All right. Welcome, everyone, to the Jefferies 2026 Global Healthcare Conference. My name is Roger Song, senior analyst covers medtech. Now we have J.D. with me. Hello. Thank you, Roger. Awesome. All right. J.D., why don't you do a state of R for Nektar? You have rezpegaldesleukin as the lead and then across multiple indications, and then also anything else you want to highlight before we dive in? Yeah. Well, we're definitely heavily focused on regulatory T cell biology. I think it's one of the components and approaches medically that we've really taken and really defines our pipeline. It defines our work. I think we've established a level of expertise and certainly the most clinically mature data in this field, demonstrating that rezpegaldesleukin, with its Treg mechanism of action, is efficacious in multiple disease settings such as atopic dermatitis, alopecia areata, and even other conditions such as lupus and others, and psoriasis, where we've shown clinical activity with rezpegaldesleukin. We've sort of used that as a springboard. We've used that to propel our R&D pipeline. We've been focusing in a TNFR2 agonist as a target because that also interacts with a Treg compartment. It interacts with a different subset and classification of Tregs. The two major types are native, the ones that form in the thymus, and the induced, the ones that can form when T cells that are undifferentiated undergo a fate change and decide they could become induced regulatory T cells, and that's the target of the TNFR2 program. We're advancing multiple assets based on that target. One is a traditional bivalent antibody, and another is utilizing the fact that it works as a single arm, which allows us to also make bispecifics out of that same molecule. We call that NKTR-0166, and we plan to have an IND from at least one of those programs next year. I know that the majority of questions today will be focused on rezpegaldesleukin, and that really takes up the majority of our resources in the company. Yeah. No, it's good to be focused. On the other side is you do have a platform, I do notice you have a couple other earlier pipeline. A plug here, we just put out a atopic dermatitis landscape update with a doc survey, a U.S.-based dermatologist. I think rezpegaldesleukin could come up pretty positively from the result and then the piece, basically you are one of the three major pipeline product in development can take up a meaningful market share in the future AD, which is translated to $multi-billion opportunity. With that, you already reported the phase II-B data recent and then with the induction and the maintenance. I know you're well waiting for off treatment result next year, you're also full speed ahead into the phase III. Just tell us, okay, before you start the phase III, what are the gating factors? Any steps you need to go through before you can start the phase III? Yeah. We've already had our end of phase II meeting for atopic dermatitis with the FDA. We had that last year. We've also carried that out and received general health authority advice across all other regions, such as Europe and other different regions as well. We're in a very strong position where we now have all of the health authority feedback that we need for global trial to execute. To that regard, we're extremely excited to say that the first clinical study sites open this month in our phase III atopic dermatitis program, and we expect the first patient to randomize in July. Probably the early part of July with sites opening this month. That whole program, which now fully kicks off in earnest, has a number of pivotal trials that make up the clinical development plan. Two of the studies, which are each 510 patient trials, and they're duplicate trials in adolescents ages 12 and up, and they'll be evaluating rezpegaldesleukin versus placebo. It's in moderate to severe patient population. We'll be treating people for 24 weeks of induction with the 24 microgram per kilogram twice a month dose of rezpegaldesleukin, which was the most efficacious dose in the phase IIb study. It's pretty clear. A 24-week duration of induction, as we've shown that extended duration further improved the response rate over the 16-week induction that we used in the phase II-B study. We'll also be advancing patients that are responders into a maintenance portion, where they'll be re-randomized to receive either placebo or monthly or quarterly dosing of rezpegaldesleukin out to one year. I think, as many of you are familiar with our 52-week maintenance data that we presented for rezpegaldesleukin from that phase IIb study, we saw excellent performance in both the monthly and the quarterly dose regimens, which was, I think, a real differentiating feature for rezpegaldesleukin. Certainly has a very convenient dose presentation and frequency. Those two bio-naive patient studies will begin enrolling patients first in July, second in August. In the third quarter of this year, the bio-experienced patient population study will kick off, and that will be a study, again, in 510 patients. All of the patients had failed a prior IL-13 or other biologic, or they could have failed a JAK inhibitor as well. Again, same study design, same ages 12 and up patient population, same six-month duration, and so forth. We're very excited about this clinical development plan path that kicks off, and we expect the first top-line induction data from the first of those bio-naive studies around the middle of 2028. Okay, great. Yeah, exciting to see a new mechanism into the phase III for atopic dermatitis. As I mentioned earlier, we looked through all the pipeline. Yes, we have a long list of the pipeline drugs, but very, very few is actually exploring new biology. A lot of them are validated biology at the mono or multi-specific. I think rezpegaldesleukin turns out to be a very good new mechanism. We know OX40, but seems to we have some issue there. I think rezpegaldesleukin float to the top in terms of the stage development. Yeah. No, we definitely agree with you. One of the big differences is rezpegaldesleukin is an agonist. All of those other drugs in the pipeline you mentioned, they're all antagonists of various kinds. Most of them of the IL-4 and IL-13 pathway or combinations of that, and IL-31 or TSLP or some other kinds of targets that have sort of been somewhat recycled through the atopic dermatitis pipelines over the years. We think rezpegaldesleukin represents a whole new class that can be positioned in a new way because it actually affects the patient's immune status internally. By agonizing regulatory T cells, it switches the cellular complement that people have each time they take an administration of rezpegaldesleukin, and that can have really lasting effects. I think we've seen that represent itself in terms of how broadly rezpegaldesleukin can act. It can help with people that have a comorbidity such as asthma. Which is a different presentation of Th2 inflammation in a different tissue, but rezpegaldesleukin can impact both atopic dermatitis and asthma. We've seen rezpegaldesleukin impact Th1 mediated inflammation in the setting, for example, in alopecia or in lupus, and also Th17 mediated inflammation such as in the setting of psoriasis. It has a mechanism that has much more breadth because of the cellular mechanism of action that rezpegaldesleukin induces with Treg biology. Because sort of Tregs help regulate the underlying inflammation at its source, at the actual initiator and the trigger of inflammation, it also has demonstrated very high durability. We saw that with very infrequent dosing in the phase II-B, and even in our first phase I study, we saw six months of durability after a 12-week treatment cycle that we published a few years ago. We think all of those things really could position rezpegaldesleukin as a whole new approach to treating these diseases. With this phase III campaign that we're kicking off in atopic derm and in alopecia in the first and second quarters of next year, we think that's going to really give us a chance to establish rezpegaldesleukin as this whole new approach to treating diseases. Good. I think we'll touch on the alopecia areata in a moment. Maybe just a couple detailed question related to the phase III. I think you give us the design, how you will do the phase III, which is very rational. We notice maybe this time because the global trial, you expand the patient population into some of the demographic, maybe such as APAC or Asian. How you think of the baseline will be different compared to the trial you have been enrolling in terms of Th2, Th17 driven atopic dermatitis? Yeah. Well, firstly, we've evaluated the PK/PD profile in patients of Asian descent, and in multiple forms of that kind of genetic lineage. We've seen the same PK and the same Treg induction profile, so that was important. We've also, in other clinical trials, such as in our lupus study, we had a number of patients from multiple Japanese sites. We've also have experience in studying in that region of the world. You are correct, there are some slight biological differences in atopic dermatitis. There's a little bit more, say, a Th17 complement in the disease, patients of Asian descent, than there is in Caucasian, for example. We feel very confident because we've seen with rezpegaldesleukin, the activity to work broadly across different forms of inflammation, in contrast to a targeted therapy, say like an IL-13 blocker. It only blocks the IL-13 pathway, and that's it. If the patient escapes and maybe develops a Th1 or a Th22 component to their disease, right? It's very difficult for an IL-13 inhibitor to block that. With rezpegaldesleukin, we've seen activity in very different kinds of T cell inflammatory conditions including Th1, including Th17, and others. We feel very confident that the genetic differences in that patient population in the APAC region will not be an issue for rezpegaldesleukin, and we're very excited to be including numerous sites from the APAC region in this global study. Yeah. Okay, great. Then it's interesting you will separate the biologic naive versus experienced into different phase III. I think it's a very rational design because all the signals so far is biological naive. I understand that we ask the question many different ways that you're very confident about the activity post biologics, but it will be a lot cleaner if you can design the trial, just separate them whatsoever and then just show the result. I think we should have a lot higher confidence for the naive and then the experienced, we still have a pretty good kind of a chance you're going to hit that. That design, maybe an interesting question is how you think this design going to differentiate your label compared to the current atopic dermatitis drugs? I believe you are the first one to do that in a controlled setting. Yeah, that's exactly spot on. We think it's a very strategic and kind of like the most recent 2026 onward approach to doing this. The other drugs that are approved have a very kind of what I'd call a general indication statement. They're indicated as a drug for the treatment of atopic dermatitis, with maybe some body weight and age range kind of additional parameters around that statement, and that's fine. That's a general statement. As we're prospectively running a large study in a basically a biologic failure, like a second-line indication, then if we win in that study, we can be a lot more aggressive with our label aspirations because we would be the first to have demonstrated prospectively efficacy in that patient population. Which can now start allowing us to add that into our indication statement. When you start thinking about the landscape and the fact that there is quite likely a biosimilar Dupixent coming, and you start thinking about the landscape of patients needing to probably step through agents having a label that includes a second line and being a different MOA, rather than another version of the same MOA, all of those things we see as being very compelling, differentiating elements. That was a lot of our thought process in separating the patient populations, and in doing that kind of a study design. There aren't a lot of benchmarks because not many standalone placebo-controlled, randomized parallel design studies have been done. We'll be among the first, if not the first. Yeah. In the way you powered the trial, do you have a little bit lower expectation or assumption on the effect size, compared to the naive patient population? I know the power, the size is probably not driven by the efficacy because it's a placebo control, so it's highly likely you're going to hit the statistic. Yeah. With 510 patients and more patients on drug than placebo, yeah, they're very, very highly powered studies for sure. A lot of the size of the studies really is intended to ensure that the size of our safety database is really robust. Right. We want to enroll and randomize more patients and take more patients to one year plus of safety follow-up, especially for this first registration, because that really sets rezpegaldesleukin up very well, right, in future indications to come. Yeah. When we kind of going back to that bio experienced, it is 510 patients, you're absolutely right. It's very, very well powered study, not highly powered, an overpowered study. When we designed the statistics around it, we were being more conservative, for sure. Right. It's not been studied in that patient population. From first principles, we don't expect any difference in efficacy based on the mechanism of action of rezpegaldesleukin, and also based on the fact that the patients that will be in that study will be a mixture of people that were intolerant to the prior therapy, so it would've failed, not necessarily for efficacy, but for safety reasons. Others will have failed for efficacy. You also have a range of different kinds of patients that would be eligible for that trial. Yeah, we were more conservative in the statistical design. Yeah. In terms of the mix of the patient, you say intolerant, failed. What's the distribution work going to look like? Do we have a min-max for each subpopulation? Also, if it's a fail, would you include those patients fail, like a refractory patient versus a relapsed patient? Yeah. I think a good kind of proxy is, there was a trial run by lebrikizumab called the ADAPT study. It was published recently and that gives a good example of what the patient population is in a kind of a Dupixent failure. As I said, the four common types are people that are immediately intolerant or that acquire an intolerance. Those are the kind of the two most common on the safety side of patient. Then another are the efficacy side, people that were never responders or people that lost a response or maybe had an insufficient amount of response, like maybe they just barely reached an EASI-75 but couldn't hold it. In our study, we also will include people that will have failed a JAK as well. I suspect there'll probably be less patients that are in that category because likely people that failed a drug like Dupixent are more likely to join our study probably than to take a JAK. Those will all be the kinds of patients that we would expect to have enroll. If they only failed JAK, that's not a population. Well, if they followed the FDA-approved treatment guidance, right? JAK has to be used second line. Yes. Right. We would expect that people are treated correctly, so then they would've had to have failed an IL-13 first, and then they would've taken a JAK. Maybe they were, again, intolerant to the JAK. It's very common to have adverse events that cause you to discontinue a JAK inhibitor and so on. Got it. I understand, statistically you can be conservative on the effect size and then when you design. On the clinical perspective, based on your feedback from your advisor, will physician or patient expect how much lower or if at all in terms of the effect size after they fail Dupixent? For example, Dupixent is 40%-50% response, and then when they fail using other drug, they are expecting have a lower response. No you know. Yeah. Again, there aren't a lot of benchmarks here, but in the ADAPT clinical trial, so the efficacy to Lebrikizumab was very close in that patient population as it was in the phase III. Again, with some caveats, that was a single arm open label study. That's really the only benchmark data that exists. Obviously, when you run a study, it's very important to set expectations for patients, and the informed consent does that partly, and then the investigator and their discussions with the patient do as well. We'll make sure that we work very hard to set expectations for patients so that there's high adherence and that that trial is done as best as it can be done. Yeah. Got it. On the statistical method, how would you handle the missing data and the discontinuation data? Is also the multiple imputation like you did before? Mm-hmm. Yeah. That's really commonplace nowadays. That's what the agencies recommend. Missing data is treated with imputation. People that had any kind of prohibited uses, they become intercurrent events, right? For example, if they discontinued a lack of efficacy or a prohibited med is taken, they would be non-responders. This will be similar to lebrikizumab, I think is more modern statistic. Mm-hmm. Yeah. They're not treating those missing or discontinuation patient as a non-responder. They're just treating at the multiple imputation. Yes. Correct. They treated them as missing, and missing data was imputed. Okay. Got it. The last point on AD is you will have the one year off treatment data next year, but you will start phase III this quarter. in the next month. How those data going to inform your pivotal or potential label and maybe the real world use. How are they going to use that data? Yeah. Well, I think one of the things that's important is that the IEC has just issued a kind of a consensus guidance on the definition of remission, right? It was a JAMA Derm article. That's very interesting because now the field is starting to come up with a consensus criteria of what could constitute remission. Now, there hasn't really been any study that's used it yet. We might be one of the first that assesses our data using that kind of new metric. For us, what we think about, whether respeg is able to be withdrawn for extended periods of time or if you take it four times a year or something, that's all a win, right? That's great. That's a great medicine, right, as far as I'm concerned. If the medicine can be stopped or taken with very low frequency, we see that successfully. That's one thing. On the other thing is we also see that respeg has shown tremendous durability, right? We've seen that in the phase I study that we published in our "Nature Communications" paper in 2024, and this coming one year off drug, right, is going to be the next data point. I do think that for us it's related to how respeg can be a whole new class of agent because ultimately we believe that if we fix the underlying immunological problem, it should create lasting benefit for the patient, and we know that can reflect itself in preference, whether it's doctors that would prefer it, patients would prefer it, and other things like that. In terms of how we're dealing with that in our clinical development plan, in the design of the phase III program, we have multiple periods where the patients are re-randomized onto placebo throughout the duration of the program. For example, in the phase III, after the induction period ends at six months, responders are re-randomized to either placebo or monthly or quarterly. That's the first withdrawal assessment. At the end of one year, responders, again, can be re-randomized to stay on drug or to placebo. We'll be assessing off drug from 12 months. We'll be assessing multiple patient populations and multiple time of exposure relative to drug withdrawal. That'll all be ultimately a big part of our entire package. That's your phase III. You have 24 weeks and then the 52 weeks, and after 52 weeks, you have another year off and then the treatment? Yeah. Into the LTE. LTE. Okay. Yeah. Yeah, patients again are re-randomized to either stay on drug or to move on to placebo. They will still be blinded for those? Mm-hmm. Yeah. Oh, okay. You have an entire two years trial. Correct. One year is LTE. Yeah. The phase III is one year, and this LTE is the study afterwards. Got it. Yeah. Again, for us, this is very important because respeg has shown this extended off drug potential. We want to continue to explore that as part of the phase III program and into registration. Yeah. You don't need the second year data to support the initial approval and the label, but that data will be updating the label once you get the data. Yes. Correct. Yeah. Then the initial label will be based on the first 24 weeks data, or you need the whole year data? Yeah. You like to have in the safety database for the BLA package, right, you want to have as long of a duration as possible, but we start to set the clock on the number of patients that have one year of exposure, right? That starts to become the anchor point of your safety database. Then it's typical to do an integrated safety summary where you pool all of the safety data in the program to date all together and continue to bolster that safety database. When we move through the alopecia phase III, we'll continue to do the same thing. Yep. Okay. The one year is for the safety for the initial BLA submission. Mm-hmm. Yeah. for AD. Okay. For AD, you know I think the phase II, including the long-term 52 weeks data looks pretty good. I think investors probably are just a little bit confused about the statistical method you're using, the multiple imputation. I understand it's early data, so you don't want to be too conservative on the discontinuation patient. You have the rationale to choose that. How should we think about the phase III design if you do the same thing? How are you going to expect the discontinuation of missing data going to end the phase III? Yeah. In the exact same way. Same as in the atopic derm. Again, it's very common practice to use imputation for missing data, and that's our approach again. If patients have an intercurrent event, which means that they took either a prohibited med or they discontinued due to progression, then they're non-responders. Yeah. Any patient that has missing data, however, their missing data can be imputed. Of course, patients that don't have missing data are not imputed. They're actual data. And then- Again, very similar. Yeah. Then your expectation for the discontinuation patient will be lower than the phase II? Oh, indeed. Yeah. Firstly, if you just look across in almost any indication between phase II and phase III, the discontinuation rate always drops in the phase IIIs. That's driven by a few things. The first is that when you do the phase II, you don't approve concept. Nobody really knows what will happen. You might also have dose responses, which includes lower efficacious doses and stuff. All of those change in phase III because you've established efficacy and you've established your dose. That's one big difference. Another big difference is that in our phase II study, we didn't offer a crossover or an escape arm, like we did in atopic AD, and that was really for resource kind of issues for us. In the phase III program in alopecia, that won't be the case. We'll offer patients a chance to cross over at the end of the induction. That also has a big impact. I think the biggest impact is when we ran phase II, we just didn't know. Yeah how long it would take and how efficacious respeg might be. That was really a proof of concept study. Now we do. We can set expectations adequately with patients, with doctors. We can put it into the informed consent. This is how long this trial will take, and this is what you might experience. All of those things completely change the mindset, and they have a very positive effect. We fully expect that the discontinuation rate will be much lower in phase III than it was in phase II. Got it. Then the phase II, this 24-week off treatment data later this year, do you need that data package for your end of phase II meeting with the FDA to design the phase III? Or internally, how much you want to see from that data? No, because firstly, we've already had the end of phase II meeting with the FDA. Yeah, that's right. I can share a little bit that our intention, as we've guided, is we'll be doing one registrational phase III study. It will be also in adolescents age 12 and up. We will be doing it for a 52-week duration for the primary endpoint. The dose that we'll be using is 24 micrograms per kilogram, dosed twice a month, all the way for 12 months. Through 52 weeks. That's a key element, and that's what we've guided to, and that's our objective. In the first to second quarter of next year is when we'll kick off that registrational program for respeg and alopecia areata. The off-drug data that comes in December of this year, that helps us inform how we'll conduct the long-term extension in the alopecia study. In alopecia, one year and the endpoint, and then patients will roll into an extension study. In the extension study, what we're looking at is do we continue dosing Q2 week, or do we use a lower frequency regimen, such as maybe monthly or so on? In atopic derm, remember, we used that phase I off-drug data to give us confidence to use monthly and quarterly dosing in the maintenance portion of the phase II. In alopecia, we'll use this 24-week off-drug data to inform that same kind of decision and design of the LTE. Yeah. Excellent. Okay. Last minute. rezpegaldesleukin, you have a Q&D with the investigator, and then I think you're also thinking about other indication, including maybe lupus or any others. How much you can say right now for the indications? Yeah. Firstly, the type one diabetes study is being run by TrialNet, they're underway. With the start of the study, multiple sites are open, patients are being screened, there's a chance that we could have data in 2027 from that. That's one item. For us, in parallel with that, we also are exploring other opportunities with respeg, particularly in this period of time while the phase III studies are moving before the first phase III study reads out in the middle of 2028 for atopic dermatitis. We've seen activity in multiple indications, like I mentioned earlier. Other cutaneous indications, for example, cutaneous lupus and so on. We're exploring what's feasible in terms of a study that can be clear enough and interpretable enough, also executed in the correct timeframe. For example, choosing a 12-month endpoint is perhaps not the right indication. Also just sort of mapping those things, and we'll give more guidance on that later this year. Excellent. All right. Thank you, J.D., for joining us, and thank you everyone watching NSC. Thank you.