Arrowhead Pharmaceuticals, Inc. (ARWR) Earnings Call Transcript & Summary

Thanks so much for coming today. We're here for Arrowhead's Analyst R&D Day, and we have a lot to cover. So I'm just going to jump in. This is our safe harbor statement. Please refer to the risk factors in our SEC filings. We will be making forward-looking statements in today's presentations. So we've got a really good panel. Arrowhead folks as well as a few external folks who are experts in their disease areas. We have Michael Benatar from the University of Miami, who'll be talking about ALS and SOD1 specifically. We have Matthias Salathe from University of Kansas, who will be talking about the pulmonary programs, both inflammatory as well as the muco-obstructive program; and Ira Goldberg from NYU, who'll be talking about the cardiometabolic. So like I said, there's a lot to cover today. This is the basic flow. We'll go over CNS. We'll go over some of the expansion that we're doing to the TRiM platform to get to new tissue types. Then we'll have a Q&A session in between each kind of topical area, short Q&A, about 10 minutes. And so we likely won't be able to cover all the questions. We will have a Q&A session at the end of the day. So if you don't get your questions in for the topical areas, just hold them until the end. We'll go into the pulmonary programs. After that, we'll talk about some of the early clinical programs that we have and then we'll go into cardiometabolic, which will include some data as well as our strategy for the clinical development and then some of our strategy on the commercial build-out. And then as I mentioned, we'll have Q&A after that. And then lunch will be available if anybody wants to stick around. So now I'll turn it over to Chris Anzalone, CEO of the company. Thanks.

Well, thanks, everyone, and welcome. All right. well, again, welcome. It's great to see all you here today. We have a saying at Arrowhead that every day matters. And it is an expression of our commitment to patients. Every day that we can shave off development time is a day earlier, we can get important medicines to the patients who need them. It's also an expression of our commitment to pushing science and finding new ways to treat patients and to serve patients. And so I think today is a really -- is a good expression and a good representation of these twin commitments. Okay. So who are we? I think everyone in this room knows who we are. We are an RNAi therapeutics company with a large pipeline of 12 clinical programs, 7 of which are wholly owned fibre ered. These span from early, to mid- to late-stage clinical programs targeting both rare as well as high-prevalence diseases. It's a fast-growing pipeline. We think we can push 2 to 3 new drug candidates into the clinic every year. Everything we do is based on the TRiM platform. We have spent an awful lot of time developing this platform. It's a modular platform. We think it enables us to be potentially best-in-class where we go, and it also allows us to get into new tissue types to follow diseases where they are. And then finally, we have the financial resources, I believe, to push these programs to the patients who need them. An important part of our model is partnering. We've got good strong partners in Amgen, Takeda, Horizon, GSK and Royalty Pharma, and we anticipate continuing to partner some of our programs, now we'll do that judiciously. We will always have a good stable of wholly owned assets, but this partner strategy allows us to bring in nondilutive capital as well as to find homes for those noncore assets of ours. And we have an initiative called 20 in 25, where we expect to have 20 individual drug products either in clinical studies or at market by the year 2025. So we're building a different kind of biotech company. As I said, everything we do is based on this TRiM platform that is structurally quite simple. It's composed of targeting ligands, linker chemistries, stabilization chemistries along the backbone of the RNAi trigger and PK enhancers as necessary. Importantly, we've got libraries of all these components and so allows us to optimize these candidates. So this -- we've created this modular system, this structurally simple system in order to address multiple cell types, again, to go where diseases are. It also provides a continuity or confidence in the platform. We've said this in the past a lot that once we see clinical validation in a certain cell type, we have an expectation of continued success. A cell doesn't care what sequence you're knocking down once we see a well-tolerated ability to knockdown a gene target. We think we can do it over and over again. Again, it gives us the confidence that future candidates should function as planned. And I think that gives us a greater hit rate than is normal in the industry. And then finally, we move quite rapidly from idea to clinic in the liver at least, we've been able to go idea to the clinic in the shortest 12 months, and I think we can continue to operate in this fashion, not only with -- we think it has. In 2017, we had the TRiM platform, but we had 0 drug candidates in clinical studies. 6 short years later, by the end of 2023, we will have brought 18 drug candidates into clinical studies and let that sink in for a second. It is an astonishing accomplishment; in 6 years, 18 brand-new drug candidates into clinical studies. I don't think that we've seen that in the industry before. And so these 18 are addressing different cell types. We will -- by the end of this year, we will have brought candidates in the clinic addressing hepatocytes, solid tumors, pulmonary CNS and skeletal muscle. We are seeing that platform continuity or confidence. Only 2 of those potential 18 candidates have been discontinued. We're treating many people. We've got over 3,500 people and counting in clinical studies with these drug candidates, and we're moving rapidly. We expect to be in 4 Phase III studies by the end of 2023. And so imagine what we can do in the next 6 years. So today's focus will be as follows. We will have updates on some of our clinical programs, the cardiometabolic, pulmonary, C3 and PNPLA3. We'll also be talking about what's next, that will be CNS as well as potential systemic delivery for CNS and also addressing adipose tissue. So here's our pipeline. We'll be talking about ARO-APOC3 and ARO-ANG3 in the cardiometabolic side. ARO-PNPLA-3, our wholly owned NASH drug. ARO-RAGE, the first pulmonary drug that we have data on. We will not be talking about ARO-MMP7, it's too early. We don't have enough data out there yet. We also will not be talking about ARO-MUC5AC. As some of you may know, that is upregulated by -- into the tune of around 30-fold in patients. And so it's much easier to study in patients. I expect that we'll have patient data as well as some MMP7 data towards the end of this year. And we'll also be talking about ARO-C3. So with that, now I'd like to turn it over to Dr. Benatar.

Great. Thanks so much. It's a great pleasure to tell you a little bit about therapy development for SOD1 ALS. So these are my disclosures. I do serve as a site investigator for some industry trials and do some advising -- consulting around trial design and my federal and foundation research funding. So what is SOD1-ALS? ALS is amyotrophic lateral sclerosis, a neurodegenerative disorder. Mutations in the SOD1 gene were the first described genetic cause of this disease. And overall, they account for around 2% of all patients with ALS. And they're found both in patients who have a family history of disease, familial ALS, and also in patients who don't have a family history of disease and what's called sporadic ALS. SOD1 ALS is enormously heterogeneous. There are over 200 different genetic mutations. I mean the SOD1 gene that have been described. In the U.S., one of those, the A4V mutation accounts for about 50% of all SOD1 cases. There is marked phenotypic heterogeneity across the spectrum. That most common genetic form of SOD1 in the U.S. is very aggressive with a median survival of 12 to 15 months. But some of the SOD1 mutations may be associated with survival as long as 20 or 30 years. By and large, mutations in the SOD1 gene produce a toxic gain of function. So although the initial thinking, when this was first discovered, is that this might be a loss of SOD1 mutase activity, superoxide dismutase activity. Turns out the mechanism is overwhelmingly a toxic gain of function through development of a misfolded protein. So where do we stand in the broader therapeutic landscape? You perhaps know that the FDA recently provided an accelerated approval for tofersen which is an SOD1 antisense oligonucleotide developed by Ionis and Biogen. And the idea of this is that it's an ASO, it recruits RNase H and lease the degradation of the SOD1 RNA, and this leads to no production of the misfolded protein. It is not allele-specific and knocks down both wild type and the mutant form. So what do these results look like? And I think there's important lessons here as we think about developing therapies in the SOD1 space. So this is the effect of this SOD1 ASO and clinical function, and this is looking at this outcome measure called the ALSFRS-R, that's the ALS functional rating scale. It's a core functional outcome measure of disease severity in ALS. And the primary outcome was measured at week 28. And you can see the group in red started on drug and the group in blue started on placebo. And although there's a separation of 28 weeks, there is no clinically meaningful or statistically significant difference in the outcome between those 2. When you look at the end of the open -- not the end, but week 52 of open-label drugs, so an early treatment versus a late treatment as the placebo group switched over to active drug, you can now see a clinically meaningful difference and a statistically meaningful difference between those 2 groups. Perhaps even more important than this is the impact on biomarkers. And I'll show you, first, the impact on CSF levels of SOD1, SOD1 protein, which we expect to knock down. And you can see there's about a 30% reduction in the level of the SOD1 protein. This knockdown occurs over about 12 weeks and is sustained out at week 28 and out at week 52. And as patients go on to active drug from the placebo group, we see the same effect. Perhaps even more important and relevant to the regulatory decision is the impact of this ASO on this biomarker called neurofilament light. And I'll tell you more about this, but we see about a 60% reduction in neurofilament light levels. Importantly, that's in blood. This is plasma levels. And you can see once placebo patients go on to drug, the same thing happens. And it's this reduction in neurofilament light that supported the accelerated approval by the FDA. So the FDA reviewed this reduction neurofilament light as a candidate -- or not as a candidate as a surrogate biomarker reasonably likely to predict a clinical benefit. And that was based on the preponderance of evidence that exists in the peer-reviewed literature about what we know about neurofilament light. So what is NfL? Well, the neurofilaments broadly, these are major structural components of nerve cells where nerves degenerate in ALS and other neurodegenerative disorders. They're released into the spinal fluid and into the blood. There are light and heavy chains, there's also a medium chain. But perhaps most relevant is that these are very reliably measured. We have very robust technology and a variety of different assay platforms to measure neurofilament and this can be done, as I say, reliably in blood. So I wanted to just take a sidetrack for a moment and talk a little bit about biomarkers because I think they're very relevant to therapy development in ALS broadly and certainly in SOD1 ALS. And when we talk about biomarkers, we need to think in terms of the FDA's language around thinking about the context of use for biomarkers. And many of these contexts of use are relevant to developing therapies for SOD1 ALS. So I'll draw your attention first to the idea of a predictive biomarker. A predictive biomarker is one which tells you which patient is most likely to benefit from which treatment. And the genetic causes of ALS serve as those predictive markers. So the example here is the mutation, the SOD1 gene, tells us who is likely to benefit from an SOD1 knockdown strategy. We lack this approach in nongenetic forms of ALS, but for the genetic forms, we have those predictive biomarkers in the genotype. Then there are response markers. These are markers that we can measure that tell us that there's been a biological response. There can be -- these can be pharmacodynamic. They can also be surrogate endpoints. I mean the examples I'll give you here are a biomarker that's elevated but stable or a biomarker that's increasing over time as disease progresses. And the FDA's view on this is that if you give an experimental therapy and you can lower the level of the biomarker or blunt that rise, that serves as a response biomarker. And this is what we see with the tofersen data with neurofilament light. Then there are prognostic markers, and NfL has this role as well. So prognostic markers are those that we can measure at baseline or at randomization and are helpful in trial design, typically shortening study duration and reducing the number of patients needed in order to demonstrate a clinically and statistically meaningful benefit. And these are markets that might be relevant to predicting survival or functional decline. And neurofilament light actually is a very powerful prognostic marker and was incorporated into the analysis that were done as part of the tofersen program. And then lastly, I'll say a word about the role of neurofilament light as a risk or a susceptibility biomarker. And the idea here is that this is a market that can be measured in a population and which tells us something about the future risk of developing disease. And what we know is that in SOD1 mutation carriers, if neurofilament levels are elevated and these are presymptomatic individuals who've not yet developed ALS, if their neurofilament levels are elevated, they are much more likely to go on to develop ALS within a relatively short period of time, which gives us a paradigm for thinking about disease prevention. Let me just go back. Yes. Okay. So what then be summarized, we know about neurofilament light as a biomarker for therapy development in ALS broadly and in SOD1 ALS specifically. What I've schematized out here are the rises in neurofilament that we see over the course of disease. This in green is a slower progressing patient, in red is a faster progressing patient. We know that neurofilament levels rise presymptomatically as a marker of axonal degeneration. They continue to rise in early symptomatic disease and then plateau. So typically, by the time we see patients who are affected, they have a stable level of neurofilament that acts a little bit like a spedometer. It tells us how fast disease is progressing, that's why it has prognostic value. And because it's stable, it has value as a response market. If we can knock it down, tells us that there's been a biological response. Let me just shift gears and tell you a little bit about how we might use this in a prevention paradigm, and we're doing this currently with tofersen. So we are taking SOD1 mutation carriers in this trial called ATLAS. These are unaffected individuals. We are following them with their neurofilament light levels. When neurofilament levels go up, if they remain presymptomatic, we randomize them to drug or placebo. And the primary outcome in the trial is the development of clinical ALS, at which point they go into open-label drug. If they develop disease without a rise in neurofilament, they go on to open-label drug. And what's relevant about this is when the FDA granted accelerated approval to Biogen, they identified this trial as hopefully providing confirmatory evidence of efficacy that will hopefully lead to full approval. So what then is the unmet therapeutic need for SOD1 ALS given that we now have an FDA-approved SOD1 ASO? So firstly, in the trial, and we don't yet know in the real world, but certainly in the trial, there was disease progression despite tofersen. There's also the potential for a greater therapeutic effect with more marked lowering of CSF SOD1. I'll remind you in the tofersen program, that reduction is about 30%. We think there may be value in lowering it further but not lowering it to nothing where there may be some potential harms. I think there's also value in developing a therapeutic paradigm that entails less frequent intrathecal dosing. The way tofersen is dosed is there are 3 loading doses, 2 weeks apart and then monthly dosing, and that has to continue forever. So if you're an affected individual and this works or if you're a presymptomatic individual, that would require lifetime intrathecal dosing with the frequency of monthly. And I think over time, that's going to prove to be challenging. What are some of the lessons from the tofersen drug development program that I think are relevant to future trials? I think perhaps most importantly is the value of NfL as a response biomarker that's acceptable to the FDA. That's built upon the foundation of a robust scientific literature. I think there are things that we can learn about trial eligibility, and we can get into this in the Q&A if there's time, how we think about the mutational spectrum, the use, I would say, not of the pre slope of the ALSFRS-R and thinking about disease duration with a premium on treating early. We should also incorporate NfL as a prognostic market prospectively, either via stratification or via dynamic randomization and that will make trials more efficient. And then we need to think about trial duration. I think one of the lessons from the tofersen program is that 6 months may be too short and pivotal studies should probably be longer. So what then are the potential paths for a future Phase III for your compound? Well, I think we have to think about geographies where tofersen is not available or won't be available. We should think about head-to-head comparisons, either noninferiority or superiority design for efficacy and thinking about superiority with regard to safety and patient tolerability. I'll remind you of that monthly intrathecal dosing with tofersen. I think it's probably premature to know how the Phase III should be designed, but the full array of options, I think, is really critically dependent on the results of the initial Phase I study as well as longer-term efficacy and safety data, which will emerge over the coming months and years as tofersen gets into the clinic. And so I will stop there. Thank you very much.

Thank you, Michael. That was absolutely terrific. So good morning. I have the privilege today to share with you the data that the team at Arrowhead has been working on for the past couple of years to develop our new CNS targeting program. And I have to share also, I've been in the RNA therapeutic space for more than 20 years, and this is the most exciting thing in terms of impact to patients that I've ever done. There are more than 50 million people worldwide affected living with neurodegenerative diseases. It's the leading cause of disability and there's almost no therapies available. So on the right side, we're listing some of the most prominent examples of these diseases. These include ALS, Alzheimer's, Parkinson's, probably every single one of you know somebody that's been impacted by one of these diseases. So they have different clinical symptoms and impact different regions of the brain, but they all share a common feature of being caused by abnormal protein aggregation, which results in toxicity to brain cells. Now this has been a very difficult mechanism to drug using traditional modalities. But with RNAi, we can get right to the root cause and knock out the toxic protein. And so this is a really powerful tool to have, and it's coming at a time, which you just heard about of really rapid progress in our understanding of the genetic cause of these diseases and the identification of biomarkers that are really enabling more effective clinical development, really increasing the probability of success. And so it's in that context that we developed the new CNS targeting TRiM platform. So this is just outlining some of the key features. In terms of the design, it is a simplified lipid conjugate. And with this, we can achieve potent target RNA reduction throughout the brain and spinal cord and to all relevant cell types in preclinical models after intrathecal injection. So just as we've seen for the liver and lung TRiM platforms, we have a long duration of action, which can enable infrequent dosing, potentially half yearly. And for safety, we can say that for our first program, we've now completed the GLP tox and our NOAEL was the highest dose tested in both the rat and the monkey. So you just heard about SOD1 ALS. ALS, of course, is a devastating progressive motor neuron disease that can be fatal within a few short years, and SOD1 mutations are one of the most common genetic causes of ALS. And those mutations, again, result in abnormal aggregation of the protein, making it a good target for RNA therapeutics. And you heard about the biomarkers that are available in this space to tell us that we've hit the target, so SOD1 protein in the CSF is used as a surrogate for SOD1 protein reduction in the brain and spinal cord and so we'll know right away that our platform is performing and hitting the target as expected. And of course, NfL is a very powerful marker. And as a reasonably likely surrogate to predict clinical benefit, offers a lot of opportunity for efficient clinical development. And so as the accelerated approval of tofersen a few months ago is a really exciting accomplishment, milestone. It's having -- it's a huge advance for patients, but there's definitely room for improvement. So the limited efficacy with only 30% reduction of SOD1 in the CSF is correlated with a lack of -- they fail to demonstrate clinical benefit in the pivotal trial. So there's a need for better efficacy. And this monthly lumbar puncture is an enormous burden for patients. So longer-acting therapy really has a potential to make a difference. So we think ARO-SOD1 has the potential to achieve both of those things. So let me show you some of the data. This is -- first, we're looking here in transgenic rodent models. So these are rats and mice that are expressing a mutant human SOD1, and they do, over time, develop ALS. And on the left is the transgenic rat. We're looking at simply at a dose response at SOD1 RNA reduction after a single intrathecal dose and this is 1 month post dose. And you can see that we get at the top dose, maximal efficacy 95% reduction in SOD1 RNA. So very potent, deep knockdown. The ED50 or the amount the drug needed to knock down SOD1 by 50%, we calculated 33 micrograms, which is lower than what was reported in literature for tofersen. On the right is the similar experiment in the transgenic mouse. And again, maximum 90% reduction in SOD1 RNA and our ED50 of 13 micrograms is 5x lower than what was reported in the literature for tofersen. So we're getting potent knockdown in brain regions that matter. And if you look in the -- now we're looking at those biomarkers in the CSF that you've heard so much about now in the rat. And on the left is SOD1 protein in CSF, and this is now 3 months after a single administration intrathecal. In this dose range, we're seeing 75% to 90% reduction in SOD1 protein. And on the right, of course, I've mentioned these animals develop disease, so neurofilament levels are elevated. And at this time point, 3 months post dose, we see neurofilament completely suppressed. And so as a marker of disease activity, this shows we're really suppressing disease development. And you can see that more clearly in the transgenic mouse where we've completed a survival study. So here, we're looking at motor function measured by these behavioral tests, grip strength and rotarod and survival. And so here, we've given a single administration to the mice at about 9 weeks old, and we also made the tofersen ASO and put it in head to head. So if you just focus on the left here, the black line is the control group, and we're looking at survival. So what you see is that on average, these mice survive 155 days. The tofersen is the yellow line. It extends that by about a month, but ARO-SOD1 treatment extends that by about 4 months. So the lifespan of these mice went from 5 months to 9 months, pretty significant. You see very similar patterns with the grip strength in rotarod performance, just shifted a little bit in time, about a month's preservation of function for about 1 month with tofersen, 4 months with ARO-SOD1. Now in the nonhuman primate, we've also looked at the activity of ARO-SOD1. And on the left panel, you can see the mRNA reduction throughout the brain -- different brain regions in the monkey at 1 month after a 45-milligram dose, and this is intrathecal lumbar puncture. Now if you just focus on the first 6 bars, you get -- you can -- those are the spinal cord and cortex, those are the region's most relevant for ALS. And you can see 80% to 95% reduction at the RNA level in those brain regions. But importantly, if you look across all of these brain regions, we are seeing really profound 80% or more reduction in many different regions relevant for those other diseases that were on that list that I first showed you. So cerebellum for ataxia, hippocampus for Alzheimer's, substantia nigra for Parkinson's. Even the caudate and putamen, these are deep brain structures that are known to be difficult to reach, we're still able to see 50% reduction. So this tells us that our TRiM platform is ready to be applied to many different indications. So on the right is we also looked at the what cell types we're getting the RNA delivered to -- the siRNA delivered to. So this is a section of a monkey cortex and we are staining for the siRNA. This is an in situ hybridization method. The pink is the siRNA. And you can see that those large neurons are taking up plenty of pink siRNA. But also, you probably can't see it from out there, but the yellow stained astrocytes and the blue stained microglia also have plenty of pink dots in there. And so this is important because all of these cells are known to contribute to the pathogenesis of SOD1 ALS and many of the other indications on that first slide. So we've also looked at the dose dependence in the monkey. I'm a pharmacologist, I want to know where we are in the dose response curve. And again, here, we're just focused on the spinal cord and the cortex and looking in a dose range from 5 -- if we lower the dose to 5 milligrams, we're still seeing really profound deep knockdown in most of these brain regions. So this is very potent in the monkey. And then finally, the duration of action. So we've looked in the nonhuman primate at up to 6 months now. And on the left, we're looking actually at SOD1 protein, not RNA now. So at 1 month, so day 29 here, looking at the spinal cord and cortex, you see we get around 50% reduction in SOD1 protein. But by 3 months, we're getting 80% to 90% reduction of -- at the protein level in all of these brain regions. And this is actually a -- SOD1 is well known to have a long half-life. So it was expected that it would take some time to reach the maximal knockdown level. But importantly, you can see that, that knockdown is sustained out to 6 months. And the SOD1 protein in the CSF which, again, is our translational biomarker that we'll be able to monitor in patients is meant to reflect the protein knockdown in the brain. And it does, you can see it follows a very similar time course. We get about the maximal reduction 2 months-or-so post dose, and it's 60% to 70% reduction is sustained for 6 months. So this compares very favorably with the tofersen published data showing they need with -- it required 175 milligrams administered to achieve 50% reduction. So with this data in hand, we're now moving to our first clinical study. And this is designed as a placebo-controlled single ascending dose in symptomatic SOD1 ALS patients. The primary endpoint is safety. The secondary endpoints will include PK and PD, or biomarkers that we've been discussing in the CSF, the SOD1 protein for target engagement and neurofilament as a marker of response to treatment. We'll have also include some exploratory clinical endpoints, ALS FRS and measures of motor function and lung function. Patients will be monitored for up to 6 months, potentially longer, if necessary, depending on the duration of action. So now that you've seen the data, we hope you're as excited as we are about what we can do with this platform. We think ARO-SOD1 could be a best-in-class therapy for SOD1 ALS with better efficacy and longer duration of action, a better option for patients. So as I mentioned, the GLP tox is complete, and we're now looking forward to a CTA submission in the next couple of months. But we're also looking beyond ALS to the broader opportunity to bring disease-modifying therapies to neurodegenerative disease patients and we are now moving forward a portfolio of programs. So more is coming. Thank you.

All right. Thank you, Christie. Can everybody hear me okay in the back here? Good. So while we're very excited about the work that has gone into the intrathecal program, I just want to highlight some of the additional platform expansions that we're making including a program designed to deliver siRNA across the blood-brain barrier. This is an early stage program. But what we're showing here on the left, this is data in the mouse. With IV administration, we're achieving upwards of 85% knockdown in various brain regions, including the cortex, the cerebellum and the striatum as well as the cord with good duration of effect shown on the right. We've also taken this platform, which is a ligand targeted platform into the monkey. And you can see here the knockdown that's being achieved in the cynos upwards of 72% knockdown in the monkey. Importantly, this is being seen in the caudate, which is a deep brain region that is more challenging to reach with intrathecal routes of administration. We think that this platform could be administered either intravenously or with a subcutaneous injection. Here in the mouse, we're showing subQ versus IV. And you can see that with both routes of administration, we're getting between 80% to 90% knockdown, maybe slightly favoring subQ administration, at least in the mice. So we have a ligand-targeted platform. This is still in early development, but we see at least in animals and ability to deliver siRNA across the blood-brain barrier. We see that this route of administration, either IV or subQ could have potential significant advantages over an intrathecal route if we're able to bring this into the clinic. And we also see an ability to target the deep brain regions that are difficult historically to hit with an IT route of administration. This may be important for diseases, neurodegenerative diseases involving the deep brain such as Huntington's. Another new platform that we've been working on is another ligand-targeted approach to deliver siRNA to adipose tissue. Adipose tissue collectively is the largest endocrine organ in the body responsible for synthesis and secretion of numerous adipokines and cytokines that are involved in not only obesity and type 2 diabetes, but other conditions, including dyslipidemia, inflammatory disease and even some malignancies. Here, we're showing data in the mice. This is targeting a gene that we can measure knockdown in the tissue, but also in the blood. And on the left, we're looking at protein knockdown in the serum, so achieving upwards of 90% knockdown in the blood with knockdown in the tissue at the tissue level of about 88%. So corresponding nicely to what we're seeing in the blood. We've taken this platform targeting the same gene into nonhuman primates. And here, after a single 5-milligram per kilogram dose, we're achieving knockdown -- maximum knockdown of 98% with duration out through week 31, so we can maintain 85% knockdown or better for 31 weeks. The development of this platform is a little bit further ahead than the blood-brain barrier platform. We've taken this into non-GLP tox studies and at a high dose at 120 milligrams per kilogram, we've seen no mortality in rats, no adverse changes in body weight or adverse changes in labs, including chemistry, hematology or coagulation parameters and no adverse findings on histopathology. One other area that we've been working on is the area of dimers. So 2 siRNA triggers covalently linked but each targeting a different gene. And what you can see here, the dimer is in green at the bottom, 6 milligrams per kilogram of a dimer targeting 2 different genes, so target 4 and 5, we're able to achieve deeper and better duration of knockdown when compared to the 2 same siRNA triggers administered separately. So we see a lot of opportunity here for targeting the 2 different genes in the liver, also potentially 2 different genes in the lung. And these are both liver targets that I'm showing here. So I think we're going to take a break for Q&A.

Okay. So we got about 10 minutes or so for Q&A. And there's a lot we covered CNS as well as the platform expansion. So Brian and Jason are passing microphones around. We got some up here, Brian.

Great. Luca Issi, RBC Capital. Maybe a couple of questions here on the CNS. Maybe Christie, can you just talk a little bit more about the chemistry here. I believe Alnylam using C16 chemistry there to improve the lipophilicity. And I think they're using vinylphosphonate to improve potency and actually have tighter binding to the risk complex. So wondering if you're using a similar chemistry here and if you can comment more broadly on that? And then maybe the second one. I understand SOD1 sets a very favorable regulatory bar here, but just wondering the rationale behind going after this indication given the relatively small populations in the United States.

So is this working? Yes. So on the chemistry part, so I think we are also a lipid conjugate but we do have different design features compared to what Alnylam has. But I think we're similar enough that we think there are recent clinical data should read through to us, and we're really excited to see that. And then the other part, you want to say? Well, I think SOD1 is how we've developed the platform. And so I think we've learned a lot about what we need to do, and it's really just -- it's a launching point for us.

Yes, I would agree with that. And I'll add just on the chemistry front, we do have our own similar chemistry that facilitates risk loading. It's not the vinylphosphonate, but so as Christie said, I think we should be similar in terms of potency. And then in terms of why SOD1? I mean we have lots of targets that we're looking at. SOD1 was the first to find a sequence. And it also has a lot of advantages in terms of a measurable biomarker, a well spelled out clinical and regulatory pathway. And we think there's room for improvement still over to tofersen in terms of depth of knockdown and duration of effect. So while there -- the market size and the population in the U.S. may be small, there's still, I think, a lot of ways that we can improve upon what else is out there.

Mani Foroohar, SVB Securities. We'll start on the adipose side. It looks like you're measuring a serum biomarker, which I presume is in adipokine or something else involved in signaling. To what extent does that read across to the intracellular knockdown and efficacy of risk loading inside the adipocyte? There's a lot of steps that up and down regulate signaling from what I presume is ghrelin leptin something like that. Like how translatable is that serum number to what's actually going on in cells? And when can we expect to see something that looks more like an intracellular metric or some more direct measure of knockdown on a single cell basis?

Sure. Yes. I don't know that I can give you guidance on when we would show single cell knockdown and maybe Erik or Tao can cover that if we have it. The first slide that I showed was the knockdown in the blood and then also the knockdown at the tissue level in the protein. So that might get you closer to what you're interested in. And we have the similar data in the monkey, and we haven't disclosed what the target is as of yet.

So I guess the other question is, to what extent should we think about the targeting to different types of fat tissue. Obviously, when you talk about adipocytes or fat tissue, it's -- you go to mitochondria meeting and there's 50 different types of fat tissues they talk about. How should we think about tuning for different types of fat tissue or this is -- or should we think of it almost exclusively as subcutaneous fat targeting? Like how should we think about the nuances of where you're actually getting?

Sure. Yes. So this platform should be able to target all types of fat, both the subcutaneous and visceral fat. In fact, the samples that we looked at in the rodent those were visceral samples. I hope that's what you're asking.

And how do you guys think about target selection for that delivery mechanism in terms of what kind of indications you'd be chasing and where that fits into your broader pipeline?

Sure. Yes. So we have actually a fair number of targets that we're looking at now. We're not ready to disclose those yet, but suffice it to say that I think the targets are in some of those disease areas that I highlighted, including metabolic disease, inflammatory disease and so on.

This is [indiscernible] from Jefferies. So did you compare the neurofilament light chain reduction and by distribution for the MUC tofersen related to the siRNA? And what sort of difference are you seeing there to explain the survival benefit?

I'm not sure I heard the question completely.

Could you repeat that question?

So like the -- did you compare the neurofilament light chain reduction and biodistribution for the MUC tofersen related to the siRNA SOD1 that you used.

I see. In the rodent models you're talking about.

Right. So why do you see the survibal benefit? That's the...

We have that data I think what -- the neurofilament will rebound as the activity of the compound wanes, and that will map very closely with the graphs that I showed, I think, is the answer. Yes.

Great. Ted Tenthoff of Piper Sandler. So when it comes to CNS and obviously advancing towards the clinic with SOD1 ALS. Is the plan here to achieve proof-of-concept initially and then look into other disease areas? Or will you already be starting to prioritize other CNS trends? And as it comes to sort of taking these further, is this an area where you would envision partnering either on a product-by-product basis or even doing a large maybe CNS type deal? Chris?

Sure. So the first question first, are we waiting to see proof of concept in the clinic before we move forward on other targets? The answer is no. You'll hear about more targets later this year. We've got -- we have a pretty good staple of targets, I think. With respect to partnering, I sort of view CNS in a similar way as I view pulmonary. It's a target-rich environment. As we've said publicly in the past about pulmonary, we don't see 2 or 3 drugs, we see 9 or 10 or 11 drugs. Same thing with CNS. And so I think there is room there for some partnering as well as for retaining wholly owned assets. It's hard for me to imagine us not commercializing some staple of CNS drugs. And so we will certainly hold on to some. But again, we are so good at bringing new drugs into the clinic that we've got plenty of ammunition to also bring in the right partners for some of those.

Eliana Merle, UBS. Can you comment a little bit more on the safety particularly anything in terms of chronic tox preclinically of siRNA and any theoretical concerns that you're thinking about for siRNA versus ASOs? And if you can comment maybe any perspective on APP and some of the recent updates in the space there?

So I'm guessing that, that question is specific to the CNS platform. I think all we can say is that based on what the data we have in hand that comes from the GLP tox study where, as Christie mentioned, the top dose was the NOAEL. So we feel like we have a good room to work in terms of margin of safety in the Phase I study. And the -- my understanding is the problem that Alnylam ran into with APP may have been related to frequency of dose in their GLP chronic tox study. I think that it's -- there are pluses and minuses to be in first. And one of the minuses is that you get to figure out things like that the hard way. Suffice it to say that we will be sure to have sufficient duration as allowed by our pharmacodynamic effect in our tox studies.

I think we had one more up here. Madhu? Last question for this session.

Yes, maybe following on the adipose program, what is your line of sight for kind of developing adipose targets? Like I mean, do you think that there's a kind of in the time frame of, say, the blood-brain barrier seeing -- do you have any regauge of where you are in terms of given the predictability of RNAi preclinical drug development, do you think you have a line of sight as when you could actually get an adipose-directed drug into the clinic?

Yes, you can say that we have a target that we've identified, and we're in the process of screening triggers against that target. I can't guide on exactly when those would make it into the clinic. But we do have a program, I guess, is what's important.

All right. Thank you. And we'll turn it over to Erik, who's going to talk about pulmonary. Thank you.

Thank you. Thanks, Vince. I just wanted to take the opportunity to reintroduce the latest generation of our pulmonary delivery platform that we introduced in this meeting a year ago as well as our clinical stage pulmonary programs. So the latest generation of our pulmonary delivery platform utilizes the same algorithmic approach to siRNA design but it employs enhanced modification chemistry to improve the depth and duration of knockdown in vivo. It utilizes the same integrin alpha v beta 6 targeting [indiscernible] to drive epithelial cell uptake in the lung. And what this translates to is that via inhalation, we see increased potency of our siRNAs and target knockdown. And this is related to preferential delivery and uptake of those siRNAs by epithelial cell types in the lung versus nontargeted cell types such as pulmonary macrophage. And this is driven by transient internalization of the delivery receptor. We've looked for any evidence of potential downstream receptor pharmacology associated with receptor internalization such as elements of TGF beta access like smad, protein phosphorylation, we see no evidence of that. So that's a transient event. Also related to the platform, we have confidence that after inhalation, we see good delivery through airway mucus. We know the physicochemical properties of our pulmonary delivery platform is compatible with transit through the mucus. The conjugates are a very small size in order of 3 to 10 nanometers. So an order of magnitude or more smaller than the pore size of the mucus itself, they harbor a net negative charge, which is favorable with respect to potential electrostatic interactions with airway mucus and these conjugates are soluble and aqueous solution. And as we've shown in the past, we have multiple lines of evidence for efficient delivery through the airway mucus both in vitro, in cultures with musin on top as well as various in vivo models of mucus hypersecretion. So moving on to targets. Our first target to speak to today is the receptor for advanced glycation in products or RAGE, and this is a fascinating target for inflammatory lung disease at a very high level. This is a -- the pro-inflammatory pattern recognition receptor highly abundant in lower airways in type 1 alveolar epithelial cells, but expressed at very low quantities throughout the body and other organs. It is activated by a very broad milieu of different pro-inflammatory ligands. So these are sugar modified proteins and lipids otherwise known as advanced glycation in products, a variety of immune cell alarmens, in particular, HMGB1. And signaling through this receptor culminates in canonical NF-kappaB pathway activation and results in secretion from the airway epithelium, a variety of pro-inflammatory second messenger cytokines, mucin secretion and so forth. As we'll hear from Dr. Salathe momentarily, this receptor is very important to amplify and sustain chronic inflammation through other inflammatory pathways. The knockout phenotype of the mice has been well studied and they show striking resistance to a variety of allergic and inflammatory airway stimuli, but despite this very promising biological validation, it's proven very difficult to drug with small molecules. The structure of the receptor is highly related to immunoglobulins. So there are very many potential interfaces for interacting ligands, which, of course, makes traditional small molecule antagonist design almost impossible. We also like this target because it is processed or it's cleaved in the lung where it sheds a soluble version of the receptor or sRAGE into the lavage fluid and into circulation, and that can be measured as a surrogate for target engagement. So what we're showing you here is a summary of some data that was presented at ATS last year. This is our rats that received a single inhaled aerosol dose of a RAGE targeting conjugate at 0.5 milligram per kilogram deposited dose. And in blue, we're following the whole lung gene expression of RAGE. And as you can see, within 3 days of dosing, whole lung RAGE knockdown is greater than 90%, and it remains in this range for multiple months post-dose. Serum sRAGE in the blood detected as a target engagement marker, takes longer to clear as the burden of receptor in the lung is processed in cleaved and cleared and it takes about a month lag to see the full knockdown in serum sRAGE. If you look at the actual receptor itself, at the protein level by immunohistochemistry at day 36, you can clearly see here complete depletion of protein. So moving from there, establishing good knockdown in the rat. We moved to rat models of allergic asthma seeking to phenocopy elements of the knockout mouse here. We use sensitized rats that are exposed to a fungal allergen, Alternaria alternata to provoke lung inflammation in those animals. We can study that lung inflammation in a number of different ways, but simply, we're showing here bronchial or alveolar lavage collections. In response to fungal allergen challenge, these sensitized rats have profound pulmonary inflammation evidenced by immune cell infiltration, so eosinophils, neutrophils. We see elevations in a wide range of cytokines such as IL-13. And similar to what was seen in the knockout mice suppression of RAGE expression by RAGE silencing effectively reduces all of these inflammatory markers, both eosinophils and notably neutrophils. Moving to cynomolgus monkeys in nonhuman primates. Here, we gave a single inhaled aerosol dose of our ARO-RAGE clinical candidate, and we see very good deposition and activity of the drug across all anatomical regions of the lung, resulting in 90% or more silencing with a single inhaled dose. Some new data here that we wanted to share. This is a dose response in cynomolgus monkeys via inhalation. These are single inhaled doses with animal sacrifice at day 29 post dose. We're looking at doses ranging from 0.13 to 0.47 deposited dose. And looking first at lung expression of the target itself, this is membrane in brown, full-length of receptor RAGE, and we see a good dose-dependent reductions with the highest deposited dose resulting in 88% knockdown at the protein level. Now looking at our surrogate biomarkers of target engagement in this case, BAL, sRAGE. We see BAL sRAGE pretty closely match what's happening in terms of full-length receptor knockdown in the lung. However, in the cynos, serum sRAGE appears to underrepresent the depth of lung knockdown of full-length receptor in these animals. So we think at least in the cyno monkey compared to the rat that there may be some extra pulmonary sources of sRAGE that will drive a higher baseline. So moving on from RAGE to MUSIN5AC targeting for various muco-obstructive lung diseases. So to briefly introduce the mucins in the lung, there are 2 major genes that encode gel forming mucins that are expressed to form the mucin gel in the airway that rides on the airway surface liquid. These are very large, heavily glycosylated proteins. And the 2 genes are MUC5B, which is constitutively expressed. It's the major component of airway mucin comprised of 90% or more. And it is required for mucociliary clearance. Knockout of this is lethal, and our siRNAs are highly selective for MUC5AC and do not silence MUC5B. MUC5AC is a minor component at baseline. And at the mRNA and protein level comprising 10% or less, you can see here an airway epithelial goblet cells, there's very little accumulated protein, but under settings of inflammation or allergen challenge. There is a very profound upregulation of this mucin. It accumulates in secretory granules and the goblet cells, but what does that mean for patients? So what that means for severe asthma and a variety of nuclear obstructive lung diseases. There's a wonderful literature in this space for a very challenging target is that if we look at airways and healthy individuals, this is a cross-section cartoon, the predominant component of healthy airway mucin is MUC5B, a very minor component with MUC5AC. But in settings of chronic inflammation such as asthma, there's a much larger MUC5AC component and the biophysical characteristics of mucin 5AC when it's combined with MUC5B, lead to a much stickier and harder-to-clear mucus that results in impaired mucociliary transport and sometimes even plugging in fatal asthma. So the genetic basis of this target was first uncovered a number of years ago in 2015 by Burton Dickie's Group, who made the first knockouts here and a bit of a surprising finding here was that in these animals that produce -- produced no airway MUC5AC is that in these animals they -- these animals were resistant to airway hyperresponsiveness. So if they were given an allergic challenge and then provoked with broncho constriction, they found they were completely protected from airway hyperresponsiveness, showing that it was a combination of the sticky mucin in airways combined with bronchoconstriction that was driving airway dysfunction. So moving to our mouse models of allergic asthma. We're able to study this in the same way. We stimulate mice with house dust mite allergens or IL-13. I just want to point out that in normal mice at the mRNA level or an airway immunohistochemistry, there's very, very little MUC5AC expressed. This can be driven very high with an allergic stimulus, hundreds of fold increase in MUC5AC expression. You can clearly see the burden of this in airway goblet cells. And what we find with our siRNAs is that we're consistently able to get in the 70% to 90% range of silencing of MUC5AC at the mRNA and protein level, and we've translated this through to cynos. But really what's important is the consequence of MUC5AC silencing on airway function. And we've studied this in a large animal model of allergic asthma. These are sheep that are sensitized to ascaris nematode antigen, and they can be provoked into an allergic asthma attack by giving them this inhaled allergen, and what that looks like over time is a very stereotypical set of physiological responses that are seen in humans having an allergic asthma attack. They have an immediate sort of late phase response that occurs 4 to 8 hours post challenge and then a lingering airway hyperresponsiveness that's very difficult to treat. So in sheep, what we're looking at is airway resistance or lung resistance in the hours post inhalation of this challenge, and they start from having very good lung mechanics to 4 to 8 hours post challenge, airway resistance increases by 100%. And in this late phase response, we see a dependency of this response to mucin 5AC because we see a dose-dependent reduction in late-phase response. But critically, the really challenging piece to treat is the lingering airway hyperresponsiveness. So within 24 hours, these sheeps tend to -- their airway mechanics return to baseline but they're still highly sensitive to bronchoconstriction. So if we give them a bronchoconstrictor, so have them inhale carbachol, we can measure the number of breathes those animals need to take in order to get a 400% increase in airway resistance. So this is modeling an asthma attack. So in the sensitize sheep that have not received ARO-MUC5AC, it takes them about 20 breathes of carbachol to get that 400% increase. If they've had that ascaris challenge or that allergen challenge, 24 hours later, without treatment, it only takes about half the number of breathes to get that 400% increase in airway resistance. If those animals have been treated with ARO-MUC5AC, they look identical to their nonchallenged control measurement. So this really highlights the important role of MUC5AC in airway hyperresponsiveness. Dr. Salathe will speak in a bit greater detail here on MUC5AC. We expect treatments for MUC5AC to be broadly applicable to a range of muco-obstructive lung diseases. We're just highlighting here clinical data from patients that have provided induced sputum samples and looking at total mucin concentration in that sputum in healthy people, musing concentration is rather low, but it's almost exclusively MUC5B, as I've mentioned before. And we look at patients from asthma and COPD, all the way up to patients with cystic fibrosis or primary ciliary dyskinesia, there's more and more mucin. But critically, the major increase, the major contributor to this increase in the sticky, hard-to-clear mucus is MUC5AC. And then finally, transitioning to our third clinical program, our newest program. This is a program targeting matrix metalloproteinase 7 for idiopathic pulmonary fibrosis. So this is our first entry into interstitial lung disease. MMP7 is a protease. It's secreted by injured epithelia. It's one of a very large family of proteases with very diverse gene functions. Notably, although it's expressed at rather low levels in the lung, in IPF, it is one of the most upregulated genes in lungs of those patients. And it's, in fact, been used regularly as a validated IPF biomarker. It predicts disease severity and progression. And it also contributes multiple roles to IPF pathogenesis promoting inflammation, aberrant epithelial cell repair and fibrosis. The knockout mice are robustly protected from standard models of bleomycin injury. So that piece comes together there. But despite these understandings for a number of years, matrix metalloproteinases have been really difficult challenges for traditional drug discovery simply because to achieve these effects, you need to have highly selective inhibitors against specific members of this family. And these all share essentially is the same catalytic domain. So that's an almost insurmountable challenge for small molecules. So I'm just going to walk you through just very high-level highlights of data that was presented at our ERS meetings last year. You can find these results on our website in much greater detail. Just simply touching on our work to validate proof-of-concept in rats. We used the rat bleomycin injury model. With this insult, there's a profound pulmonary inflammation followed by fibrosis and that can be read out immnohistochemically, you can see accumulation of lung fibrosis. This can be scored with a standard ashcroft pulmonary fibrosis scoring system. And what we find in rats that doses of MMP7 silencing siRNAs that achieve 50% or greater silencing were highly effective at blocking fibrosis. You can see the fibrosis score of these MMP7 silenced animals, mostly shifting into the moderate to mild. And this was accompanied by significant reductions in airway and lung inflammation, improved pulmonary function and significantly reduced mortality. From there, we transitioned with our ARO-MMP7 clinical candidate into nonhuman primates in dose response studies, given a single inhaled aerosol dose ranging from 0.24 to 1.7 mg per kg, and we see robust anatomically well distributed knockdown of MMP7 throughout the lung and have confirmed the same level of knockdown in cultured human lung slices as well. So I think with that as prelude, I'll turn it over to James to walk us through some of the new clinical data. Thank you.

Thank you, Erik. So just give an update on the various clinical programs, and we'll start with MUC5AC, briefly just an update on the study design. This, like the other programs uses a single and multiple escalating design, starting in the healthy volunteers and then in the asthma patients. In the healthies, we do collect sputum samples throughout the study in both the SAD and the MAD part of the study and then the healthy volunteers in the multidose cohorts undergo broncoscopy predose and then post dose. So the healthies are fully enrolled, and we've moved on to the asthma patient cohorts where we are enrolling patients with moderate to severe asthma. And these patients are currently enrolling. Similarly, in terms of dosing, they get dosed on days 1, 15 and 29. And then we measure sputum after the third dose, no bronchs in the patient cohorts. Key endpoints for the study, safety endpoints include respiratory AEs, changes in lung function. We're able to measure in the healthy volunteers, changes in inflammatory cell counts on BALF. And then we can also look at any changes in chest x-ray. In terms of biomarkers for target engagement, we can measure MUC5AC protein in the sputum and also in the airway swabs. But importantly, as Erik has already highlighted, it's really in the patients, the asthma patients that MUC5AC is upregulated. And the healthy volunteers, the mucin in the sputum is mostly MUC5B, so very low levels of MUC5AC expression in the healthies. Safety so far has looked good with no serious or severe AEs, no dropouts from the study. No AEs due to adverse changes in lung function and no adverse changes in BALF cell count. So no indications of inflammation or inflammatory cell infiltrates on the BALF. The Chest x-rays have all been read as normal, predose and post dose, and there's been no pattern of adverse laboratory changes. So next steps for this study. So far, we've seen a favorable safety profile, and we're in the process of evaluating the MUC5AC samples in both the healthy volunteers and the muco-obstructed patients. And we continue to enroll the asthma patient cohorts, and we plan to enroll 2 additional cohorts that Javier will give some details on in a few slides here, specifically cohorts of COPD that will run in parallel to the asthma patient cohorts. And we've decided to do this because we view this as not only a large patient population, but one where MUC5AC expression is significantly upregulated and also an area where there have not been a lot of new advancements in therapies. The treatments today for COPD or similar to what was used a couple of decades ago, so we view a lot of room to improve therapeutic options in COPD. We're also investigating the possibility of adding some additional studies using novel biomarkers, including changes in air flow based on MRI in both the asthma and the COPD cohorts. Moving forward to MMP7. So this study is still in early days. We have enrolled just the first 2 cohorts in the healthy volunteers, a similar study design with single and multiple escalating dose cohorts in the volunteers. And then after the volunteers are all enrolled, we will enroll IPF patients at doses that are still to be determined. Just to point on the biomarkers that we can measure for MMP7. Similar to the situation with MUC5AC, we can measure MMP7 in the blood and also in the BALF in healthy volunteers, but it's really in the patients, the IPF patients that have this ongoing inflammatory and fibrotic process that MUC5 -- the MMP7 is upregulated. And so we really look forward -- we haven't seen any data, the data from the healthy volunteers in the study yet, but we look forward to the patient cohorts being the most informative in terms of pharmacodynamic effect. And on to RAGE. This is the study that is the most advanced, similar study design with single and multiple escalating dose cohorts starting in the healthiest. The healthy volunteer cohorts here are almost fully enrolled. I think we have a few more at this last cohort down here, B5, that's the highest dose, highest multi-dose cohort is nearly fully enrolled. And then we've started enrollment in the asthma patient cohorts and have only this first cohort, 44 milligrams is fully enrolled. The endpoints are similar in terms of safety, respiratory AEs, lung function, BALF cell count and chest x-rays. However, with the RAGE program, we do have a fairly versatile biomarker, that being sRAGE or soluble RAGE, which we can measure both in the BALF but also in the blood. As a reminder, while we can measure sRAGE, what is most important is actually mRAGE, which is the membrane bound RAGE component. This is what binds. It sits on the surface of the epithelial cell and binds to the PAMPs and the DAMPs, the pathogen-associated molecular patterns and damage associated molecular patterns and trigger the downstream inflammatory cascade in both the type 2 high asthmatic patients in both -- and as well as the nontype 2 asthmatic patients also is likely involved in COPD. What we're showing here on the right, this is data from a rat study, single-dose rat study that demonstrates the relationship of reduction or knockdown in RAGE at the mRNA level with the BALF, the serum and the membrane-bound RAGE. And I think you can appreciate that we are achieving great mRNA knockdown but more importantly, that corresponds to knockdown in sRAGE, corresponds temporarily as well as in magnitude of knockdown and that corresponds to serum sRAGE and membrane-bound RAGE. The point of this all being that we think measuring sRAGE in the BALF and in the serum is a good proxy for what is going on at the membrane level. So these are the data. This is new single-dose data. You can see on the left, after a single dose, the top dose level of 184 milligrams, we're achieving knockdown of 76%. That's mean knockdown after a single dose. So that's the red line at the bottom. We only have data out through day 29 so far. And then on the right, these were the individual participants in the study. So in terms of the individuals, we're seeing max knockdown of 91% after a single dose, again at day 29. After 2 doses, serum sRAGE, mean maximal knockdown of 80% and on the left and then a maximal knockdown of 90% seen here on the right, again, the individuals on the right. Importantly, we're seeing good duration of activity. So after 2 doses, after the second dose, we're getting duration that we think would be supportive of Q2 month dose administration, so good duration. And we still have not seen the 184-milligram data yet, and we'll share those data once they are available. The knockdown here is all with the 92 milligram. So not quite at the top dose level yet for the multi-dose cohorts. Now one of the questions we get asked is, can you knockdown the target in the inflamed lung or the muco-obstructive lung? And we don't have a lot of data yet from the patient cohorts. This is really all we have so far. This is the lowest dose enrolled in the patient cohorts at 44 milligrams. But other than this kind of hyperresponder here in the healthy volunteers that are in blue, I think the patient knockdown data for sRAGE generally overlaps well with what we're seeing in the healthy volunteers. So, so far, I think the answer to that question is yes, it looks like we can knockdown the target in the patient lung. And then finally, the last slide here, this is the BALF knockdown. Again, healthy volunteers after a single dose -- on the left, mean knockdown of 90%. This is from the top dose level, 184 milligrams and maximal knockdown on the right of 95% in BALF sRAGE. The safety profile so far has been favorable with no serious or severe AEs or discontinuations. Like with the other programs, no AEs due to changes in lung function, and no adverse changes in BALF cell count or chest x-ray or adverse changes in safety labs. So in summary, ARO-RAGE has achieved deep and durable reductions in both serum and BALF sRAGE in a healthy volunteer population with similar silencing seen in the asthma patients. We believe this is the first compelling clinical evidence of gene target silencing in the lung using siRNA. The safety profile to date has been favorable, and we look forward to presenting the full data in the healthy volunteers at an upcoming medical meeting. Like with the MUC5AC program, we're also adding a couple of cohorts to this study. Specifically, we'll be adding cohorts enrolling asthma patients with a high baseline FeNO or high fractional exhaled nitric oxide. FeNO is largely a type 2 high measure, it's driven by IL-13. So this should help us better understand the anti-inflammatory effect of ARO-RAGE in asthma patients that have a baseline high FeNO. So now I'll turn things over to Professor Matthias Salathe, who will help put some of these results into clinical context.

Right. Good morning. I sort of go backwards a little bit. I'm going to talk about the targets and how they are relevant in different lung diseases. So what I'm going to look at is both the RAGE and MUC5AC and how I see them to be important in these different lung diseases that we already talked about. These are my disclosures. So let's start with asthma. And I think this slide basically shows you the big 2 targets that we want to go at. Here is the airway in a normal human being. Here is an asthmatic airway. You see inflammation and narrowing of the airway lumen, but also mucus in the middle of the lumen, which makes it obviously much smaller. And this is just depicted in a schematic for the same. So how can we address that? First of all, asthma is a complex disease, right? This is not just a -- asthma is not asthma, it's not asthma. And you see here that there is type 2, T2 high asthma and T2 low asthma. And while we have reasonable treatments, and I say reasonable because they can always be improved, in T2 high asthma, there is almost 0 treatment in T2 low asthma and there is a huge need for that, at least for people who have related T2 low asthma. And this is a significant -- and if you think about here obesity-associated asthma, this is a growing need in the U.S. population, for sure. So here is what is available today beyond the regular treatment of inhaled steroids and bronchodilators. You see the biologics on top here is tocilizumab, which is basically the highest level of interfering with the inflammatory cascade that goes down here, and there is a bunch of other biologics on the market. And when you look at decision trees to make treatment choices for these patients, it becomes actually very, very difficult to make these decisions, who is going on what treatment. So the idea here is really what is on top of this inflammatory cascade and if we can interfere with the top of that inflammatory cascade, then will we be better off than making these decisions down here. And like I said, this is mostly for T2 high not T2 low asthma, where really almost nothing exists. So here, you see an attempt of looking what happens to patients with tocilizumab in T2 high and T2 low asthma. And really, what we should look at is these 2 panels diagonally. Here you see exacerbation risks of T2 high with blood eosinophils at over 300 and FeNO at over 25 that they have a significant reduction in exacerbation with this biologic. On the other hand, if you do not have this, if your T2 low, your exacerbation decline with the treatment is much, much, much lower. So there's huge room for improvement, especially in these group of patients. And these are sort of other intermediate patients in there. So the top of that cascade is really RAGE. And I'm not going to reintroduce RAGE now, but Erik did a great job at that. But you can see it's really stimulated by a lot of different inflammatory stimuli, allergen, smoking, pollution, other these DAMs and PAMs that James mentioned as well. And even they're really originated from advanced glycosylation end products, and that was recognized in diabetes. So sugar, obviously, is an issue there as well. So RAGE, however, is not necessarily expressed very highly in normal cohort, normal people. We can have a whole discussion about what normal means now. But if you see biopsies here from the airway in people without asthma, there is not that much RAGE expressed, but it goes significantly up, the more severe patient's asthma becomes. And that is true also in mouse models that are with allergy treated TDI and there is an increase in RAGE expression that can be at least partially inhibited with a RAGE antagonist. And this is a small molecule, and we all know this is not a great treatment at least so far in the small molecule space. Now RAGE is necessary for type 2 inflammation. You see here how [indiscernible] challenged mice with increases in cytokines. I have a hard time to read it from here, but when you look at RAGE knockout, there is a significant decrease, obviously, in these open bars between the control mice treated with [indiscernible] mice and the RAGE knockout. So RAGE sits really on top of this inflammatory cascade. And if we can block that signaling initiation then we should block a lot of the downstream mechanisms. Importantly, RAGE is also needed for sustained inflammation. So if you initiate an inflammatory cascade, that can be maintained and the maintenance comes from RAGE expression that drives this continuous inflammation even after one stimulus. And in RAGE knockout mice, this is completely blunted. So you can get an initial response, but there will be no sustenance of any inflammation. Now what's also important is in mice models of T-low inflammation, this is an model. There's a lot of neutrophil inflammation here in these mice. So it's not the eosinophilic inflammation, so it's a T2-low inflammation. And you can see in orange, I don't know why you chose these colors, so it's okay. But you see that there is a significant blunting, again, even of the T2-low inflammation in this mouse model with RAGE knockout. And what seems to be important in this area is the inflammasome inhibition. So overall, from these data, at least in these animal models, it seems fairly reasonable to target RAGE as the top of this inflammatory cascade that is not only initiating but also maintaining the inflammation in the airway and seems to work both in T-high and T-low inflammatory responses. This is not only in asthma. So we know that COPD has also neutrophilic response. And you see that the neutrophils here -- I'm sorry, that RAGE is expressed also at higher levels in patients that are smoke exposed or have COPD. There is also a neutrophil relation to this inflammation. So this goes now beyond asthma. Here, there is a mouse model with also smoke exposure with a decrease in the neutrophils with RAGE knockout and the clear decrease in cytokines. So this may be a model also for using this in COPD. Now there is a subgroup of patients with COPD and high type 2 inflammation. This makes about 15% of the COPD population has that. And there is at least one study that was just published in the New England Journal of Medicine, where there is actually a decrease but a small decrease in exacerbations in these COPD patients with a biologic in this case, dupilumab. But this is a small decrease, but the principle there is clear that if you can target RAGE in COPD, you may actually have a much better effect than it's not only in the small subgroup of patients, it might also be in the subgroup of -- no, in the other patients that don't have a T2-high inflammation in the airway. Now RAGE is not only in COPD asthma, here is cystic fibrosis. RAGE is increased in cystic fibrosis in general, and it's even worse in cystic fibrosis patients with diabetes. And as you know, CF patients live now much longer with modulator therapies, but that means also they have more complications as they approach adult age. Currently, it's about 40% to 50% of all CF patients have diabetes, and we expect that with aging of that population, there will be even more. So there is a correlation also with blood glucose and RAGE expression. And here, you see despite modulator therapy, patients who have diabetes have actually still much more accelerated lung function decline than people on modulators with CF who do not have diabetes. So there is an additional opportunity in this space. Now I like this slide because it shows you mucus plugs extracted from a lung of an unfortunate patient who died from an acute asthma attack. And you can all imagine that these mucus plugs are obviously not a good thing to have in your airways because you cannot breathe. And as Erik already pointed out, these are all related to MUC5AC, mostly related to MUC5AC. But what's interesting is why in the world do we have MUC5AC upregulating in there? And from an evolution point of view, it is likely, even though Erik used ascaris challenge sheep, right, it was the ascaris worm that came into the lung to be trapped by mucus plugs and not actually be swallowed again to have the infection with ascaris. But today, this is obviously no longer needed, and we have now problems with this mucus plugging. So mucus plugging is a huge issue. This is true in COPD. You see MUC5B is not really that much changed, but MUC5AC is go up and it is related to the severity of the disease as well. And I come back to that because this is also related to actually how many mucus plugs not only asthmatic patients have, but actually COPD patients have. So now there are a lot of studies that look at mucus plugs in the airway of COPD patients and the higher severity of the disease, the higher the number of mucus plugs. And if you go backwards, there are some notions at least from experts in the field that up to 50% of the lung function loss in COPD may actually be to mucus plugging in COPD. And if that turns out to be true, there is obviously a huge target to prevent further deterioration in COPD, if you can prevent mucus plugging. Asthma has that as well. This is a little bit misleading because this is actually MUC5B over MUC5AC. So it goes down because it's getting worse with MUC5AC. I don't know why [indiscernible] plotted it like this, but this is published. So it's the opposite of what you see here in cystic fibrosis where it also goes up because the ratio is MUC5AC to MUC5B. So in all of these diseases. And then in non-CF bronchiectasis disease, which is a very heterogeneous group of diseases, there is obviously also a MUC5AC problem. So here is me muco-obstructive disease in COPD. And I know this is a bit complicated, and I'll try to quickly go through this. On the left is the first time point of CT where people were looked at with COPD who have mucus plugs in the airways. There is low mucus plugging or no mucus plugging, there is medium and there is high mucus plugging. And this is a year later. And what's interesting about this is that most people who are highly plugged stay highly plugged. And people who are not plugged, they don't really develop plugs. So there always a target group of people, and they're very consistent with continuous plugging at the same location. And in asthma, this is the same, and you can look at that up as well. What you also see is the severity categories are going up with the gold stage of COPD. So the more severe, the more mucus plugging you have and the worse is your lung function. And this is true mostly for COPD, but you see even some smokers with preserved lung function have the same problem. Again, it goes up and the mucus 5AC to MUC5B ratio is also somewhat related to severity. But again, even smoking at that risk has a target for this where you really want to prevent mucus plugging. Now MUC5AC, when you look at mucus -- I'm sorry, airway hyperresponsiveness, you see also that, that contributes to viral exacerbations in COPD. So here are mice data with MUC5AC knockouts and you see their viral sendai virus challenges of these mice has a much attenuated response to the sendai virus, which is important. But also in humans, obviously, you cannot really challenge them with virus. I mean you can, but that's usually not done. These are patients that are having exacerbations to viruses. And you see in their sputum, MUC5AC is elevated as well. So it is also an opportunity to look at exacerbations and potentially prevent the exacerbations. So maybe we should think about airway diseases in a different way. We think about airway diseases now more like a muco-obstructive disease and an inflammatory disease and how these quadrants can be filled in with high mucus and high inflammation. And you can see that these targets, obviously, are very, very good to go after because you can basically treat any of these 4 targets in there. So in summary, I don't want to really read this throughout here. But I think there is a good reason to go after RAGE in multiple airway diseases, not only asthma, but also in T2-low and T2-high airway disease, and this goes to COPD and cystic fibrosis and potentially other diseases as well. MUC5AC is really the mechanism of how mucus plugging and muco-obstructive disease is being seen in asthma COPD and other diseases as well. And going after that target, we'll have a lot of beneficial effect if we successfully can silence that target in the lung. And again, that has no -- at least from our animal data, negative effects because for usual host defense, innate host defense, you need MUC5B not MUC5AC. All right. With that, I thank you, and I'm giving it over to Javier.

Hi, everyone, and thank you, Matthias. Well, what a section this, Erik told us about the platform and the targets and the biology and the preclinical work, chain show, the proof of concept in the clinic. And this is, as he said, the first time that RNAi has been used directly to the lung and is able to achieve significant target engagement. And then Matthias put all of this in context in the clinical context. So my task today is to give you an update of the clinical programs and where we're going from now. But before I do that, I wanted to share why we're so excited about that. And I've been thinking about this since we started all these pulmonary programs in the last couple of years. In chronic pulmonary diseases, there is really 3 fundamental mechanism that cause or are associated with most of those chronic diseases. The number one is inflammation, the other one is muco-obstruction and the third one is interstitial lung disease. Those 3 mechanisms really explain the underlying pathophysiology and/or the clinical syndromes of most chronic pulmonary disease. And what we have right now with our 3 first trigger of drugs is that we're addressing each one of those. So the first one, as you hear, for inflammation is RAGE or ARO-RAGE. And what is important is not just that we pick, I think, a very important target gene, but this drug does the effect in the right cell type where it should be, which is the epithelial airway and not the other cells. So that's number one. The number 2 is the muco-obstructive and the target we selected is MUC5AC or ARO-MUC5AC and then you need to have an effect in the goblet cells of the airway, not necessarily in the epithelial cells, and that's exactly what we're seeing. And finally, and importantly, when we look at interstitial lung disease, the alveolar epithelial cells is the key target that we need to address. And again, the MMP7, I think, seems to be a very validated target. So when I look at all these together, I really get very excited about how we can address with 3 genetic targets, a large number of pulmonary diseases. And I will get back to this concept towards the end of this talk. So now let me get into the clinical programs. And the first one I want to address is RAGE or ARO-RAGE. As you know, we started this program about 2 years ago-or-so, we demonstrate clear proof of concept of target engagement in more than one animal model. We evaluated the pharmacologic effect also in animal models, and we have short-term toxicology that enable our Phase I program. We completed or almost completed a healthy volunteers component of the ARO-RAGE Phase I study. We addressed target engagement as James presented, and we have an initial read on safety that looks good. And also, we're starting the patient portion of this study, and we are completing the first cohort, moving to the next one. And again, the goal here is to really demonstrate the target engagement in patient population start to really identify what is the right dose and the right dose interval. And of course, safety is critical as well. What was the next step. So the next step that you will see this year is the chronic tox data that will be available likely at the end of the year. The proof of concept with regard to the anti-inflammatory effect of the inhaled ARO-RAGE on top of the target engagement, of course. And finally, next year, we will start our Phase IIb study to hopefully enable a registration Phase III program. So as James said, we are adding a new cohort in -- 2 new cohorts in the Phase I ARO-RAGE study. And the reason is because we really wanted to demonstrate the anti-inflammatory proof of concept before we get into the Phase IIb study. So we're adding these 2 cohorts of patients with asthma and high FeNO baseline to really investigate where we have a significant effect. Why we believe that, that will be the case. Well, FeNO is really driven by mainly IL-13. We have significant data in different animal models, both in mice and rats showing that either NOK RAGE or inhibit RAGE production with the RNAi trigger, it really decreased IL-13 or prevent the increase in IL-13 to an allergic stimulus. So we have the preclinical proof of concept that, in fact, RAGE inhibition inhibit IL-13, which is what drives allergic inflammation manifests biocommunity as FeNO -- as a biomarker, sorry. Now the other point that is important is 2 drugs that have been approved biologics and they are effective in treating allergic asthma did show significant reduction in FeNO. So the proof of concept that a reduction in FeNO should translate into clinical benefit has been established with these 2 clinical programs. So this is happening as we speak. The protocol amendment is almost complete. This study is now conducting not just in the Phase I clinic, but in a large number of sites in different countries. We aim to complete this enrollment within the next few months. All right. So where are we going from here? So as you know, the Phase I study will be completed next year. We will have a good idea about the dose, dose interval safety and proof of concept of the anti-inflammatory or anti-allergic effects. So then we need to get into the Phase II study. The clinical development path in severe asthma has been established by the biologic recently, and we will follow a similar pattern. So the first next step will be to conduct the Phase IIb study. One of the key question is what the patient population will be, and we wanted to address both the type 2 high and the type 2 low, and we do that because we're following the biology that was described earlier today. The typical size of a Phase IIb study in severe asthma is about 500 patients. The endpoints are exacerbation, pulmonary function tests with FEV1 and of course, symptoms and quality of life and PROs. Of course, the decision we need to make is whether in the Phase II study, the primary endpoint will be exacerbation or FEV1 and the implication of that might have to do with patient selection and size of the study. So initially, we will focus on asthma with ARO-RAGE, but as Matthias mentioned, ARO-RAGE and inflammatory, of course, has a significant role in COPD and other diseases as well. Now let me switch over to MUC5AC and why we think this is really another important target and why we are prioritizing right now COPD, the first disease that we want to address with this drug? COPD is a very frequent, very common disease, about 16 million people, 9 million of them have a phenotype of chronic bronchitis or hypersecretion. Of course, this is decreased life expectancy and also has significant impact in quality of life and living. The current therapy has been the same for many, many years. LABA, LAMA, so bronchodilators and steroids, inhaled steroids. Recently, as you see, there was the first proof of concept that the strong anti-inflammatory drug can be effective in a subgroup of those patients. So again, these patients have significant morbidity and the clinical trials, again, will be following the current guidance, which is about exacerbation, pulmonary function and patient-reported outcomes as well. So James already presented the -- or show you the key features of the Phase I study. And the news is that we're adding the 2 cohorts of patients with COPD. Patients, particularly with a phenotype of chronic bronchitis and hypersecretion. And this is critically important to us because of all the reasons that you heard today about how to assess mucin and MUC5AC in normal healthy volunteers. So the protocol is complex for normal healthy volunteers because in order for them to produce mucus or to produce a sputum with the sequence of nebulization with saline and hypertonic saline in order to generate more mucus, but even if you generate more mucus in this normal healthy volunteers, it will be driven mainly by MUC5B. So of course, it's not the best model to assess target engagement. And that's why we're adding on top of the severe asthma population the COPD patients. And this protocol amendment is going on right now and we expect to have data on this early next year. And why COPD? Again, is important here, clinically, COPD and particularly those patients with hypersecretion has significant symptoms and impairments in quality of life that are related to the hypersecretion. When you think about cough, when you look at the 3 of the more frequently used patient reported outcome instrument in COPD clinical trials and clinical practice, all of them highlight as one of the key issues cough and the amount of phlegm or sputum that these patient produces and how that impact the daily living activities. So clearly, from the clinical point of view of improving patient syndrome, patient symptoms and life reduce mucus seems to be a very good idea. But it's also important mechanistically, and this was already presented or described by Matthias, but of course, having enough traction in the airway is a bad thing, having accumulation of this mucus harbor bacteria, meaning it's an environment where bacterias can over growth, and that's one of the reason why people with COPD tend to have chronic or repetitive infection that cause exacerbation, that cause progression of the disease, worsening of symptoms and so forth. So this is a cycle, a vicious cycle that the accumulation of mucus is responsible for. And the other thing that's important from the clinical and clinical development perspective is that right now, we have CT scan that can really describe and quantify the amount of mucus obstruction or mucus plug that you can see in patients with COPD or patients with asthma or any muco-obstructive disease. So that helps from the diagnosis perspective. For the perspective of selection, patient population that more likely will be a good candidate for this intervention, but also speak to the severity of the disease because the whole function of the lung is to allow the air goes through get into the alveolus and do the gas exchange. And picture, if you have this type of obstruction, the pulmonary parenchyma distal to the obstruction is not really working. So how critical is to be able to prevent the progression of and muco-obstructive for the future of the disease for the prevention of disease progression and also the improvement of the symptoms. So how we're going to move forward here. We are going to complete the Phase I study. We're going to have the data target engagement and initial clinical read, and then we're going to develop a Phase IIb study in COPD. This has also been well described. We can follow the path that is established already. The endpoints will be FEV1, exacerbation and of course, patient reported outcomes and quality of life. We will focus initially in those patients with hypersecretion. So the chronic bronchitis phenotype MUC5AC. The typical size of the study is about 500 patients. And as you know, there are many of these patients out there. we will effectively learn about pulmonary function, exacerbation and everything that will lead to a proper Phase III study design. So again, I think I mentioned most of these things that Phase IIb study will lead to a Phase III study. But I think what is important is we will eventually address other muco-obstructive disease, and I think that's also a key feature of ARO-MUC5AC, and we'll get back to that in a couple of minutes. And finally, I want to mention ARO-MMP7. Again, this is our first drug that is aimed to address interstitial lung disease. And the first one we want to address is, of course, IPF or idiopathic pulmonary fibrosis because there is a lot of data that documented that this pathway is likely to be very relevant. And one point I want to make that I think is very unique and important is we're targeting the alveolar epithelial cells because that's where this disease starts. In contrast to the approved drugs that are mainly focused on TGF beta or fibroblast actions. So this approach could be synergistic, could be used in association with whatever is the clinical standard of care, and that is really relevant as you think about such a severe disease that is almost universally fatal and the current standard of care even though it has a number of limitations, both based on the efficacy and also in the tolerability due to adverse effect, having a development of a drug like this that address the disease from a different mechanism, different administration path and also that can be associated with what it is right now, the approved drugs of the standard of care. So our initial clinical view is to go into IPF. The Phase I study that just started has, of course, normal healthy volunteers. But again, this is another Phase I study where the patient cohort will be really important because that's where MMP7 is up-regulator. So we expect to see target engagement when we get to the patient population within this Phase I study. But I also want to mention that we started to do the research and investigation to identify other interstitial lung disease where this intervention could be beneficial. So I want to finish how -- to catch up what I started by saying that we're looking at 3 different gene targets in 3 different cell types within the lung. And each one of them have kind of a pipeline of diseases that we can address not just the first one that we're focused right now, which is asthma for RAGE and COPD for MUC5AC and IPD for MMP7. But as you can see, the muco-obstructive component is important in a number of diseases, and we will eventually try to address most of them. So thank you very much. And now I think we will invite the speakers to have the Q&A session.

Thank you, Javier. So we have about 10 minutes or so for questions. And then after that, we'll do a short break.

This is Prakhar from Cantor. So maybe first, a clarification on the addition of high FeNO cohort. Are you seeing more patients on the moderate end of the T2 spectrum in the initial asthma patients and that is why you're adding this cohort? Because if I remember, you had a cutoff of blood eosinophil greater than 200 in the asthma patients. So any clarification there?

Well, the clarification is that this is a new cohort that we're including in order to have proof of concept of the anti-inflammatory antiallergic effect. When we designed the study initially, we weren't focused on that. We wanted to have target engagement and initial clinical risk. So we didn't know -- we decided not to do the FeNO subgroup at the beginning. We realize now that it's really important, it will add value as we're waiting for a much larger and longer Phase IIb study.

And one for the doctor. Where do you think a target like RAGE fits in relative to the biologics that are out there in the context that biologics can be spaced out to 4 to 8 weeks, so maybe the dosing frequency is not that much of a differentiator. Any comments there on the residual unmet need?

So I think there are 2 things. Number one, you can address issues that the biologics are not addressing right now. So that's one of the issues. Number two, the dosing is high dosing in longer intervals, but in the inhaled medication here rather than injections, even though some of them can be done at home, the inhaled medication is easier, and it gets based out much more than that with not higher dosing. So it's what disease are you addressing and where is actually not really a playing field for the biologics right now? And second, the spacing alone is not necessarily a big deal, but it's longer and inhaled.

Patrick Trucchio at HC Wainwright. Just a couple of follow-up questions. The first one is earlier mentioned that the FeNO reduction of 42% to 47% with the biologics. I'm wondering if you have an idea of what the reduction would look like for ARO-RAGE? What could this generate? Is this the data that's expected to be collected with the initial readout? And then also, how does the FeNO reduction correlate to the reduction in exacerbation or other approval endpoints in asthma? And then separately related to ARO-MUC5AC. How are you going to be evaluating the initial target engagement with ARO-MUC5AC and COPD patients? And is that the data you expect early next year?

All right. So the first question, what we expect to see with this ARO-RAGE in comparison with the biologic in terms of the decrease in FeNO? We're expecting to see something very similar. We know that FeNO is driven by IL-13 and the allergic reaction to that. We had a proof-of-concept in preclinical animal, but in 2 different models, inhibition of RAGE clearly suppress IL-13. So the translation of that should be decreased in FeNO. When you look at the magnitude of IL-13, I think it's likely to translate to that. I wouldn't say that it is true across the board or maybe Matthias can speak to that, that all biologics decrease FeNO. But definitely, the 2 that I present here, dupilumab and tezepelumab decrease FeNO significantly in the 40%, 50% range, and that translate into clinical benefit, both in pulmonary function and also in exacerbation. Whether the magnitude of FeNO decrease relates to the clinical benefit, I don't know. But the FeNO is a specific feature of those patients with allergic asthma or type 2 high, and in those patients where you see the maximum clinical benefit on this anti-inflammatory drug. So I think at this point, I feel comfortable with making that translation. But I don't know Matthias, if you have...

Yes, I feel comfortable with that as well. However, you select people with high FeNO, right? I mean it's a special population. And the hope here is -- the expectation is that ARO-RAGE is going way beyond that. But then the question is what are the biomarkers? That's a different question.

And then finally, about COPD and MUC5AC. So again, MUC5AC is happening right in patient with a disease situation. It can be a severe asthma, it can be COPD, bronchiectasis disease or non-CF or CF bronchiectasis disease, so that's a fact, and it's a magnitude of order or more than 1 magnitude further of increasing MUC5AC expression. And when I say that is if you picture the mucus of a normal person, he have a minimal amount of mucin that is produced by MUC5AC. In contrast, in patients with COPD or severe asthma, that's the opposite. So number one, they produce a lot more mucus than a normal healthy lung. And second, that mucus -- component of mucin is trigger MUC5AC. So we are very confident that we're going to be able to describe target engagement in part because also the assay is behaving very well, and we're working on that. So we don't have an issue around the assay. The issues not having samples that have enough mucus and mucus that have enough mucin driven by MUC5AC. All that will be reversed when we study patients with COPD.

Great. I have a few questions. Maybe the first one on alpha v beta 6, can you just talk about a safety risk associated with that? Obviously, Morphic Therapeutics has a paper out there, suggesting bladder cancer signal in their preclinical data. They actually argue that, that is on target. I think Biogen used to have antibody that was discontinued. What gives you confidence that you can actually engage alpha v beta 6 with no major safety issues. That's one. And then two, if I recall it correctly from last year, you're also planning to measure RAGE in sputum. I don't know if I've seen the data today. So wondering if you have any update on that and if you could comment on what you're seeing there? And then lastly, on the relative potency, it looks like you're using higher doses for 5AC versus RAGE. Can you just talk about the relative potency between the 2 triggers and maybe bigger picture, what gives you confidence that you can actually get to a pretty good therapeutic window for 5AC despite higher doses? Sure. I'll start with the platform. The alpha v beta 6 integrin targeting. Obviously, this is a delivery receptor. We're seeing transient internalization of the receptor with periodic dosing of monthly to 2 monthly, every other month, something like that, small molecule inhibitors that are intended to really affect the pathophysiological pathways that might be driven by this receptor require 24/7 dosing, BiD dosing and full coverage there. We are aware of these pathways. We've looked downstream. We've looked for alterations potentially in TGF beta signaling the most sensitive readouts there are phosphorylation status of Smad 2 and 3, we've looked at that. We see no evidence of that changing. So as a delivery receptor, we're quite confident that, that's not contributing to any problems there. I think the next question -- was it clinical question?

Sputum.

Sure, yes. So the sputum -- the question on sputum RAGE. We don't have all the sputum RAGE data yet. I think similar to the situation with MUC5AC, I think rage and the sputum in the healthy volunteers at least in the sputum has been sort of inconsistent. And I think the serum and the BALF, what we're seeing there, both in the placebo and in the active expression is really consistent. So I think those are the best biomarkers we have. And then the other question was on MUC5AC dose. Yes, I mean I think the doses are similar. There is some difference between the 2. And so clearly, with RAGE, we're seeing proof of target engagement at those dose levels, and I would expect that to translate into other programs as well.

Mayank from B. Riley Securities. So good to see first asthma patient data today for ARO-RAGE. You also touched on the NHP data where you showed the BALF flooring relative to serum RAGE. So that phenomenon of comparable pharmacological effect. Could you talk about as you look to generate additional patient data for higher cohorts, any early thoughts on what lowest therapeutic dose could look like and what sort of frequency you could be targeting? And then the follow-up question I had, also in Q4, you will have the chronic GLP tox data for ARO-RAGE. So wondering how that reads into the other programs that you -- either you're -- I'm assuming you're also running chronic GLP tox for other programs also. So how does that read through into other programs?

Yes, sure. I'll take the last one first. I mean I think we'll see what the chronic tox data show. And depending on the results, I think that would not definitively read on other programs. But if we have the safety margins that we'd like to have, that would certainly be helpful for other programs. In terms of the question about the asthmatics and what doses might we see a downstream effect? I think that's why -- one of the reasons why we added the FeNO cohorts to help understand how the knockdown levels at different doses translates into changes in markers of inflammation. And so we have FeNO to measure. We have other markers in the study in the patient cohorts. And then what was the third part of your question?

I think you addressed both. And then just a follow-up for Chris, maybe. As you know, a company like Amgen also had to partner with AstraZeneca to sort of recognize the full potential of tezepelumab. So I'm just curious how you're thinking sort of 2, 3, 4, 5 years down the line of the broader portfolio or at a program-by-program level given the capabilities that you would need to recognize the full potential here?

Yes. As I mentioned, the pulmonary space is a target-rich environment. So we see a lot of drugs coming out of that. And so I do expect that we will do some partnering there. We are not interested right now in partnering in these 3 candidates. We're going to continue to develop these, and we will bring more candidates in the clinic over the next several years. So I don't know what we're going to partner, but I assume that there will be some partnership activity in the pulmonary space, but certainly not everything we can -- we do expect to commercialize pulmonary drugs.

Madhu Kumar, Goldman Sachs. I mean a couple of questions for Dr. Salathe and also for the whole team. First, and the natural question that comes up with all the animal studies is knockout 0, wild-types 100%. Where do you think the tipping point is? Like where is it a level of knockdown that effectively represents complete loss of function of RAGE as evidenced in the kind of efficacy data seen preclinically?

That's always a very difficult question to answer. And I think there are a few considerations. Number one, a lot of RAGE is expressed in the parenchyma, in the alveoli from an asthma and COPD perspective, at least for now, we're not really interested there, right? We are interested in the airway. So if our biomarker is sRAGE, if we don't knock it down in the alveoli a lot, whatever that means, right, we still see sRAGE in the serum even though we may have a complete knockdown in the airway epithelium. And so I think the answer to this can only be addressed with biomarkers and physiological responses. But if you see these knockdown effects in the human beings, this suggests there is a broad coverage and knockdown of the RAGE at least. I can't really answer, but it looks good to me because I would expect that the inhalation is mostly targeting the airway, less so the alveolar space. And obviously, we cannot do -- we cannot really do long [indiscernible], although you could, but that's very invasive. But that will be only the only way. I think the second area is really easier to look at inflammatory responses and physiological measurements. But I'm encouraged by these significant knockdowns.

Okay. A question again for Dr. Salathe, but also for the team. So you mentioned kind of the addition of the FeNO cohort, makes sense in eosinophilic disease, we know FeNO is a good biomarker. How are you thinking about biomarkers in the T2-low population, this neutrophilic population? Something between you can knock down RAGE and you can see a function on clinical outcome. Is there some kind of thing in between you can look at and say, "Hey, we're hitting the target enough to translate into clinical benefit?"

Do I go ahead? Okay.

No. I'm going to set you off for something. I mean you can be invasive and you can do PALs and look for cytokines and inflammation and cells. That's one thing. I also believe that there are now technologies to use, for instance, et cetera, to very low levels to actually measure previous activities, cytokines, et cetera, on a consistent basis. So some of that obviously needs to be developed further. But I think there are noninvasive ways to look at that as well, plus you can do the more invasive look at with bronchoscopy.

Yes, which I don't think we're going to do that in patients, the bronchoscopies, and we can explore some other approaches. Now if you look at the precedent for the biologics for some of the COPD developed drug or the nonasthmatic drug, there isn't really anything life, which is very reliable, correlate with the disease, correlate with the mechanism -- the underlying mechanism of the disease, I don't think that level of biomarker or surrogate exists for the non-type 2 or the COPD population, who will need to be more clinically based and target engagement should be our lead indication to move on to the next level. Now the good news is we're not going to wait until the very end. Right now, our plan is on Phase I to include both populations and to power the study accordingly.

Okay. One last one on the mucus depletion in the MUC5AC program. So one can be the internal optimist of the internal pessimist on how you think about MUC5B and MUC5AC, right? You're right, the MUC5AC goes up a lot more on a relative basis. But there's just physically more MUC5B and if that goes up, you can make it at the math you guys put out there that most of it is going to be MUC5B still even with the increase. So as we think about muco-obstructive diseases, how much of it is about the absolute decline in mucin versus the quality of the mucin that's being targeted, where like depletion of MUC5AC is qualitatively different than depleting MUC5B or just total mucin?

Okay. I'm going to take a stab at that, and I'm not sure how long it will take. But -- so when you look at some of the work, I mean, MUC5AC to MUC5B ratio goes up. But you're absolutely right, 5B goes up as well. And so if you look at COPD patients, there is an increase in the mucous solids, meaning the combined mucus in that sputum as well. The problem with MUC5AC is that it's much more sticky. It sticks to the airway epithelium itself. When you look at David Earls paper that was mentioned once in the JCI, it actually doesn't even secrete fully. It's adherent to the cells and therefore, has this plug capability that much less of an issue occurs in MUC5B. So the idea is that if you don't have this plugging, you also have secondary inflammation. So if you can prevent to MUC5AC, you may still have a MUC5B elevation. That is a true statement. But it should be, and I say should because that's difficult to measure, it should be clearable. And the mouse is not really a good animal model for this because mice clear sort of with air liquid interface without even coughing, et cetera. But if you don't have a sticky mucus that is adherent, you may also have a better cough mechanism to clear it out. So that's the theory behind it. And if you prevent mucus plugs, you get a decrease in inflammation as well, which helps you overall in terms of the MUC5B elevation as well.

Great. Just a quick question. You mentioned earlier the dual trigger strategy in potential lung. Does it ever make sense to combine RAGE and MUC5AC or are these different enough patient populations where one or the other would be better served going after the specific diseases things.

So as you saw, I think in my last slide, we're going to target both, let's say, asthma and COPD to DC with both drugs. Now that it's proved that an anti-inflammatory approach with biologic might help people with COPD or patients with COPD. And we think that muco-obstructive is a significant component. So both drugs will -- might be developed for the single indication. And I think the other concept clinically that is evolving and Matthias presented is that quadrant that shows maybe the patients in the future will not just be divided by asthmatic or COPD, but based on the phenotype with regard to the immune component and the muco-obstructive component of the disease, and that could help clinician to decide which is the best approach. Of course, there is overlap. And of course, today, both diseases are treated with multiple drugs that do different things. So it's not unthinkable that at one point you can combine whether that will be part of our long-term dimer strategy, I don't know now, but it could be because it's a really good point.

And actually, just to be clear, the dimer data presented today was hepatic delivery not lung delivery. Okay. We're running a little bit behind, so let's take a quick break. We're going to do 10 minutes, but let's come back in 5. So let's be back at 11:26-or-so. Thank you. [Break]

We're just about to get started, so come back and take your seats, please. So we're going to bring James back up to talk about some of the earlier programs, the ARO-C3 and the PNPLA3 programs.

So we'll just briefly review some of the earlier-stage liver programs. First, starting with ARO-C3, which we're developing for complement-mediated renal diseases. Specifically, we're focusing on C3 glomerulopathy, which is a rare disease. There are only about 50,000 or so of these patients globally. However, half of these patients progress to end-stage renal disease within about 10 years. And while transplant is an option in these patients, the disease actually recurs in the transplanted kidney in about half of those patients. There are no approved therapies for C3G. And we think that C3 is really an ideal target for this disease because it's the accumulation of the downstream C3 split products in the setting of overactive alternative pathway that really drives the pathogenesis of this disease. So if you sign with C3, you should be able to prevent the formation of those split products and prevent the downstream damage to the glomerulus. Similarly, IgA nephropathy is another area of interest. This is not a rare disease. There are about over 1 million patients globally, and this accounts for about 40% of all cases of glomerulonephritis. This disease is also largely driven by over-activation of the alternative pathway of the complement system, and this is supported by biopsy evidence, also the increased amounts of C3 split products in the blood. That's actually prognostic in IgA nephropathy. So high split product level means a worse outcome in these patients. And then also genetic studies where variance -- genetic variants that increase alternative pathway activity have a higher risk of developing IgA nephropathy, and those that decrease alternative pathway activity actually are protected from IgA nephropathy. Importantly, we think that urine protein reduction could be a pathway for accelerated regulatory approval in both of these diseases, and there's some precedent here for using proteinuria as an approval endpoint for accelerated approval at least in IgA nephropathy. This is some of the data from the Phase I study. This was a healthy volunteer study of ARO-C3. And you can appreciate on the left, after a single dose, we're seeing 82% mean reduction at the top dose level. Again, this is C3 in the blood in healthy volunteers with good duration, duration of about 16 weeks after the single dose. And then on the right, after 2 doses, up to 88% knockdown with also a duration through about week 16. These levels of knockdown translated nicely into measures of alternative pathway activity such as Wieslab AP. And with the top dose level, we're getting up to 99% reduction in Wieslab on the left; and about 91% reduction in AH50, which is a measure of hemolysis, on the right. So far in this program, the healthy volunteer safety profile has looked good with no SAEs or dose-limiting toxicity. Importantly, we've seen no evidence of infections with encapsulated organisms, which is important for complement inhibitors. And the most common AEs are headaches, upper respiratory infection and inflammation at the injection site. In summary, ARO-C3 has achieved knockdown of up to 88% in the blood, and we think the duration can justify quarterly or less frequent dose administration. The safety profile has been favorable. And the knockdown that we're seeing has translated nicely into reductions in measures of alternative pathway activity that we think are competitive with other alternative pathway targeted therapies that are either approved or in development. Additionally, we think that the small infrequent injections that we can use should have advantages over other alternative pathway inhibitors that require large doses, large subcu infusions or even BiD oral dosing. The patient cohorts are open in the Phase I/IIa study. We have cohorts that are enrolling patients with C3G and IgA nephropathy. Moving on to PNPLA3. And so as you know, this is the program that we initially licensed to Janssen as part of our hepatitis B deal we did with J&J several years ago. And this program will be coming back to Arrowhead due to strategic decisions at Janssen, not due to any issues or problems with the drug, and we'll share some of the data here in a few slides. But first, just to touch on the target, so we really like this target. We think this was a great siRNA target. We thought it was a great target when we worked on this program years ago and did the licensing deal. Normally, PNPLA3 serves as a lipase. It sits on the surface of the lipid droplet in the hepatocyte and metabolizes triglycerides. In the setting of the I148M PNPLA3 variant, this variant codes for a nonfunctional protein that's resistant to proteasomal degradation. So it just kind of sits there on the surface of the lipid droplet. It doesn't do much other than block other functional lipases that -- whose job it is to metabolize the triglycerides in these lipid droplets. And so the lipid droplets just kind of get bigger and bigger. This leads to hepatic steatosis, downstream inflammation, eventually NASH and liver fibrosis and cirrhosis. Since this is an intracellular target, it's difficult to modulate with an antibody, also difficult to hit with a small molecule because it sits -- again, it's an inactive protein. It doesn't have any active site or cleft or receptor for a small molecule to bind to. For these reasons, we think it's great for an RNAi approach, and we're not alone. There are several other competing programs that are in development as well as the ARO-PNPLA3 program. The I148M variant is associated with not only increased liver steatosis that you can see here in the black bars, these are the homozygous patients that have fattier livers than the individuals who don't carry the variant or even compared to the heterozygous individuals, it's also associated with increased ALT and an increased risk of not only NASH, but also NAFLD-related hepatocellular carcinoma. And then one of the most striking figures here, I think, is the increased risk in liver disease-related mortality in the homozygous here at the bottom, so the big gap on the Kaplan-Meier curve compared with even the heterozygous individuals that are closer to the top. As we all know, there's a lot of NASH out there. NASH is, of course, very common. What is probably less well known is that this variant is also common, about a 45% allelic frequency. And so homozygosity, as based on this common allelic frequency, it's pretty common to find homozygotes. The -- there are about 12.5 million of these patients. Homozygous PNPLA3 I148M variant patients with NASH, about 12.5 million of these patients in major pharmaceutical markets and about 4.5 million in the U.S. So this is not some orphan disease, rare disease indication. This is really a pretty sizable genetically defined population that should be amenable to an RNAi approach. This is the study that Janssen conducted. They had plans to do a single and a multiple escalating dose study. They stopped after the SAD component of the study. They enrolled NAFLD patients with high liver fat, so 8% or greater. They have a confirmed genotype, either heterozygous or homozygous, and they enrolled those separately. So they had heterozygous cohorts for the PNPLA3 variant and homozygous cohorts. And then the ALTs were relatively normal in the study. They excluded anyone with ALT at baseline greater than 1.5x, upper limits of normal, and they excluded anyone with baseline liver fibrosis. Baseline characteristics here. I think the key takeaway is the level of liver fat in the homozygotes compared to the heterozygotes. And then at least in the U.S. population, the variant is very common in a Hispanic population. So this is the key data takeaway. After a single dose at the top dose level, they were seeing reductions in liver fat using MRI-PDFF of up to 40%. And at the lower doses, you can see about a 20% to 30% reduction in liver fat from baseline that has good duration. So we're still getting the data from Janssen for the highest dose cohort. But at least at the lower doses, a single dose reduction in liver fat that's of 30% that lasts 24 weeks is pretty good. And so we would expect to see or hope to see the liver fat reduction at the 400-milligram dose level in the homozygotes extend out here through week 24, but we'll have to get those data. In terms of safety, there were really no issues in the study, no adverse changes in triglycerides or in LDL-cholesterol, no severe serious AEs or adverse changes in labs or vital signs, and most of the AEs were mild. Importantly, in this Phase I study after a single dose, although with long duration of effect, there were no -- not a signal around GI toxicity or GI AE. So not a lot of complaints of nausea, vomiting, the types of things that are commonly being described with some of the other NASH programs in later stages of development. So we look forward to continuing the development of this assay. We're currently working on the design of a Phase IIa study that would enroll patients with NAFLD and high ALT at baseline. And of course, we could measure them after multiple doses, changes in not only liver fat, but also changes in ALT and changes in other noninvasive measures of fibrosis and NASH activity, so things like FibroScan, ELF, Pro-C3. And then if this study looks favorable, if the results from the study are favorable, we would then progress to more of a larger classic NASH Phase IIb study, a larger study with the typical histologically driven NASH endpoints. And so we'll pivot here from the early-stage liver programs to the cardiovascular programs, and I'll hand things over to Professor Goldberg.

So I'm having a little hoarseness on my voice, but I think I'll do fine. So I want to tell you a little bit about the kinds of things I see in the sort of lipid universe, something about an acute disease and then a bigger picture about lipids and cardiovascular disease and what people who do what I do are thinking. So we see somebody like this down on 34th Street about twice a month. Somebody who has triglyceride levels that look like cream or about 8,000 milligrams per deciliter, 8% fat. They get this finding called eruptive xanthomas. And if you look in their eyes, their bud looks like cream of tomato soup. And these people get admitted to the hospital with an inflammation of their pancreas called pancreatitis. And they spend anywhere from a few days to weeks in the hospital. It has a number of complications and is associated with mortality. It is the third most common cause of pancreatitis in the United States and is not a rare condition since we see it a couple of times every month. So where is this fat or triglyceride coming from? So it comes from 2 places. It comes from the fat that you eat. It gets reassembled in your gut, goes into your blood and then circulates. It also comes from your liver where it's basically re-put together from either fat that's made in your liver, which is adopted from carbohydrate that's converted to fat or from lipid that returns. Both forms of lipid that circulate in particles, either in chylomicrons microns from your gut or VLDL from your liver, require the same enzyme. This enzyme, lipoprotein lipase, basically degrade the triglyceride or liberate it and allow the fatty acids to get into tissues. So you could see if you have a defect in this enzyme, your triglycerides go up and up. And if you have a genetic defect, they can be very, very high and cause a pancreatitis syndrome. Also, this is the key to preventing triglycerides from getting up. If you get this enzyme reaction to work better, triglycerides will drop. So there are 2 approaches now to lowering triglycerides, both of which really focus on that lipase enzyme. One is an inhibition of protein called APOC3, a small circulating protein that's on VLDL, chylomicrons and HDL on a lot of lipoproteins. And it's long been known to be an inhibitor of lipoprotein lipase. It turns out it also has other actions that block the uptake of lipid particles from the liver. The second approach has been to inhibit an inhibitor. So there's an inhibitor predominantly made in the liver called angiopoietin-like protein 3, or ANGPTL3, that works as a complex in your bloodstream to actually block lipoprotein lipase. So the other approach has been to inhibit the inhibitor. This protein also looks like it has other actions. And the other actions have to do with reducing LDL particles, and I'll mention that in a couple of minutes, especially in the setting where the normal pathway, the LDL receptor pathway is defective. Okay. So the C3, the story of APOC3 actually came from a genetic study that was done now I think it's like 15 years ago down in the University of Maryland, where they studied Amish in Lancaster County, Pennsylvania. The investigators found that there were Amish people who had very low levels of triglyceride, very low incidence of cardiovascular disease, and they had a defect in this protein called APOC3. So very quickly, a number of companies jumped on this, and they made ways to inhibit the APOC3 with the idea that they would inhibit the APOC3 and triglycerides would come down, which should exactly happen thinking that APOC3 was the inhibitor for lipoprotein lipase. Then believe it or not, as somewhat of a control experiment, they actually reduced the APOC3. This was done by Ionis. They reduced the APOC3 in patients who had enzymatic molecular defects in the lipoprotein lipase. So it proved that C3 actually had another role, and that other role was to block lipid uptake in the liver. So even if you don't have any enzyme, this therapy works and will lower triglycerides, as you can see, about 80% in people with the rare genetic diseases where the usual therapies do not work. So these are rare patients. And there are rare patients with the molecular defects in lipoprotein lipase or the other components of that system. More often, the patients we see with the really high triglycerides who come in, the more common things are somebody has diabetes and they have a heterozygous defect, and you put diabetes or obesity or some other condition on top of the defect and their triglycerides will go up very high. And those people are more common, the patients we see who come in with the high triglycerides in the pancreatitis. The rare molecular defects, well, they come to me, but they are not generally what people see. But the bigger issue for triglycerides is that almost 30% of the adult population has some degree of hypertriglyceridemia. And is that an issue? So for over 50 years, since the time when I was a medical student, people would ask this kind of question. They would say, how important are the triglycerides for cardiovascular disease? And some of it was driven by things like this, this paper that came out in 1973 from a group in Seattle that raised the question that triglycerides were as common an abnormality as high cholesterol in people with cardiovascular disease, at least in Seattle. This paper, by the way, was written by Joe Goldstein, who went on to win a Nobel Prize for describing the LDL receptor. But subsequent to this, there have been studies looking at how high your triglycerides go after you eat, so called postprandial hypertriglyceridemia, just triglycerides in cardiovascular disease. And more recently, genetic abnormalities, all of which have raised triglycerides are associated with more cardiovascular disease. So what's missing? What's missing is an intervention trial showing reducing triglyceride reduces cardiovascular disease. So if you followed this literature, this is the most recent attempt to do that kind of trial. So this is a trial called PROMINENT. It used the new form of fibrate called pemafibrate. And this paper came out last fall in New England Journal. They looked at the right people. They were people with high triglycerides, many people with diabetes. I think most of them had diabetes. And what they found was even though the therapy reduced triglyceride levels, it did not affect cardiovascular disease. In fact, it was equipoise. Whether you're on the drug or you are not on the drug, you had the same incidence of cardiovascular events or so-called MACE. But what did this do? This trial, what it did is it showed no overall change in cholesterol, no overall change in ApoB even though triglyceride went down. So what this trial did is it converted the triglyceride particles called VLDL into the cholesterol particles [ called ] LDL. And it was equipoise because this trial proved that the VLDL and LDL were equally atherogenic, equally atherogenic. And the problem was that the LDL went up during the trial. So this is not the way to do it. Fibrates are not the way to do it. So this may be the way to do it, we'll see. So you'll hear more about this trial from Javier. So this is what ARO-APOC3 does, but lowering APOC3 in general will do this. You could see a marked drop in triglyceride, 220 to 59. But also LDL does not go up. Actually, it goes down and ApoB goes down. So this is a different kind of approach and doesn't have the issue of converting one atherogenic lipoprotein into another. Okay. The second way to drop triglycerides is to use this other approach, ANGPTL3, and they're silencing RNAs in the antisense that work for this as well as monoclonal antibodies, and you'll hear a little bit about this. And so you have 2 different approaches or 2 different kinds of therapies. The C3 works better in the rare people with the molecular defects, though, in the lipase cascade. What else is unmet? So mostly, we can take care of cholesterol pretty well. We put people on statins, ezetimibe, we now have PCSK9. Except for the person like this in the middle, who is a young person who has a homozygous loss of LDL receptors, and they get cholesterol levels of 500, 600, 800, 1,000. 1,000 is not as high as they go. They get plantar xanthomas, that's those things on their hands. And they could get myocardial infarctions at 15, 12, 5 years old and bad aortic valve disease. We don't really have a way to treat them. They're treated with like a dialysis system. When I was at [ Comrie ], we used to send these kids for liver transplants, which actually does work. So what are they doing? They're missing the receptor, the LDL receptor, and that's why the drugs that mostly require up-regulation of the LDL receptor that's had statins work in PCSK9, if you have no receptors, they don't work, but ANGPTL3 does. And it's not clear why, but it's clear that this particular molecule activates some other pathway that we don't know yet, which is an alternative LDL removal pathway. So this is a little bit of an unmet need. The last thing I think you should know about -- I went back. I got to go backwards one. There we go. The last unmet need is this molecule called Lp(a). It's a risk factor. It's a small extra molecule stuck on to LDL. It's a potent risk factor. It's made in the liver, and it will be treated with the liver approach. So what am I trying to quickly tell you? C3 inhibition treats familial chylomicronemia syndrome, something we did not have therapy for before. Unlike fibrates, C3 does not -- C3 inhibition lowers triglyceride, does not raise LDL. So it may give us a different approach to try to answer the question, how important are triglyceride particles? ANGPTL3 is another approach. And Lp(a) is another thing that we will be treating probably in the near future. And I'll stop with that. And I'll let Javier tell you about the programs here.

Thank you. So I want to start making a comment and similar to what Chris did at the beginning, saying that in 6 years, we brought about 18 molecules to the clinic, which is unheard of and probably this is the only example. In the clinical development group, we took 2 molecules in 3 years from Phase I to either Phase III or enable Phase III in 3 other indications. So 2 molecules, 4 indications, ready to get into final registration model within the last 3 years. So I'm going to give you an update of the cardiometabolic program, both molecules ARO-APOC3 and ARO-ANG3. You are familiar with these studies. The PALISADE study, the FCS study, we communicated a few weeks ago that we completed enrollment of that study, 75 patients. So it's a 1-year study, and we are already working on registration model in many different aspects of what it would take to get to that point. The SHASTA-2 study or ARO-APOC3 in patients with severe hypertriglyceridemia. Of course, we completed enrollment last year. And we are -- I'm going to present now the final result up to the primary endpoint that was at 24 weeks. So all 100% of patients complete 24 weeks in this study already. The new study on mixed dyslipidemia is similar. We completed the enrollment, and we completed the week 24 data, which is the primary endpoint, and I will present this data today as well. And then the VISTA program, which is the ANG -- or the ARO-ANGPTL3, HoFH, the study -- Phase II study GATEWAY is underway, but we already presented, I think it was last week or the week before in Europe, the result of the Phase II study. And again, we're already thinking about how to get this drug to the patients that need them. And finally, the ARCHES-2 study in mixed dyslipidemia with ARO-ANG3, again, the study is complete. I will share with you the data and how we're thinking about moving forward. So I'll start with FCS. And I would remind you something that you probably already know, which is with this study, 4 patients with FCS in the Phase I study 1001. So it was a small group, only 4 patients. And then as you can see, these patients received 50 milligrams of ARO-APOC3 at baseline and at week 4, and they achieved an about 85% reduction in TG. A very similar magnitude of reduction that you see on the other group of patients that also has severe hypertriglyceridemia, but now the genetic component of FCS. So we do have, in this case, I think a significant level of proof-of-concept based on the mechanism of action and the initial data that we developed on the Phase I study. On the right part of this slide, I wanted to share with you the baseline characteristics of the FCS study that is fully enrolled. Patients in average were about 45, 46 years old. About half of them are men. The median TG level, about 2,100, and that's expected. Genetically or clinically confirmed, of course, all of patients in order to qualify for the study had to be in either genetically confirmed, where that 44 of the 75 are genetically confirmed, and then the others are clinically confirmed. And this is important because this was the outcome of our interaction with the FDA when we submit the Phase III protocol in which initially we were focused only on patients with genetically defined FCS. And the feedback was like there are a number of patients who have clinical feature of FCS, based very high TG levels and preexisting history of pancreatitis their self or family members. And those patients should be considered FCS if that's the clinical feature because some of these genetic mutations might not be known. So we genotype all these patients. And when we get to that point, we will disclose that information on how the data looks like. And also importantly, and this back to what Dr. Goldberg told us today, the instance of pancreatitis and severe pancreatitis in this population is very high. And thus, the goal of therapy, reduce TG levels and prevent the event of pancreatitis. So now I will switch over to the SHASTA-2 study. So ARO-APOC3 in patients with severe hypertriglyceridemia. We enroll in this study patients with triglyceride levels greater than 500 at baseline, but no greater than 4,000. The key endpoints, of course, are all the lipids, target engagement assessed by APOC3 and mainly triglycerides and, of course, safety as well. We randomized the patients into 3 groups: 10 milligrams, 25 milligrams and 50 milligrams at day 1 and at 12 weeks after the initiation of the study. And of course, we compare this with a group of patients on placebo for each of these subgroups. The baseline characteristics are as expected. Patients in average, when you include patient with more than 500 of TG, the average is about 650 or 670, and it was well balanced across the different treatment group. These patients tend to have low LDL cholesterol in the 60s, as you can see here. And again, all the other parameters are consistent. There were more males. In this case, about 80% were males, 20% females. And the BMI is in the upper end of overweight versus obesity. So here are the 2 key endpoints on this study. The first one is ARO-APOC3, and you can see some level of dose response. And the first assessment was at week 4, and you already see a very significant decrease in APOC3, achieving the nadir after the second dose at week 16 with a reduction of about 76% for the 10-milligram dose, 86% for the 25 and 87% for the 50 milligrams. So it's very consistent, the effect of 25 milligrams and 50 milligrams with regard to APOC3 reduction. And the same is true for the triglyceride reduction, which you see already a very profound effect with a dose as low as 10 milligrams, and then we see an increased efficacy at the 25- and 50-milligram dose. The table here in the bottom of the slide shows the baseline level of TG, that was about 670 to 700. And the post-baseline at week 16, that went from 696 to 150 approximately. And that was even more profound in the 25-milligram approaching 114; and even more in the 50 milligrams, 94. The normal level of TG, and I'm not talking about people with severe hypertriglyceridemia, but in general, it's lower than 150. So the average patient or the average response put this patient with very severe hypertriglyceridemia into the normal range. And another way to look at the magnitude of the efficacy and how we will translate into a clinical benefit is to look about the proportion of patients who at baseline have very severe hypertriglyceridemia. And in order to simplify this, we selected those patients who at baseline have TG 80 or greater, which is the definition that people use in Europe, and we wanted to use that to highlight that these people have very high risk of pancreatitis. And if you look at both the 25- and the 50-milligram dose, at week 16, almost all patients are below the threshold at which pancreatitis occurred. So this is the clinical benefit that we want to demonstrate. And there is not too many examples of a treatment that can achieve this type of efficacy that hopefully translate into the improvement in clinical outcomes. With regard to safety, and we presented this last year, there is really AEs that are related to this patient population and comorbidities. There is a very well balance between placebo and treatment groups. All the SAEs were not related to the treatment, and there was no patients who died in this study. The finding that we also reported last year is that in the 50-milligram group, the small group of patients who have uncontrolled diabetes at baseline, the inclusion criteria was up to 9% of HA1C. So there was a number of patients with fully controlled diabetes, some of them have worsening on that diabetes or in A1C diabetes control. Now what we did with that when we came about this finding last year is we amended protocol because we noticed that there was no cases that investigator adjust the diabetes treatment. So we adjust the protocol to create flags that go to the investigator and say, you have a patient that have an increase in A1C more than a given threshold. I mean please consider to do something to improve diabetes care. And that was twofold: one is assure and compliance with treatment or adjust the approach to treat diabetes in that particular patient. Once that happened, most patients decreased the HA1C. So I think that's a very reassuring point with regard to this finding we reported last year. So how we're thinking about the Phase III program? And we're -- right now, we're still working on the briefing document to engage with the FDA and the European Medicines Agency to define our Phase III registration program. We're thinking about doing 2 Phase III studies, the SHASTA-3 study that will be really similar to the Phase II study, so patients defined by TG greater than 500. We aim to enroll about 600 of those patients and to have a randomization ratio of 3:1. Why? Because the main endpoint in this study will be TG level and the safety aspect of the drug. So we want to enrich the number of patients on treatment to reassuring about the safety aspect of this drug as we move forward. But it is also very important to address the clinical outcome in this condition, and that is prevent pancreatitis. So we're going to do a second Phase III study that will be enriched for patients who have high risk of pancreatitis defined by having a TG level greater than 880 and a past history of pancreatitis. The fine-tune details of that is still underway. But the approach to this that is important is the primary endpoint for those studies, which is TG, will be done at 6 months, and that follows regulatory guidance. So we had the opportunity for an early file, but also this program will document the clinical benefit of this intervention, both in prevention of pancreatitis and likely to improve quality of life and symptoms in patients with SHTG. All right. Now let me move on to the new study or the ARO-APOC3 study in patients with mixed dyslipidemia. You also are familiar with this study. We enrolled patients with baseline TG between 150 and 500, LDL cholesterol greater than 70 milligrams per deciliter and non-HDL greater than 100. All these patients had to be on optimal steady statin therapy. So we have a running Phase II reassuring that, that was the case, and all of those patients were on statin and properly treated. Key endpoints are the typical PV biomarker related to lipid profile and, of course, safety. We randomized patients into 4 groups here. One group received 10 milligrams at baseline on week 12. Another group, 25 at baseline on week 12. And the third group, 50 milligrams at baseline on week 12. The final cohort, we enrolled patients into a 50-milligram dose every 24 weeks with the intention to see how long the effect lasts after the first dose. The baseline characteristic, as expected, based on the inclusion criteria of patients at baseline, have a TG level of about 220, LDL cholesterol in the 100 levels, non-HDL about 150 and ApoB tracking close to LDL cholesterol. And again, all the groups were well balanced across the board. Primary endpoint was APOC3 reduction. And once again, you see this similarly to the other population, a dose ranging effect on decreasing ARO-APOC3 that goes from 69% at week 16 with a 10-milligram dose to 90% with the 50-milligram dose. Here in orange, you have the group of patients or the cohort that received 50 milligrams every 24 weeks. And as you can see, the effect doesn't seem to last as much as when you give it every 12 weeks. On the right-hand side of this slide is the triglycerides data. And again, consistent with what we showed before, even though these patients have a much lower level of TG at baseline about 230 or so, the decrease is about 60% to 73% based on the different doses at either the nadir week 16 or at the end of the dose interval week 24. So we replicated essentially the same results that we did before in the SHASTA study. Here, the non-HDL data, which I think is one of the critical one and reason to believe in this, we see again a dose range in effect in which the 10 milligrams reduced non-HDL about 20%, 25% for 25 milligrams and almost 30% for the 50 milligrams. And when we look at the LDL, we see no changes in the 2 lower doses, 10 and 25 milligrams, and we're seeing a decrease of LDL of 10% to 11% with the 50 milligrams every 12th week. Then on cholesterol, again, this was calculated, but I think it speaks to the point that Dr. Goldberg mentioned between LDL and the VLDL particles as atherogenic lipoproteins. We observed a 50% to a 60% reduction in remnant cholesterol, which essentially is the LDL. And again, we see some dose response and it's consistent at the nadir and at the end of the dose interval. ApoB is reduced with all doses anywhere between 10% and 20%. So we also see a dose response in the context of decreases of ApoB. With regard to safety, this behaved exactly as the other study. These AEs that were seen are all related to comorbidity. There was no reported SAE that will attribute to ARO-APOC3. And we did see a similar event with regard to the increase in HbA1c as we saw in the first study. We applied the same protocol amendment, and we observed essentially the same phenomenon that patients tend to come back to their baseline A1C once they adjust the treatment or enhanced compliance with the treatment. So yes, Dr. Goldberg showed this data, and this is what will get me really excited about this program to develop ARO-APOC3 for the treatment of mixed dyslipidemia and prevent cardiovascular disease. And the way I think about that is look at the first row that represent the average value of all these lipoproteins and TG in a typical patient with mixed dyslipidemia treated with optimal statin therapy. So this is a large population that has already been addressed, to a significant degree, the LDL issue, and they still have this profile. Once you intervene with an ARO-APOC3, APOC3 goes from 15 to 2 TGs, as you see now, it's 59. The high level is -- or the upper limit of normal is 150. Non-HDL is now 107 instead of 150. LDL decreased to 98. ApoB now is 75. Remnant cholesterol is extremely low. And HDL increased by 65%. So now this person have a lower LDL and higher HDL. So you see that and look at that as a whole and fast-forward 5 years, a person that have persistent cardiovascular disease and have this profile versus a person who have persistent cardiovascular disease, so high risk for it and has this profile of atherogenic lipoprotein. And this is exactly why we think that this drug deserves a proper cardiovascular outcome trial to really demonstrate that treating this risk factor will translate into prevention of atherosclerotic progression and eventually cardiovascular outcomes. I also wanted to share with you that we're working now on selecting the clinical research organization, the CRO, and the academic research organization. That's going to happen within the next few weeks. And yes, we are a mid-sized company. And we're getting ready to run this large study. Okay. So ARO-ANG3, the ARCHES-2 study, mixed dyslipidemia. This study has exactly the same inclusion criteria as the MUIR study on ARO-APOC3. We randomized patients to 3 treatment groups, 50, 100 and 200 milligrams, based on week 12. And the endpoints are the typical for these type of studies. And I will just summarize the data from the highest dose, 200 milligrams, to make it simple and just too many numbers here. But this is the profile that ARO-ANG3 offers in patients with mixed dyslipidemia: 76% reduction in ANGPTL3; 60% in TG; 18% reduction in LDL; 60%, remnant cholesterol; non-HDL went down by 36%; and ApoB decreased by 22%. So once again, you see a very favorable lipid profile. And we're still evaluating how to move this program forward because really the data is very impressive as well. Importantly, we did not see any increase in liver fat or any issue with LFTs in liver fat decrease by about 30% in the 200-milligram dose group. Again, the adverse event were exactly the same as we see before, really related to the comorbidities seen in this patient population. And we did see a similar increase in A1c. And the pattern followed what I just mentioned before. So now Dr. Goldberg already talked about HoFH. It is a very significant disease. It's very infrequent. It's very rare. But those who have it have a very short lifespan. And this starts in early childhood. The treatments are very complex. There is a proof of concept that even ANGPTL3 does have a very positive effect based on the antibody to ANGPTL3. And Dr. Goldberg presented that data. But we're really looking to see whether we can have an intervention that will be more friendly to patients, and particularly children and as effective as this antibody. So we conducted a Phase II study, open label, patients with HoFH. In this case, they were genetically confirmed. We enrolled 16 patients, primarily I think were 18, randomized to either 200 million or 300 milligrams every 12 weeks. And as you can see, a very significant decrease in LDL cholesterol. And patients have like 120 to 170 milligrams decrease in LDL cholesterol. So that represents about 38% to a 42% reduction. And this is, I think, very consistent with what has been shown with the antibody therapies. And that's through both 16 weeks and 20 weeks. So looking forward to continue this program. And I wanted to finish again like how we started and tell you where we are right now with the cardiometabolic program. FCS is on its way to registration. The study is fully enrolled. Last patient, last visit will be in the second quarter of next year. And at that point, we will initiate the work to get to an NDA filing hopefully before the end of the year. So we're already working, not only chasing the clinical aspect but in all the other aspects that are required to file a new drug. And that for us, of course, is the first time we're doing it. And we're very excited as a company to be doing this. The SHASTA study, I mentioned how we're thinking about the Phase III. We're going to go to the agency within the next quarter to conduct the end of Phase II meeting that defines the Phase III strategy. And the intention is to start the Phase III programs -- or studies soon after we have agreement on the plan with the FDA and with the European Medicines Agency. Similar pattern with the cardiovascular outcome trial with ARO-APOC3. We're going to start this by a few weeks to be able to complete the work and to have the briefing document proper. By that time, we're going to have the structure in place to run this study once we complete the end of Phase II. Again, we're going to start our cardiovascular outcome trials soon thereafter. And finally, ARO-ANG3. Right now, we are moving the program on to again a Phase III registration study. And this study or this program will start probably early in 2024. We are still considering how to move ARO-ANG3 in the other patient populations. With that, I'd like to invite now my colleague, Tracie Oliver, who will talk to us about the commercial opportunities. Thank you.

Thanks, Javier. So a couple of takeaways I want you to have from my presentation is, first of all, ANGPTL3 and APOC3 are validated targets for the treatment of dyslipidemias. And so our programs have been largely derisked going forward. As you saw from Javier's presentation, for both ANG, APOC3 -- sorry, for ARO-APOC3 and ARO-ANG3, we are starting with really very ultra-orphan indications and then plan to rapidly move towards larger indications that have blockbuster potential. Also, I'll show in this presentation how we plan to differentiate ARO-APOC3, which will support a very strong value proposition for physicians and payers but more importantly, for patients. And then finally, talk a little bit about how we plan to move forward in commercializing our first drug. So this schematic shows sort of how we think about the dyslipidemia market, where you start with very high levels of LDL-C with familial hypercholesterolemia or very high levels of triglycerides with familial chylomicronemia syndrome and where we think our drugs can have the most impact. So right now, we're targeting FH for ARO-ANG3 and then mixed dyslipidemia out to FCS with ARO-APOC3. And these market segments combined represent about 43 million adults in the U.S., so very, very large markets eventually, although we will start, as I said earlier, with HoFH, about 500 people in the U.S.; and FCS, again about 500 people in the U.S. For ANG3, our focus, as Javier has said, is really on HoFH. It's an ultra-orphan indication associated with significant morbidity. As Dr. Goldberg has laid out, the efficacy of ANGPTL3 inhibition in HoFH has been well established by evinacumab. Where we see our product being differentiated is that our mechanism of action doesn't rely on LDL receptor activity, like the PCSK9 inhibitors. And so we anticipate we'll see a stronger and a more predictable response to ARO-ANG3. And then relative to the other pieces -- sorry, evinacumab again is we have a very small injection volume subcu versus IV for evinacumab and then versus the PCSK9 inhibitors every quarter dosing versus monthly or every 2-week dosing that you have with Repatha and PRALUENT. So we see this product as being highly differentiated for this patient segment. And we know that we have a lot of potential in refractory hypercholesterolemia and also mixed dyslipidemia. But I'm not going to talk about that today. I'm going to focus more on ARO-APOC3, which is furthest along in clinical development. So as Javier said, we have completed our enrollment in FCS. This is an ultra-orphan disease associated with significant morbidity. Our PALISADE study is not limited to genetically confirmed FCS, which we see as an advantage. While we know that this will be -- this will yield limited revenue just based on the size of the market, it will be an important indication for establishing us as a commercial-stage organization. And because SHTG is somewhat related to FCS, we know that we'll be able to leverage a lot of our learnings and insights from the FCS market into the SHTG market, which is much larger, about 4 million people in the U.S. And a market potential of over $1 billion. And then finally, going into mixed dyslipidemia, where Javier demonstrated earlier, we have impact on both triglycerides as well as LDL-C and ApoB. There's about 6 million to 10 million people in the U.S. currently not at target for LDL-C and triglycerides. We had this unique opportunity to address this patient population because of our ability to lower triglycerides, ApoB and LDL-C while increasing HDL-C. So we are planning a CVOT trial. And again, the potential in this market is over $1 billion as well. So if we look at how we're going to differentiate APOC3, really we're focusing on a couple of different areas, where we're moving beyond just reducing triglycerides as the surrogate biomarker to really establishing clinical outcomes associated with reducing those triglyceride levels. As I said earlier, our PALISADE study in FCS includes both genetically confirmed patients as well as clinically diagnosed patients, so patients with triglyceride levels greater than 880 milligrams per deciliter and a personal or familial history of pancreatitis or hospitalization due to abdominal pain. And this is different from the other products that are currently being studied in FCS. And I think what it probably means for us is that we will have a broader patient population eligible for our therapy when we finish the study. Dr. Goldberg and Javier both spoke about hypertriglyceride-associated pancreatitis. This is a real risk factor, a real risk for patients with triglyceride levels greater than 500. And I know that this graph is a little busy. But what I want to show is that the darker blue colors are the higher triglyceride levels. And these are patients who have had 1 episode of pancreatitis in the previous 12 months or more than 2 episodes of pancreatitis in the prior 12 months. And what you can see is that there is an increase in the incidence of pancreatitis as triglyceride levels go up but dramatically increases in patients who have a recent history of recurrent pancreatitis. And so the study that we're planning, the SHASTA-4 study, will move beyond just showing that we can lower triglycerides but hopefully show that we have a clinical impact on patients by reducing triglyceride levels. And as Javier showed, the vast majority of patients actually shift to the left to have triglycerides less than 500 milligrams per deciliter, which puts them below the threshold at which pancreatitis is associated. And then also, again this is hard to read, but what I want to impress upon you is it's not just pancreatitis that is impacting the morbidity of patients. These patients with high triglycerides, be it severe hypertriglyceridemia or FCS, have significant burden of illness. And each of these bubbles represents a symptom. The size of the bubble represents the proportion of patients who report this symptom, either from a daily basis to a monthly basis. And what you can see is there's a vast number of symptoms impacting these patients: physical, emotional and cognitive. Patients experience abdominal pain, joint pain, nausea on an almost daily basis. They worry about food. I don't know what you guys had for breakfast this morning. You probably exceeded the daily fat intake just with your breakfast that an FCS patient can have. They worry about food. They worry about the next episode of pancreatitis. They worry about somebody else has prepared their food, how much fat is in there. And they suffer from depression, social anxieties because they don't want to go out and be with other people, doing the things that people normally do, like have a beer and a pizza. And then finally, cognitive symptoms, they report the sort of what they call a brain fog or impaired cognitive thinking. And by adding in patient-reported outcomes to our studies, we'll be able to hopefully address these physical, emotional and cognitive symptoms associated with elevated triglycerides in these patient populations. And then finally, our CVOT study will really look at the sort of cumulative impact of APOC3 on the lipid and lipoprotein profiles by reducing the residual risk posed by these atherogenic lipids and lipoproteins. And we're really excited about the CVOT study because we believe that reducing triglycerides, LDL-C and ApoB will result in a positive study. And then finally, just a quick note on where we're focusing our attentions right now. I've talked about how we plan to differentiate ARO-APOC3 in terms of generating early clinical outcome studies, outcomes with pancreatitis as well as the cardiovascular outcome study. We're preparing the market. There's a tremendous amount of clinical inertia and misunderstanding about triglycerides. And one of our physicians called triglycerides the Rodney Dangerfield of lipids. It gets no respect. And so just educating physicians on the impact of high triglycerides and increasing the urgency to reduce it will be a key focus for us as well as working with payers to understand the risk posed by high triglycerides and developing a compelling value proposition that will support reimbursement and access. And then finally, we're preparing the company for our first launch. This is very exciting as we move from a clinical-stage company to a commercial-stage company. We are in the process of hiring out our commercial team as well as medical affairs teams and expanding all of the internal capabilities across the key functions internally. And we've been thinking a lot about our go-to-market strategy for U.S. and ex U.S. We have not finalized it yet. But what we're trying to do is build into an optionality so that we can take what we are doing in the U.S. and leverage it where possible ex U.S. We're also gaining our investment to major milestones such as reimbursement in the EU market as well as new indications. And so you'll see us grow from a small organization to a much larger organization as the market dynamics change. And that is it for me. Chris?

All right. We have done our best to exhaust you. I have a handful of slides and then we'll be done. So I think the update today represents a ton of potential value. Our late-stage clinical programs are moving rapidly towards commercialization. And these are ARO-APOC3, ARO-ANG3. We didn't talk about fazirsiran. We didn't talk about olpasiran. But those are already in Phase III studies. Our earlier clinical stage programs are showing promise and a clear path as represented here by ARO-C3 and ARO-PNPLA3. C3 has multiple indications we can go after, data so far look compelling. PNPLA3 is, I think, the only truly genetically validated, genetically targeted approach to at least a population of NASH. And we are excited about that program. Our pulmonary program appears to work. That franchise appears to work. ARO-RAGE has seen up to 95% knockdown. We expect ARO-MMP7 and ARO-MUC5AC to follow. And we expect a large number of additional targets to follow after that. Platform expansion continues. We talked about skeletal muscle today. But we expect to either partner ARO-DUX4 or file a CTA over the next month or so. We expect the CTA for our next skeletal muscle targeted program by the end of the year. CNS is an exciting place to be. There are a number of good targets. Our preclinical data are quite good. We're excited about getting into that space. And I expect our first CTA over the next couple of months. CNS with systemic delivery is a truly transformational thought. Our early data are compelling there. We are looking forward to continuing to develop that and see if we can make it into the clinic with that sort of approach. And then finally, adipose, this is the largest endocrine organ in the body. And we see a ton of interesting new targets there. So many of the areas that we are addressing have been neglected in the past but are now sort of back in vogue. And that's not strategic on our part, a lot of that is just good luck, I think. Cardiovascular disease is one that has been avoided for years because of the cost associated with the CVOT, the length of the CVOT. But given the continuing risk factors, given new GWAS analysis, there is new interest in the space. And we are, of course, addressing that with ARO-APOC3 and ARO-ANG3. NASH. NASH was hot until it wasn't. It's a very large opportunity. It's a very large problem. It turns out it's a very difficult disease as well. And it is probably heterogeneous. We have recently seen promising data. And I think the market is back to being interested in NASH. And again, PNPLA3, I think, is a very unique approach to a subpopulation of NASH. Pulmonary. There have been very few inhaled drugs ever approved by the FDA. And so it's been a very difficult space to be in. It is clearly a space where there's an awful lot of opportunity. And we are there now with ARO-RAGE, ARO-MUC5AC, ARO-MMP7 and more going forward. CNS, as we talked about, is a place where people were not jumping over themselves to make bets because it was such a difficult space. Recent data have been encouraging. Our nonclinical data are encouraging. We are excited to be there. And then finally, adipose, again the largest endocrine organ in the body. If I was here talking about an obesity drug 6 years ago, no one would have listened. This is an important problem. And I think that being able to address the adipose issue enables us to get into that space as well as other metabolic conditions. So we mentioned early on in the presentation today, we mentioned about the last 6 years. And so what about the next 6 years? As we talked about, the 6 years between '17 and '23 have brought -- or will have brought 18 clinical candidates into clinical studies. We think during the 6 years, from 2023 to 2029, will bring an additional almost 20 new drug candidates into clinical studies. And given that RNAi is an increasingly validated technology and given that TRiM is an increasingly validated set of platforms, we expect the majority of these around 40 drug candidates to actually make it to a patient. And I'm reminded of when Elon Musk and syndicate acquired Twitter. And I didn't follow this, but I heard about it. I think I saw a picture of it. He had a video of him walking into Twitter headquarters with a sink and the caption was, "Let that sink in." Well, let that sink in. We are not a large company. And we're talking about, over the course of a fairly short period of time, upwards of 40 individual clinical candidates in clinical studies and a majority of those making it out into patients. It's astonishing. And it is an exciting place to be entirely. We don't look to benchmarks, we look to be the benchmark. So what about the next 6 years? We expect multiple product launches spanning small and broad markets in different therapeutic areas via wholly owned and partnered assets. We expect dozens of drug candidates in clinical studies spanning early to late-stage development across different therapeutic areas, both wholly owned and partnered. And we are really a different kind of biotech company. So how do we pay for all this? Well, there are multiple capital sources that we will be tapping to reduce the long-term cost of capital. So certainly, equity, the selling of common stock is common in our industry. We have been judicious in this. We haven't raised money in about 3.5 years. And I think we'll continue to be disciplined in this approach. Business development has been a big thing for us. Over the last 6 years, we brought in almost $1 billion of capital from business development opportunities. We will continue to do that, and I think that grows. Not only will we be doing around a deal a year, I think, going forward. But our existing deals, as they mature, will be bringing in additional capital as those milestone payments get larger. Debt, I think, is something that we can consider going forward. We are approaching the state in our company where we can tap the debt market. Creative sources via product financing or even royalty monetization as we did late last year is an opportunity. And then of course, commercial sales, we expect to be commercial in the next few years. So with that, thank you for your endurance. And we can open this up for questions.

Prakhar from Cantor. So on ARO-APOC3, on the diabetes events, specifically for the 50-milligram arm, how many patients had diabetes at baseline? And were there any imbalances between the dosing arms on patients with baseline diabetes, given the higher events? And do you plan to take the 50-milligram dose forward?

So about 60% of patients with this type had diabetes at baseline. So a small fraction of them had an excursion with Ha1c. So it wasn't the majority, it was the minority of those patients. But it was different than the placebo group. With regard to -- so the second part of the question was?

Were there any imbalances?

Oh, so no imbalances, no significant imbalances. Our baseline on this type are relatively large and properly randomized, so randomization works. And then whether we go with 50 or 25 milligrams, we're working right now on that, both from the pharmacology perspective in terms of proving the modeling and population PK/PD modeling. And also, in the end, it's a clinical judgment. We need to look at the benefit. We need to understand that, talk to our colleagues and experts to really decide which one is the dose that provides the best benefit/risk for patients with mixed dyslipidemia.

Will Pickering from Bernstein. For your cardiovascular program, could you describe the rationale for prioritizing APOC3 rather than ANG3 in mixed dyslipidemia? Just looking at the impact on LDL and TG, it looks like it's a bit more balanced in ANG3. And then in SHTG, could you share your expectations for prior auth requirements in that indication, whether we can expect something kind of challenging like in the early days of PCSK9?

All right. So with regard to why we select the ARO-APOC3, we think that the profile that ARO-APOC3 deliver is incredible for patients with mixed dyslipidemia. It does address that residual risk. Now does ARO-ANGPTL3 does something very similar? And the answer to that question is yes. Can we do two cardiovascular outcome trials? And the answer to that is slightly no. So we're thinking about how to move that forward because that drug has a very profound effect in lipids and lipoproteins and atherogenic lipoproteins. And so we need to think about we have an indication for refractory hypercholesterolemia, for heterozygous familial hypercholesterolemia and how we go into the major indication eventually. So those are choices you need to make sometimes. We have two very good drugs. Both of them address what the unmet medical need is in the residual risk of cardiovascular disease once you have LDL control. And I think that may be part of the explanation. Now we are focusing beyond LDL. Do you want to add to this?

In terms of prior authorizations for SHTG, that's always a risk. We know that payers are always trying to reduce costs as much as possible, which is why we're really focused on demonstrating not only just a lowering of triglycerides but also the clinical benefit associated with that with the reduction in incidence of pancreatitis and then obviously with the CVOT study. So at the end of the day, who knows what they'll do, whether they'll do step edits, prior authorizations. But what we're doing is trying to put together the strongest value proposition that we can to avoid that.

Joel Beatty from Baird. For the APOC3 CVOT, what relative risk reduction do you anticipate?

Well, the one that you need to get a viable drug. So when you look at what is out there, we have now every single cardiovascular outcome trial in our working space, starting with the 4S study. And I was in medical school when that happened, or maybe I was in the residence already. And I always look at that because that was the #1 or the first study that actually showed the reduction of LDL translating to clinical benefits. And that was in about 30% or 35% reduction in people with persistent disease. Since then, we have many study in drugs. You have PCSK9. And if you look at the PCSK9 Phase III study, the reduction was about 13% or 14%. And they need 27,000 patients. Whether that has to do with a shorter relativity study with the patient population they selected, we don't know. So I hope to see more than 20% risk reduction. But that's how we're going to -- I'm thinking right now. I don't know if, Ira, you want to make any comment about that.

No, it's a very complicated calculation.

Well, the study, just high level, it will be an event-driven trial. The main work that we're doing now and Dr. Goldberg is helping us with that is to really select the patient population to really understand who are the patient population that the profile of ARO-APOC3 will change their cardiovascular risk. So that is number one. And of course, from there, to assess the effect size is something that we're working on. We're doing some modeling work to try to estimate the event rate based on that population we select and the potential effect size. And again, the study, as most of the studies, will be an event-driven study. So I hope we get it right.

This is Madhu Kumar, Goldman Sachs. So one really granular question about SHASTA-4 and then a bigger-picture question about APOC3 and triglyceridemia. So in SHASTA-4, you mentioned the idea of patients have prior pancreatitis. Is it one event, two events? Or do you have that kind of sorted out yet? And then the kind of bigger-picture question really for Dr. Goldberg, as we think about the use of this drug in severe hypertriglyceridemia, what do you think it will take not for people like you, but for like Joe and Jill Schmo, practicing physicians, who see a patient who has very high triglyceride levels to prescribe this? Is the pancreatitis outcome really important for them more so than you or less so than you were the same? Like how do you think about kind of the broad use in terms of pancreatitis prevention versus just triglyceride declines themselves?

I mean, you're right, sometimes patients will have one episode of pancreatitis. They will become so sick that they will take care of themselves and not have it happen again, except they have a genetic problem. But taking care of yourself is very hard. It's sort of like weight loss and diet and whatnot. And it also totally changes the way people live. And so I have -- I actually have a number of genetic LPL-deficient patients who have triglyceride levels that run 400 to 600. They exercise an hour to 2 hours every single day. They eat 0 fat. It is their life. And I have the patients who don't get pancreatitis because of that. So what happens in the community? Well, in the community, when people have recurrent pancreatitis, it's expensive. It's a big problem. And I think they were ripe for this drug because it's easy to give. Javier knows I've got patients who I've tried to solicit the drug kind of out of the usual standards for the trial because they don't exactly fit for the trial. This is not going to be so hard to take. It used to be, even when GLP-1 drugs came in, people asked the same question, "What will community doctors do?" Well, there's a drug that you have to give by injection. So they'll get used to using this. I think it will start with big medical centers, then it will spread out to the community. I mean, doctors adjust, they even order them.

And in regard to the question about the inclusion criteria, right, on the SHASTA-4 study, so we're working on that right now. The goal is to have enough number of events to show the clinical benefit. I anticipate that effect size will be very large. And I think we have some evidence of that. So it would be a balance between how severe the patient needs to be and how fast we want to enroll the study. Because our ultimate goal is to get to the market as soon as possible. And that's why this product has these two stages, the TG 6 months with the two studies and then the pancreatitis risk reduction with the SHASTA-4 study up to 2 years. So the final tune of the inclusion criteria, whether it's 1 episode, whether it's 2 episodes, whether it's within the last year, 2 years or 5 years, is what we're working now to really assess what is the event rate, what's the sample size. And that will answer that question.

Ellie from UBS. And thanks for the comprehensive R&D Day. Sorry if I missed it, but I just want to clarify on the pulmonary programs just exactly what data we're getting this year. I know you're adding the FeNO cohorts, the COPD cohorts. Just will we get data from either of those patient cohorts this year? And then I guess, just how should we think about what data we will be getting in terms of some of the patient cohorts this year heading into early next year? And then second part of the question, what you're looking to see. I think you answered it a bit with FeNO and asthma. But maybe just commenting a little bit more in terms of COPD and what some of the target biology kind of markers might be there in terms of MUC5AC.

With regard to clinical new data before the end of the year, do not expect the FeNO or the COPD. We're working on the protocol amendments now. It takes a couple of months to get that protocol amendment approved and a few months to get the study done. So I would say Q1 probably of next year. It's likely that we will see the COPD mucous target engagement data and the FeNO and the RAGE asthma cohort to that space next year. This year, we're going to have more normal, healthy volunteers and more of the patient cohorts, not the FeNO or not the COPD. So you will see more patient data. You won't see these last two cohorts. And with regards to the MUC5AC Phase III in COPD, I think the #1 goal by far there is to look at target engagement. It's the right patient population to do that for the reason that we expressed before. Those are the patients who develop mucus. And they have more production of mucin base or MUC5AC overexpression. So I think at this point, that will be the key feature. Now we're measuring pulmonary function test, exacerbation, quality of life, PROs to assess the level of cough, sputum expectoration and all that stuff. So it's going to be a relatively small study for those type of endpoints. But while looking at those endpoints in programmatic ways, maybe we do have a surprise there.

Patrick at H.C. Wainright. First, just with PNPLA3, are you going to be looking for a new partner for that program? Secondly, how do you think about pricing and reimbursement in HoFH relative to the biologics, assuming you achieve that target product profile? And then regarding ARO-APOC3 with severe, high triglyceride, appreciate the opportunity for the early file with the 6-month endpoint. I'm just curious how payers think about that. Would they be looking for some of the maybe longer-term follow-up data? And finally, realizing very ahead here looking at revenues further out but with the longer-term strategic plans, can you kind of frame for us what revenue targets might look like over the longer term?

So I'll take the first and the last. And then Tracie can take the middle part. So PNPLA3, we think it's a great target. And so we are not looking to partner that at this point. And the final question, we are not prepared to give any sort of revenue guidance at this point.

I'm trying to remember the two middle ones. I think the first was HoFH and pricing. We have not set pricing for that yet. But it's an ultra-orphan indication. So you can assume it will be somewhere between 0 and that. And then what was the other question? About...

6-month, how would payers...

So we anticipate that currently all payers -- the data that they have right now and that we expect they'll have is strictly limited to lowering triglyceride levels. And so will they wait for a clinical outcome associated with that, like they did with PCSK9 inhibitors, the CVOT studies? I don't know. But we're not aware of any trials like we're planning for other drugs. And so I think we'll be able to satisfy payers quickly in terms of what's the clinical benefit of lowering those triglycerides.

And just in terms of the sequence of events that will happen with the SHASTA program, so one study is 3x bigger than the other one. We expect to have fully enrolled the severe -- the pancreatitis risk study by the time we're filing. Then you have about 9 months to a year to get approved. And so I expect that soon after the approval, we will have that pancreatitis data. And of course, we're going to do a supplemental NDA. But in the meantime, that can go into the dossier and be part of our value proposition to pay as well, right?

Luca Issi, RBC Capital Markets here. Maybe Chris, on DUX4, can you just expand a little bit more on what's go/no-go decision here to your file of CTA versus partnering? And maybe what gives you confidence that you can make that call one way or the other within the next 30 days? And then maybe for ANG3, Dr. Goldberg, I know Javier already touched upon this. But obviously, Ionis and Pfizer had an antisense oligonucleotide going after ANG3 that was ultimately discontinued, given it actually caused dose-dependent hepatopathy accumulation and ALT elevation. What's the ongoing hypothesis for what drove that signal? And what gives you confidence that actually that signal was off-target and not on-target? And then maybe last one, going back to crossing the blood-brain barrier, maybe either James or Christine, can you just talk about how you're crossing the blood-brain barrier? We've seen a few others using the transferrin receptors. But some of these companies are now running into anemia signals. So wondering if you can comment on whether it's the same approach that you're using or you're using a different approach.

All right. I'll start with DUX4. So we're in a good spot with DUX4. We've done acute tox. We've done chronic tox. The CTA is written. We just need to present. And we're happy to do it. We think it's a compelling drug candidate. We think the market is clear. We think the target is clear. So we're happy to do it. As I said in the conference call, we are prepared to do that, I guess, in early April thereabouts. And then we've got some inbound interest. And so we just wanted to let that play out. And I think we'll know over the next 30 days or so in which way we're going to go.

Yes, about the liver fat. So APOC3 deficiency or any of those treatments is not associated with liver fat, right? The ANGPTL3 issue is the patients were initially described in St. Louis as being different from all the other patients who have low cholesterol. So ANGPTL3 deficiency in humans, because it's really low levels of cholesterol, triglyceride, HDL, everything and were described as being different from the patients who are missing like ApoB or the MTP protein because they did not have fatty liver. And so that you would inhibit the protein and have something totally different from what happens in humans suggests it's most likely an off-target effect.

And then I can cover the last question. So yes, we -- I think all I can say at this point is we use a ligand-targeted approach. And we're not ready to disclose the ligand at this point.

All right. I think that's all we have time for today. So thanks, everybody. We appreciate you coming out here today. Thank you.