent past. But given the data we've seen so far with CDK2 inhibitors, can you talk a little bit about how you think the degrader could be differentiated based on the data that you have? And what the strategy of development is? And then I have a follow-up.

Yes. Thank you. So just to be clear, our discovery engine has been also very focused on immunology. We have programs that we were working on from the earlier days and one of our program wasn't CDK2. And so with our strategy shift to focus on developing immunology drug, we decided that it was best to place a very exciting CDK2 program from a development standpoint in the hands of a partner that was committed to that space. So that's a bit to the strategy. The reason why we have that program because we firmly believe that small molecule inhibitors of CDK2 are really not able to selectively target CDK2. They all inhibit to a large extent, CDK1 that pharmacologically active doses to different degrees. And that it leads to clinical doses that are probably not optimally blocking CDK2, again for the risk of hitting CDK1. Another important aspect for us to develop this drug was to have a brain-penetrant asset so that we would also address potential brain secondary tumor metastasis from breast cancer. And so our degraded program -- molecular degrader program is highly specific CDK2 also reaches the CNS. And we believe it's going to -- has the potential to be best in class. If I look at the small molecules out there, it's by far superior. Obviously, I'm not aware of other programs that are in early discovery, early development. So I can say obviously, for sure, but with regards to the development, that's a question you have to ask Gilead. We can't speak for them on that particular front.

Okay. And just following up on Tazeen's question about itch relief, and this is something that we've heard from KOLs too that it's really important to see rapid itch relief. Will we get a sense of that when we see the data, the rapidity of response [Audio Gap] fourth quarter?

Yes, we will. I mean, as Nello already mentioned, looking at pruritus NRS is a key part of the sort of suite of clinical endpoints. And we'll be looking at it fairly regularly [Audio Gap] as we'll be looking at EASI. So we'll have a good sense of the kinetics of impact on itch as well as on EASI, that will be certainly part of the profile that we share once we have those data in Q4.

Yes. I think you'll see, hopefully, that will be the case. But it would be like we'll show day 7, day 14, then 21, day 28. So you'll be able to see the kinetics of all of these parameters.

[Operator Instructions] The next question is from Mayank Mamtani at B. Riley.

Congrats on the progress team. Any color you're able to provide on the baseline EASI scores of the patients you're enrolling or have enrolled? And I wonder always about the screen failure rate for atopic derm trial sites. And maybe just remind us how you're measuring degradation in skin tissue. There's obviously a couple of ways to do that. And lastly, anything you've learned on the degradation from the four-months GLP tox studies you completed?

So you asked 4 questions in there. So that's a way to...

I will not ask a follow up. I promise no follow up.

Yes, but you asked 4. Let's see if I remember. So the first one was -- the EASI -- yes, we're not going to comment on the baseline EASI, but I will refer you to entry. So the baseline criteria for entering the studies EASI above 16 -- 16 or above. There is obviously itch as well. There is PSA more than 10%. So we have a strict criteria that really overlap with what has been done with dupilumab. On the -- again, on the failure rate, again, I don't know if we'll speak when we released the data. All I can say is that our team is watching the study very closely. And we've worked very, very hard to make sure that patients that enter a study actually have atopic dermatitis, which would be shock that, that could be possible if you don't watch the study closely that their disease is active. And obviously, that their lab work is in line with making sure we're not taking sick patients on our study. So I think when you take all of that, that results into obviously screen failures that, again, I'm not able to comment on today. On the degradation in the skin, as we've done in many of our studies, we are fortunate enough to have patients on our study be willing to take biopsies, which, as you know, it does add an additional layer. That's why we're so impressed on how we were able to enroll patients again quickly because we ask patients to undergo biopsy at baseline and day 28 to measure STAT6 with the mass spec. So that's how we're going to measure it.

Anything you learned from the GLP tox study on degradation?

I mean, all we learned in these studies is that we -- obviously, at this tox doses there is no STAT6 anywhere to be found that we degraded completely, that's maybe all I can say. Yes, if the question is, does the STAT6 degradation wanes off after sometimes? Obviously, the answer is no. We see STAT6 degradation all throughout the study.

The next question is from Jeet Mukherjee at BTIG.

I know we're a long ways out, but can you speak to payer willingness to cover therapies in the dermatology space that given alternative administration format like an oral option with DUPIXENT-like efficacy for KT-621? Or is there a bar truly superior efficacy versus standard of care options?

Well, no, it's a great question. We believe that when you make the case for an oral option, first, you will hear from prescribers that actually you don't even need to have dupilumab-like activity for expecting a substantial adoption in the market. It just speaks to the fact that there is a need of flexible, easy to prescribe, reimbursed and take medicines. So the reason why we see dupi-like in appeal is because all the data we've seen so far speaks to that. And so that's why our bar has always been there and hopefully will continue to be there. Again, I think when you make the case for having a therapy even with the same activity, you're telling actually insurance companies and prescribers and patients that this drug has a much bigger impact on their quality of life and asking for a lot less, right, in terms of visit to the doctors, testing, needing or lack thereof of cold storage of the drug, needles, injection site reactions. So I think that's the value case that a drug like this will have. And especially if you compare it to -- for example, the only drug there right now is approved in AD, it's a drug with a black box. And that drug actually is doing quite well. I think that speaks -- and then it's a drug that requires testing before you start the therapy. And so it speaks to the hunger that this market has for an oral drug. And I think you've seen in all markets, oral drugs and multiple effective therapies are needed to expand access and penetration. Now the -- especially atopic dermatitis market is really dominated by a single player, I would say, mostly with dupilumab, but it still has less than 15% penetration. I would say, if you look at all moderate-to-severe, it's less than 10%. And so I think we need this option to expand access dramatically in the U.S. and all over the world.

The next question is from Jeff Jones at Op Co.

One question from us on IRAK4. Can you provide any additional detail behind what drove the exchange of 474 for the 485 candidate, given the specificity differences, was there something that was being seen with 474 that was concerning?

Great question, Jeff. Thanks for asking about this. So just to remind everybody, the decision was made by Sanofi to focus all the resources of the IRAK4 collaboration on KT-485. Based on preclinical data KT-485 seems to be superior to 474 on both potency distribution, and we demonstrated also lack -- complete lack of the subclinical QT filing that we have seen with 474 in our clinical studies. I will also reiterate that particular finding was self-resolving with continued dosing, meaning that it will go away as you continue to dose. And we didn't learn anything even in the ongoing Phase II studies that spoke negatively with regards to the safety of the drug beyond what we've already shared. So I think it was really focus on the fact that 485 overall seem to add a better profile, and we believe both clinically and maybe commercially more competitive. But since you asked me about actually IRAK4, I thought it's also interesting to see how the landscape is evolving. I don't know if you guys have seen AstraZeneca starting -- about to start the big Phase II study in COPD after they've run a small earlier study, which we haven't seen data for, but they have shared that they're going to share their data for their IRAK4 inhibitor in COPD. So that's another indication that we at Kymera thought IRAK4 could be well positioned for. So it's exciting that a big company is -- I think it's 400 patients. No, it's more than 100 that, 1000 patient study. So anyway, just the field continues to learn and evolve. And we're excited to have a great asset out there that hopefully could also go towards that direction. But that's something that we need to obviously discuss with Sanofi.

The next question is from Andy Chen at Wolfe Research.

On IRF5, is there a reason for degradation in cytokine reduction and all that would not translate into humans? It looks like your STAT6 degrader has more than translated? Wouldn't all of that read through to IRF5? Or is there still something special about that molecule that makes you think that you're still maybe semi-concerned and maybe the de-risking steps are still ahead of you? And also what are the top 2, 3 safety signals that you'll be watching for in humans?

Yes. Great question. So for IRF5, I mean, we're at this stage, to be honest, we've been here for a while where all of our programs have translated really well. You can argue whether you like the target and the biology that translate, but all of our programs have translated really well in the clinic. So we expect IRF5 to translate just as well as KT-621. Also for IRF5, KT-579 in non-GLP tox, we've seen no adverse event of any type, and we went up to 200-fold above the expected 90% degradation human exposure. We are in the midst of buying the enabling studies. I'm confident we'll continue to see an exceptionally well-tolerated drug. So we're excited about that drug. We're working already, really hard not only to prepare for the healthy volunteer study that will start early next year, but the team has been spending in the past few months working on and planning our patient study that will start soon after the healthy volunteer study and we're prioritizing indications. We're talking to KOLs and refining protocol. So yes, we are working under the assumption that the translation will be happening just as well as it did for 621.

So we're just about up against time, operator. We'll try to just move really quickly through these last few.

[Operator Instructions] The next question is from Ellie Merle at UBS.

Just another one on IRAK4. I guess can you elaborate a little bit on what we've seen clinically with the first-generation IRAK4. And I guess, what gives you the confidence in the efficacy of this target in AD and HS. And I mean, I heard your comments on AstraZeneca. I mean any difference now in terms of how you might be thinking about the opportunity set across indications now for IRAK4. And then also just a follow-up on CDK2. Can you elaborate on some of the learnings on working with molecular glues versus heterobifunctional degraders, and your confidence in the selectivity for CDK2 with these programs?

Yes. So on IRAK4 quickly, we can't speak to what we've seen or not seen in these studies. Unfortunately, that is Sanofi's guidance on that. On the indications, again, this is again another question for Sanofi, but asthma and COPD have always been on the high priority list for that biology. Obviously, we're talking about non-eosinophilic COPD, which is a huge patient population. And with CDK2, so we -- again, we've historically said that these 2, the heterobifunctional degrader molecular glue are 2 complementary technology and they're not one, the next generation of the other, although many companies seem to go in that direction. We use molecular glues where we believe that binding site and ability to bind to the target is either not feasible or not with the selectivity, if you use binding -- specific binding to CDK2, it's really difficult to find selectivity against CDK1 and that's why we built our CDK2 degrader, which does not have any kind of cross finding with CDK1. That's why we went in that direction.

There are no more questions at this time. I would now like to turn the call over to Nello Mainolfi, for closing remarks.

Well, thanks, everybody. Sorry, we went a bit too long today. Another exciting quarter. We're here for any follow-up questions. You know where to find us, and thanks again for joining. Have a great day.