Alnylam Pharmaceuticals, Inc. (ALNY) Earnings Call Transcript & Summary

Hi, everyone. Welcome to this RNAi Roundtable. Thank you for joining us. Today, we're going to be discussing the progress that we're making across our central nervous system and broader extrahepatic delivery efforts. I'm Josh Brodsky, Senior Director of Investor Relations and Corporate Communications at Alnylam. And I'm joined today by my colleagues, Vasant Jadhav, Senior Vice President of Research; Kirk Brown, Senior Director of the CNS program; and Bret Bostwick, Director of Clinical Research. Today's RNAi Roundtable is part of a series of roundtable webinars that we have been hosting over the past few months to review progress across our various programs. The event today is expected to run approximately 60 to 75 minutes. Vasant will moderate a Q&A session at the conclusion of the presentations. [Operator Instructions] As a reminder, we will be making forward-looking statements during this webinar, and we encourage you to read our most recent SEC filings for a more complete discussion of risk factors. And with that, I'll now turn it over to Vasant.

Thank you, Josh. My name is Vasant Jadhav. Let's begin with the introduction to RNAi therapeutics' new class of innovative medicine. We are harnessing the nature's pathway of modulating the gene expression, but with the exogenous siRNA for the intended target. So essentially learning from the nature. But what we are most pleased with the programs that we have done so far is that the subtitle here. That is this approach is now clinically and commercially established with transformational potential and more importantly, reaching to the patients with unmet medical need. Now all of this, this journey of about 20 years or so, was based on this very focused R&D strategy on the next slide. And this involved kind of a three-pronged approach by having or by selecting genetically validated targets, going after the tissues of interest and very critically finding the delivery solutions. Now anytime you are optimizing or taking the new technology and developing it having as fewer variables as possible is important, and that's where we wanted to go with the genetically validated targets. And in terms of the delivery, on the next slide, after going through many, many different approaches, we now come down to 2 main ones for the siRNAs. Now when we look at the siRNAs as a drug, I mean, they essentially like or lack the drug-like properties. They are big. They are charged. They are water loving. They are nothing like small molecules. On top of that, their site of action is intracellullar. So given all these challenges, delivery was always the biggest hurdle to overcome, to realize the potential of RNAi therapeutics. So the first approach that we have come up with is a lipid nanoparticles. Now lipid nanoparticles is something that whole world in a way has heard of quite a bit more. But this was actually the delivery system that we used for our first approved drug, ONPATTRO. And here is one of the comments from [ CN News ] that without these lipid shelves, there would not be -- there be no mRNA vaccines for COVID-19. So we believe that we have played some role here indirectly by advancing these lipid nanoparticles with our first drug. But majority of effort in our pipeline for the liver targets has moved to the gamma siRNA conjugates. These are single chemical entities, meaning they are fully characterizable. They are a single system, unlike a multi-component system with the LNPs, and we have seen great activity with this approach. And we have applied the continued innovation as we have done with LNPs, the same thing with the conjugates as well. We started with partially modified siRNAs and made the breakthrough with what we call at that time, the standard template chemistry STC, and then that developed into the ESC, ESC+ and now IKARIA platform, we've recently introduced the ultra long-lasting siRNAs, which can be potentially given once every year, a most vaccine-like dosing for different targets of interest. Now let's remember on the next slide. So with all these advances, we're also seeing some of the key features of the siRNAs, the RNAi therapeutics in liver. First of all, there isn't really anything like undruggable target. If there is an mRNA, that should be a target for the siRNAs or RNAi. We can achieve a very high level of knockdown. And this knockdown is clamped. And there is no obvious sawtooth kind of response. It's a clamped knockdown and that's very important, it's very differentiating from antibodies. We'll talk more about the durability, one of the key, key features of our platform. We're now thinking about every quarter, biannual or potentially now with the IKARIA platform, once-a-year kind of dosing. The siRNA conjugates, they can be stored at a room temperature and the route of dosing is subcutaneous. Now as we expand the durability feature here, it's also going to be important as we go into the other tissues. So a couple of other important points coming up in the next slide that RNAi is durable and reversible. Now in terms of the durability of the siRNAs, we recently published a preclinical work that shows that there is, intracellular depot, that is likely the reason for the durability of our molecules. This is likely because of the incredible potency of the siRNAs and the RNAi pathway that even the small amount of siRNA residing inside the pathways can give very long-lasting activity. This body of work was put together and published in Nucleic Acid Research recently. That article was also recognized as a breakthrough article by The Journal. And just we concluded the Oligonucleotide Therapeutics Society meeting just this week actually, and it received the Paper of the Year award from this very prestigious society in the Basic Research category, which is pretty cool considering that this work is coming from the biotech company in the basic research area now. Now we always see in this one that the RNAi activity is reversible. It is long lasting, but it is not permanent. But we have shown this in a very cool experiment here. This is in the mouse model on the right side. So these mice were given the GalNAc-siRNA, and we waited for about a week or so by which we see full knockdown or the max knockdown. And then the group of animals undergo the procedure to remove 2/3 of their liver. Now given the properties of the liver and the regeneration, it grows back just about in a week or so. And so if RNAi was permanent, we will not see the impact on the activity. But as expected, we do see loss in activity as the new hepatocytes have much diluted amount of the siRNA after the cell divisions. So it is expected, what we have seen before as well, but it's good to see in this kind of preclinical studies. So with all of these efforts, we now have this incredible pipeline of approved drugs, commercial ones, many in the late stages, quite a few in early mid-stages as well. I won't go through in all of these, but it's just an incredible pipeline that has come through the advancement of these delivery technologies. Now most of these or all of this are these are the liver-based targets. So if we go to the next slide, these advances, we believe, are going to be very helpful for advancing the siRNA delivery to the other tissues. When we look at the delivery for the other tissues, there will be 3 main components. One would be the siRNA chemistry; the second, linker; and third, the ligand, the ligand that will take these siRNAs to the tissues of interest. Now the advances that we have done on the siRNA chemistry and the linker for the hepatocyte delivery, we believe we can take all of that and now change the targeting ligand or the delivery portion so that we can reach to the other tissues. So we're taking all the advances of the ESC+, the IKARIA platform, the specificity and the potency and applying it to the other tissues. And the tissues of the interest are CNS, eye and lung. Now within this space of the extrahepatic delivery on the next slide, we have a couple of very important alliances, the first one being with Regeneron, that was about more than 2 years ago now. This is a landmark alliance with Regeneron focused on the CNS and ocular RNAi therapeutics. We are making great progress with this one. And very recently, we also announced an alliance with PeptiDream. So what is PeptiDream, PeptiDream has a technology for discovering cyclic peptides. And these cyclic peptides could act as targeting ligands to reach to different cell types of interest. We believe this will really open up the space for us to reach different tissues. PeptiDream, within this collaboration, will select and optimize and synthesize the peptides for the receptors or targets of our choice and will evaluate these peptide-siRNA conjugates in, in vitro and in vivo studies. So this leads us to what's happening in our preclinical programs and across these 4 different tissues, liver, CNS, eye and lung. To speak more about this one, as Josh was saying before that the focus for today will be a lot more on the CNS side. And so for that, I invite my colleague, Kirk Brown, to talk about the CNS delivery.

Thanks, Vasant. Within the CNS, there's a high disease burden, high unmet need for new treatments in CNS diseases. There are numerous genetically defined neurodegenerative diseases, including Alzheimer's disease, ALS, frontotemporal dementia, Huntington's, Parkinson's and many more. There are a number of genetically validated targets known. However, there are very few disease-modifying therapies for these. We believe there is a significant opportunity for RNAi therapeutics directed against these disease causing these within the CNS to have a direct impact on disease progression. In the CNS, we are driving to build a best-in-class CNS oligo delivery platform that would include a highly potent, widely distributed molecule with a very long duration of action and a favorable risk-benefit profile. We're going to be doing this using our stabilized chemistry. But instead of GalNAC, we're moving toward a C16 ligand or lipid [ mode ]. The C16 conjugate platform is designed similar to our GalNAc platform, and it's been optimized for potency, durability as well as safety. We've performed exhaustive optimization of the various conjugates looking at position as well as design chemistry as well as stabilized chemistry that is performed so well in the liver. We've also included a vinyl phosphonate on the antisense [indiscernible] to help improve potency in extrahepatic tissues such as the CNS, including enhanced risk binding. One example of these experiments is on the bottom left, where we're using a single intrathecal administration that's a direct injection to the CSF of rats in the lumbar region, looking at day 14 at various CNS tissue, including the spinal cord as well as various tissues within the brain -- in brain regions. Here, we're targeting SOD1. This is a tool kit that siRNA designed to get superoxide dismutase. And we're looking at different ligands, such as C10, C12, C16, all the way up to C18 oligo. And what we look at is tissue mRNA knocked. And what we could see is spinal cord reduction is very good in a variety of different ligands, but none of them are more potent and active than that of the C16 molecule. And importantly, the C16 also provides us very good knockdown in the brain, including the frontal cortex and hippocampus. The figure on the right, again, looking at a single intrathecal administration into rats, now at day 28 post dose, we wanted to compare the influence of these different modifications on the activity. So on the left, we're comparing a vinyl phosphonate containing molecule with C16, one that lacks that vinyl phosphate, one that lacks C16 and one that lacks both. And what is quite clear to see is the molecule that contains both of those design features, including the stabilized chemistry produces remarkably strong knockdown in the spinal cord as well as very good knockdown within the brain regions again. Something that you might see is that we're also not knocking down within the liver and kidney after intrathecal administration of the C16 conjugates. Next slide. We also wanted to compare a couple of different aspects of the drug, one of which is dose dependence and another is the different cell types we're accessing with the C16 conjugates. The figure in the middle is a dose response, again, intrathecal administration in rats looking at spinal cord and brain. And we can see it low, mid and high doses. We see a stepping down or improvement of activity with each increasing concentration of the dose. We then went to look closely at the different cell subtypes and brain regions. And here, we're taking an antibody directed against the siRNA. We see good distribution. You can see good purple color throughout the cortex. And then we wanted to look at specific cell subtypes, such as neuron astrocytes and microglia, where we see good uptake and activity within each of those cell subtypes with the C16 conjugate. Importantly, we wanted to then evaluate if we can translate from rat intrathecal administration activity to nonhuman primates. And here, we're switching gears now targeting amyloid precursor protein or APP in nonhuman primates. This is an example of a single intrathecal administration, again in the lumbar region of nonhuman primates. Now we're looking at CSF biomarkers and the targeting agent biomarkers for APP or soluble APP alpha and beta. And we're tracking those soluble biomarkers through a time course that covers 6 months. And what we can see is, consistent with what we've seen in the liver, is a rapid reduction of the CSF target engagement biomarker and a clamped pharmacodynamic effect that last the full duration of this study evaluation, which was up to 6 months. Extremely encouraging for the translation of our active molecules. We also wanted to look more closely in safety, and we've evaluated the safety of CNS C16 conjugates in a variety of different settings. We've looked at 6 months platform studies in rats. We've done acute tox studies also in rats and a non-GLP tox study in nonhuman primates, looking at clinical observations, body weight, neurological exams as well as micro and macroscopic examinations, including examinations such as Flouro-Jade. Importantly, we've completed GLP tox studies in rats and nonhuman primates for our first program in the CNS ALN-APP, with no test article-related findings in any platform or ALN-APP study. To summarize what I've shared with you, we've developed a potent, durable and highly distributed siRNA throughout the CNS. We find that a C16 conjugate allows us very robust uptake and activities. We know that the modification pattern allows us good access into risk and durable and potent silencing for at least 6 months in nonhuman primates. We've seen a platform and ALN-APP CTA-enabling tox study to support further development of the C16 conjugates into the clinic. And our CNS pipeline strategy has a goal of taking genetically validated target genes preferably with biomarkers for proof of concept in Phase I and then a definable path to approval and patient access. And as Vasant pointed out, we have over 25 preclinical programs across 4 different tissues, feeding this pipeline. And in partnership with Regeneron, we've named ALN-APP and ALN-HTT as our first programs in the CNS. I will now pass it over to my colleague, Bret Bostwick, who will describe in more detail our first program in CNS.

Thanks, Kirk. My name is Bret Bostwick, Medical Director in Clinical Research, and I'm excited today to present on behalf of the CNS preclinical and clinical teams. Amyloid precursor protein, or APP for short, is a protein that when it's first made, it attaches to the fatty membrane that surrounds nerve cells in the brain. At first, it's a relatively large 87 kilodalton protein, but it's quickly cut up and metabolized into little pieces by a series of complex proteases, and this produces a diverse set of daughter peptides. One particular daughter peptide, weighing in around 4 kilodaltons, is amyloid beta 42. And this is now famous because it's sticky. It readily clumps up in the extracellular space, forming plaques. And of course, these plaques are the hallmark of Alzheimer's disease. These same clumps of protein is what the pathologist, Alois Alzheimer, noticed in the brain of disease dementia patients all the way back in 1906. Another one of these daughter proteins that is nearly the same size, A beta 40, so it's basically 2 amino acids shorter in length, it has a curious predilection for sticking to the walls of the blood vessels in the brain in a disease process known as cerebral amyloid angiopathy. These sticky proteins stick to the wall of the blood vessel, eventually weakening it and causing bleeds or hemorrhages, which results in clinical strokes. So this 1 parent protein, APP, can be harmful in 2 distinct pathophysiologic processes. Human genetics teaches us that certain mutations in the APP gene or duplications in the gene can cause early onset Alzheimer's disease. Other particular mutations in the same gene can cause cerebral amyloid angiopathy. There's even additional mutations in the same gene that causes an overlap between the 2 phenotypes. Let's start by discussing Alzheimer's disease. Alzheimer's disease is the most common dementia in the world, 5 million affected in the U.S. and 30 million affected globally. It's so prevalent that almost everyone knows somebody who has been affected by this disease and almost everyone knows the toll it takes. The life expectancy is 4 to 8 years from diagnosis. And this is a time period where not only the patient, but also the caregivers and families, are heavily impacted by the disease. At the macro level, the prevalence, combined with the morbidity, puts an enormous strain on our health care system and the economy. And all this is exacerbated by limited progress in finding disease-modifying therapies. APP is a genetically validated target for Alzheimer's disease. Mutations that increase APP production or alter APP metabolism cause early onset forms of AD. On the other hand, mutations that decrease APP processing are protective against the development of Alzheimer's disease. These genetic bookends support the idea that APP has a volume knob, and the volume has turned up too high in Alzheimer's disease. APP is also an ideal target for our emerging CNS clinical programs as it has a comprehensive set of flu -- and neuroimaging biomarkers we can monitor these in the clinical trial setting. So this not only allows for the understanding of target engagement, but may also provide an early look at downstream biology and potential indicators of disease progression. There are multiple opportunities under the Alzheimer's disease umbrella. Different subtypes of AD can be further defined by age of onset or disease stage or the presence of certain genetic predispositions. This figure compares where ALN APP acts relative to other classes of therapeutics. To our knowledge, ALN-APP has the potential to be the first therapeutic for AD tested in human trials that is a genetic medicine, targeting either the DNA or RNA of ACP. We think this could afford several key advantages relative to the ability to modify disease. I would like to highlight 2 important components of the rationale today. First, APP mRNA is the proximate substrate for all post-translational amyloid proteins and peptides. This includes the well-studied pathway that's at the top of this figure, where A beta is produced and secreted outside the cell. But APP-mRNA is also the sole precursor for the less well-studied pathway that produces other APP fragments. And these fragments are increasingly being recognized as important disease contributors. The second important part of the rationale focuses on where these proteins accumulate and aggregate, although much focus to date has been on the A beta that's secreted outside the neuron forming plaques, there are also important APP fragments that accumulate inside the cell. And these may be drivers of [ tile ] phosphorylation and deposition. So while BACE inhibitors or gamma secretase inhibitors are focused on a particular proteolytic pathway, ALN-APP is well upstream of that process. It reduces, not some, but all APP protein fragments. In addition, the other APP fragments, which don't require beta secretase cleavage and therefore, are not impacted by BACE inhibitors, they will also be lowered by ALN-APP mechanism of action. In this sense, ALN-APP has the potential to be the first therapeutic that comprehensively lowers all APP proteins and peptides, both inside of and outside of the neuron. Supporting this hypothesis are several preclinical experiments that we'll cover in the next few slides. Shown here are the data from a human neuron and human astrocyte co-culture system, which has the ability to distinguish between intracellular and extracellular protein levels via measuring either the culture media on the left here or the cell lysates on the right. The X-axis represents the in vitro concentrations of either control duplex, the APP siRNA or a BACE inhibitor. And the Y-axis is the percent of soluble APP beta protein remaining. As you can see on the left, both APP-siRNA and BACE inhibition lower the extracellular soluble APP beta level in a dose-dependent fashion. The graph on the right begins to demonstrate some of the mechanistic advantages of RNAi. At the 1, 3 and 10 micromolar concentrations, BACE inhibition seems to plateau in the percent knockdown, while the siRNA continues to demonstrate significant dose response, peeking around 90% intracellular knockdown. This type of experiment makes us hopeful that APP-siRNA may be able to impact intraneuronal amyloid levels in a profound way. Shown here is another preclinical experiment, this time highlighting the importance of what's happening inside the neuron. These cells and these experiments come directly from patients who have autosomal dominant forms of Alzheimer's disease. And in these cells, an abnormality is seen in the intracellular space. On the far left, you can see reference to previously published work, revealing that a wide variety of different APP or presenilin mutations can cause a striking enlargement of RAV 5 positive early endosomes in human iPSC-derived neurons. These same studies have shown that this is one of the earliest anomalies in Alzheimer's disease, preceding A beta deposition, and this research has zeroed in on the beta C-terminal fragment, not amyloid beta, as the driver of the endosomal phenotype. So we hypothesized that an APP-siRNA, because of its ability to also lower the beta C-terminal fragment, may be able to improve the endosomal swelling. In the top panels are cells treated with only a control siRNA, and the bottom panel, after treatment with APP-siRNA. These representative images show that swollen enlarged endosomes are significantly smaller when treated with APP siRNA. When comparing to the previously published literature, the magnitude of the rescue is approximately on par with that of wild-type neurons. This data reminds us that much APP processing goes on beneath the surface of the cell and within the membranes of the endosome. And to impact endolysosomal dysfunction, you have to have a mechanism that gets inside the cell. Now transitioning from in vitro cells to in vivo animal models, this experiment was conducted in an AD mouse model. This model has 3 APP mutations, and these mice get characteristic parenchymal amyloidopathy. They also have behavioral changes, which are manifested as traveling longer distances and increased rearing frequency. And in this model, the disease onset is typically around 4 months of age. In this experiment, mice received either a single intracerebral -- they received a single intracerebroventricular injection with either APP-siRNA or a control at 6 months. And then the phenotypic observations were made at 9 months of age. You can see that APP-siRNA-treated animals had significant reduction in distance travel and on rearing frequency. This represents phenotypic improvement with the therapeutic in this mouse model of Alzheimer's disease. Now transitioning from preclinical models to planning for clinic, I'm excited to confirm that we are planning on submitting our CTA for ALN-APP by year-end, with the Phase I expected to begin dosing in early 2022. The Phase I study will have 2 parts. Part A is a single ascending dose design; followed by Part B, a multiple dose design. The selected population for the Phase I study is patients with early onset Alzheimer's disease. This population represents around 5% of the total Alzheimer's disease population, with onset less than 65 years of age. These patients have an increased role of APP protein production at early ages, which strengthens the mechanistic fit for an siRNA. And we think this also increases the likelihood of being able to identify useful trends and disease biomarkers. Also, because these patients are younger, they have fewer CNS and non-CNS comorbidities, which makes it a more homogeneous trial population. The primary objective of the first-in-human study is to evaluate safety and tolerability with secondary objectives focused on pharmacology. As our first CNS program, these will be important learnings not just for the APP program, but also for the CNS platform. Exploratory measures include fluid biomarkers of various APP protein products, tile, phosphotile, different measures of synaptic health neuroimaging and early exploratory cognitive and functional measures. We're hopeful that these Phase I learnings will help inform additional clinical development opportunities. After the Phase I and early onset Alzheimer's disease and demonstration of safety and tolerability, there are multiple potential opportunities to expand into. Under 1 umbrella are the parenchymal amyloidopathies. These include EOAD, sporadic Alzheimer's disease, autosomal dominant forms of AD or potentially other genetic forms of Alzheimer's disease. On the other hand, the other umbrella is that of perivascular amyloidopathies. These include hereditary CAA or the Dutch type of CAA or the more common sporadic CAA. This 1 target, 2 diseases paradigm, it allows for flexibility and strategic options for the program. As additional learnings emerge during development, we hope to be able to be guided further. Let's talk a little bit more about the opportunity in CAA. Unlike Alzheimer's disease, which I think nearly everybody has heard of, CAA is highly underdiagnosed, and it's much less visible in our current health care system. So estimates are that up to 20% of the elderly population have moderate to severe CAA pathology, and this increases risk of stroke and dementia. CAA is the second most common risk factor for intracerebral hemorrhage after hypertension. Some of the neuro imaging clues that CAA is underlying our low-bar hematomas, cortical superficial cyterosis and micro bleeds. And the micro bleeds are typically in a low-bar location in CAA as opposed to the micro bleeds of hypertension, which are usually in the deep brain tissue. In recent years, increasing neuro imaging has improved awareness and it has resulted in increasing diagnosis, yet still only a small fraction of CAA is formally diagnosed today. APP is a genetically validated target for cerebral amyloid angiopathy. We know this because they are well-characterized autosomal dominant forms of hereditary CAA. Most notably is the Dutch type of CAA, which is an ultra-orphan genetically defined population that mostly resides in the Netherlands and Australia. In this disease, APP mutations result in amyloid deposits in the vessels, usually during the 30s, 40s or 50s. CAA also has helpful biomarkers to inform clinical development. This includes similar target engagement biomarkers to that of Alzheimer's disease that can be measured in CSF, but there's also neuro imaging measures of neurovascular reactivity or also called BOLD fMRI. Overall, CAA represents an important and complementary development opportunity for ALN-APP's mechanism of action could serve as a disease-modifying therapy for this very high unmet need population. We've also had an opportunity to learn from CAA animal models. In this particular model is a transgenic rat model with Swedish, Dutch and [ Iowa ] mutations. These experiments were performed in collaboration with the Van Nostrand Lab at University of Rhode Island. This model recapitulates several important features of CAA. First, it has overexpression of human APP throughout the brain. Second, on the right, you can see that there is accumulation of microvascular amyloid. And this is progressive and occurs over several months. This replicates the perivascular amyloid seen in the human disease process. Third, this model also recapitulates some of the vascular dysfunction, including the presence of micro hemorrhages in the cortex, hippocampus and thalamus, all driven by the APP mutations. On the next slide, we're sharing some preliminary data of robust target engagement in this model. These rats received a single intrathecal dose of 0.9 milligrams of APP-siRNA, and then these are representative sections from the hippocampus at day 28. The detection antibodies here include a neuronal signal in red and human APP signal in green. In the top figure, you can see the substantial human APP protein that's amassed within and surrounding the neurons of untreated animals. In the bottom, you can see robust elimination of human APP protein. This demonstrates convincing in vivo pharmacodynamics in a disease model. Additional studies in this model are currently ongoing. At Alnylam, we are excited to be moving ALN-APP and our CNS platform to the clinic. In the coming months, we hope to make history. We have an opportunity to be the first siRNA delivered to the human brain, a door that once opened, we hope many others will follow. With ALN-APP, we also have the opportunity to bring the first investigational therapeutic and clinic to specifically target APP mRNA. I hope we've highlighted today how important we believe it is to be upstream in this disease process. ALN-APP also has the opportunity to be the first investigational therapeutic preventing synthesis of all APP fragments, including the non A beta drivers of AD like the beta C-terminal fragment we showed you earlier, causing the endosomal enlargement in patient cells. Lastly, ALN-APP represents an opportunity to be the first to do all of this from the inside-out. Knowing the tremendous burden and unmet need of these neurodegenerative diseases, we're excited to bring our innovation with RNAi therapeutics to diseases of the CNS. The advances in our CNS platform and now the opportunity to bring this technology into clinical trials has been a tremendous source of pride for the Alnylam team. We look forward to the continued progress in the fight against Alzheimer's disease and cerebral amyloid angiopathy. Vasant, I'll turn it back over to you.

Thank you, Bret. This is so exciting, exciting to see that our first program is going to move into CNS with ALN-APP, what an incredible target and the unmet medical need can serve with this program. So moving on then from CNS to the other extrahepatic tissues that we are working on. So we remain very excited about the work that we're doing with Regeneron for the ocular indications as well. As part of our collaboration with Regeneron in CNS and ocular space, we are working very diligently to move the ocular efforts as well. So very briefly here on some of the data for eye, we have shown very potent knockdown in the NHPI with the TTR molecule. And the dose is high here in the range of micrograms. So 3 or 30 micrograms of siRNA given in the NHPI has profound knockdown. And not only this is very potent, and it's on the next slide, this effect is where you do ramp up out for a few months. And the impact here is seen only after the conjugated siRNA -- the unconjugated siRNA has very little activity in comparison. Now this common feature that we are seeing here in liver, in CNS, in eye, the durability appears to be the main thing, main thing in the sense that we are seeing this benefit across the board. Now if you go to the next slide, I'll talk briefly about our lung program. So here as well, we are making progress. We talked about some of our efforts on the ALN-COV. But on the next slide, what we show is that with our siRNA delivery for the extrahepatic applications, the same delivery system is also showing very good knockdown in lung, this is the data in mice. It is dose dependent in different, different cell types. And again, the durability is seen very well as well. The single dose here and the lungs data endpoint was at day 58, I believe. So this was for SARS-1. Similarly, we also saw the activity for ACE2, a very important receptor for the SARS-CoV-2. Now with lung, this is another issue where we can find different applications. There are a number of lung targets coming from the human genetics across different indications, including asthma, nasal polyps or COPD. This is consistent with our strategy -- or the R&D-focused strategy of having -- or going after the targets that have human genetics behind them and the delivery systems that have shown preclinical efficacy and safety. So I hope based on what we have all shown here, that the Alnylam pipeline will be sourced for the sustainable innovation. Our product engine is humming, not just in liver where we have a number of approved products, many in the late clinical stages and for a few in the preclinical or the early stages as well. But now we are expanding the RNAi to CNS, eye, lung. And this is -- again, this is just the beginning. RNAi is not limited to any particular tissue. It's obviously the question of resources as well. But these are the places we believe there is unmet medical need, validated targets that we can take our technology, all the advances that we have done and come up with 2 to 4 INDs per year for a large number of opportunities, and all driven by this organic growth. And we'll obviously complement this by alliances as needed, as we discussed earlier, like PeptiDream and others, but this is for advancing the RNAi therapeutics for the indications in a variety of different issues to serve the unmet medical need. So in summary, the advances that we did by staying focused in liver, those advances are paying off for delivery in the extrahepatic space as well, very, very important. We believe that RNAi therapeutics is well suited to address CNS disorders with high unmet medical need, not just CNS, but same thing in eye, in lung and other tissues as well. We believe that ALN-APP presents a huge opportunity in the Alzheimer's disease. And as Bret mentioned, we're very happy to confirm that the initial CTA submission is expected by the end of the year. So fingers crossed for making this molecule to go in clinic. And our organic product engine stands to deliver long-term sustainable innovation across multiple disease areas and organs. And as I mentioned earlier, this also provides an opportunity for other tissues, the ocular or lung disorders. Now with that, I will -- I'd over to Josh. I believe there are a couple of other comments, and then we'll get back to the question-and-answer.

Thank you, Vasant. [Operator Instructions] So I see that there is some coming in. So Vasant, I'll turn it back over to you to moderate.

Yes. Thank you, Josh. So let's begin with the Q&A session. We can start with some of the questions that we're getting around the technology. And let's start with the first one. So Kirk, what contributes to the durability of your CNS-directed siRNAs? What enables the potential for biannual dosing?

It's a good question. So we've seen potent dose-dependent knockdown across the CNS cell types in preclinical studies, which I shared today. This includes cells and regions of the spinal cord, cerebellum from the cortex and even deep brain regions such as the striatum. We've recently published the preclinical work investigating the mechanism of durability with our GalNAc-siRNA conjugates, which Vasant pointed out earlier in the hour. Our data supports this depot effect hypothesis in which siRNAs trapped within the acidic intracellular depot are slowly released to maintain functional risk levels needed for this durable silencing. While the uptick mechanism could be different with C16 conjugates, we believe it also involves the endocytic pathway and thus could provide a [ depo ] effect similar to what we have seen with GalNAc. The CNS durability data in NHP studies is very encouraging with stable, detectable siRNA and corresponding siRNA served silencing in the CNS out to at least 6 months. It will be very interesting to see how it translates to humans. I would also note that studies demonstrating that GalNAc siRNA durability tends to improve when we go from preclinical species from rat to nonhuman primates to humans.

Yes. That's a very important point, Kirk, that we have seen this improvement as we go higher in the species. So there is another question on the similar lines actually with IKARIA. So the question is, what is the difference from ESC+ to IKARIA in chemistry? And maybe I'll take that one. So just this week, actually, we shared some of the preclinical data on IKARIA at the OTS Meeting. These slides are available on our IKARIA platform. So as it goes with all our advances that we are doing from STC, ESC to advanced ESC, ESC+ and now IKARIA, I know it's mouthful, but in every version of these, the defining features of the previous template are taken into the next one. So this is exactly what we're doing with IKARIA as well. IKARIA constitutes all the features like the metabolic stability or the specificity that is coming with the ESC+ platform, combining all of these things together to give the molecules that have much longer duration of effect. And since IKARIA has a specificity and we believe very excellent safety profile based on the preclinical data, we can also have the higher dose and that together can give us deeper knockdown as well as a longer duration of activity. I hope that addresses the IKARIA question. So maybe just again, continuing on the technology theme there. Kirk, question for you. What are the limitations of the C16 conjugates? Is there a differential uptick towards the different cell types of the CNS? Or is it same? with GalNAc-siRNAs is largely 1 cell type. But here, you are looking at a very different type of tissue. What are your thoughts on that?

Great question. You're absolutely right. I mean, with liver, you're essentially laser-focused to hepatocytes with GalNAc, where the CNS has multiple cell types that are expressing targets that we're interested in. We believe that the C16 conjugation provides opportunity to reach all of these cell types and achieve effective silencing by our siRNAs. And we shared today as we've observed activity both in vitro and in vivo studies in the following cell types, such as neurons, astrocytes, microglia as well as oligodendrocytes. So we're encouraged that we're able to both access and engage risk in a variety of cell types in the CNS and importantly, see subsequent and corresponding target lowering in each of those cell subtypes.

Great. Thank you, Kirk. So maybe now moving on to a very exciting topic of ALN-APP. And Bret, there's a question here on what is different with ALN-APP compared to the other strategy of amyloid targeting by antibodies?

Yes. So amyloid-targeting antibodies, of course, are primarily targeting amyloid deposits outside of the cell, and they help facilitate the degradation. In contrast, right, ALN-APP is targeting mRNA, which is inside the cell, and it's well upstream of the amyloid process. In addition, antibodies typically are designed around a particular epitope. So whether that's A-beta monomers or soluble forms of aggregates or insoluble forms or fibrils or plaques, right, they all have a particular target, which is usually an A beta assembly state [ specie ], whereas ALN-APP, right, has a much more comprehensive lowering of really every APP protein that's downstream. So that's really the key difference is the where and then the comprehensiveness of the approach.

Thanks, Bret. On similar lines, you mentioned that mutations decreasing the APP processing are productive for AD. So what evidence is there for this and how has this been determined?

Yes. This is a good question. So I mean, I think that there's a lot of evidence on the pathogenicity side, right, of these genetic bookends. We were talking about -- and those come from clear mendelian diseases where mutations cause disease as in those that cause early onset Alzheimer's disease and those that cause CAA. But I think your question specifically asking about the protective variance in APP. And probably the most well characterized is this A673T mutation, which is also known as the Icelandic variant. And this has been found to have a protective effect when you do case-control studies in different populations of Alzheimer's. I think actually one of the first observations of this [ oligo ] was in a somewhat a finished background, who I think was -- died around the age of 104. And there was very little amyloid in the brain, which was one of the first clues. They went on to look at this mutation, the A673T in, in vitro studies where they see about 40% less production of APP. So it's really helpful to have kind of these, on the one hand, mutations that clearly cause disease. On the other hand, these mutations that clearly protect against disease, these kind of genetic bookends give us an idea that there is a volume up here, and we have a platform that is good at turning that volume up.

Great. Great. Maybe another question, Bret and Kirk, maybe both of you. I mean the discussion on ALN-APP or the Alzheimer won't be complete without remembering the approval of Aduhelm. How does that impact the plans for ALN-APP development? Obviously, this has created a lot of buzz in both ways. How do you look at that approval and what the regulatory agencies are looking in approving the drugs like this?

Yes, thanks. I mean -- so obviously, right, like everyone else, we've been closely watching the impact of the Aduhelm approval. And our team is actively looking at the impacts. I mean I think when we review the data from that trial, right, we -- I think certainly, there's still an unmet need and an opportunity for further intervention for patients. So we plan, in the Phase I, to explore a wide range of different biomarkers of disease progression, including amyloid PET, which we'll look at in the Phase I study. So that will help us inform kind of the clinical and regulatory strategy. And also, I think we can't understate that even with the approval of Aduhelm, there still remains really significant unmet need here. And so finding other therapeutics that are working by other mechanisms to hopefully slow down or halt or reverse the progression is something that we still think is very important.

Thank you, Bret. I think there are other questions on the PeptiDream as well the collaboration that we just signed. And the question is, what's the rationale behind this deal? So maybe I'll take this one. So as the leaders in RNAi space, we're constantly seeking to innovate and advance our platform to new frontiers. While a significant amount of our innovation is organically driven, coming from our internal efforts, we are also very receptive to opportunities that enhance our platform's capabilities and broaden the scope of potential value creation. So in that sense, our collaboration with the PeptiDream exemplifies the philosophy that it has a potential to create numerous opportunities to deliver siRNA conjugates, beyond liver, CNS and eye. If successful, this would allow us to address multiple areas of unmet medical need in a variety of therapeutic areas. Now I just want to clarify that. This does not include -- exclude CNS or ocular. This collaboration is open for any different receptors of our choice. Let me now look into the other questions that might help come up in the chart. There's a lot of interest on the ALN-APP side. And so maybe I'll focus on that one and go back to Bret. So what's the size of addressable market for ALN-APP?

Yes. So Alzheimer's disease, right, is the most common dementia worldwide. So as I mentioned earlier, 5 million people in the U.S. and 30 million globally. And even the early onset form of dementia, which we're looking at in the Phase I, this still represents 5% of that very large number. Cerebral amyloid angiopathy is also really common, but it's very underdiagnosed. And of course, this is the disease that causes strokes. A recent meta analysis estimated that the moderate to severe CAA is found in about 23% of elderly individuals and CAA has also found in about half of everyone who has had a low bar intracranial hemorrhage. So we're hopeful that advances in the diagnostic criteria and increased use of neuro imaging will increase the diagnosis of CAA. And we think that if that happens that this kind of -- we'll be able to help accelerate the opportunity a little bit for ALN-APP and CAA.

Thank you, Bret. I think moving along here now in the different topics, and this one would be on the status of the ALN-COV program and what's the rationale behind discontinuing this program. Maybe I'll take this one. As we showed earlier that with the advances that we have made in the extrahepatic delivery, we see pretty good activity in lung as well. And these were the basis for us to move into this program and find the ways by which siRNAs can be utilized against the SARS-CoV-2. We believe that this was a -- this is the unique space for RNAi to go after the [indiscernible] like this. We have done this one for the hepatitis B virus that is in alliance with real biotechnology. But we have made the portfolio decision to discontinue ALN-COV in development for the treatment of COVID-19. And this is largely based on the availability of highly effective vaccines and alternative treatment options. But we believe that whatever we have done in this program, these learnings will be applied for the lung targets. There are plenty of lung targets, and we talked about some of them. We didn't take the names of those, but there are quite a few in different indications for which there is human genetics behind them that we can go after. All right. I -- Pablo and Chris, I think we have time for a couple of 1 more question. Okay. Well, Bret is still in demand here. So I will end with Bret on the third question. So Bret, given the experience of the BASE and gamma-secretase inhibitors, why would you think that ALN-APP be safe and effective when these small molecules fail?

Yes. This is a good question. So beta and gamma-secretases are their enzymes, and they have a really wide range of physiologic functions, and therefore, have a wide range of substrates. So both beta and gamma-secretases are really, these membrane-bound metalloproteases. And recent proteomic studies have shown that beta secretase has over 50 substrates and gamma-secretase has over 100 substrates. So I think what's important to note when we're first considering the question is that when you're inhibiting these enzymes, you're doing a lot more than just lowering A beta. I mean, in addition, the BASE and gamma-secretase inhibitors, they reduce A beta by basically causing an increased amount of substrate upstream. So just like any metabolic cascade, when you inhibit 1 enzyme, you do decrease the product, but you also have another consequence, which is increasing the accumulation of the APP fragments upstream. And so in this case, we know now that some of these other APP fragments that are being shunted to different pathways are also negatively impacting neuronal activity. I mean, lastly, I'd say, base inhibition itself has been shown to cause a paradoxical increase in BASE 1 protein levels. And so this increase in BASE 1 protein may have confounded some of the potential therapeutic benefits. So in contrast, right, ALN-APP targets mRNA. And so it lowers the production, of course, of APP, but it doesn't cause any of those expected compensatory changes that you get when you're inhibiting a particular enzyme. So for these reasons, we expect the clinical profile of ALN-APP, it may be significantly different than of BASE inhibitors.

Thank you, Bret. Fingers crossed. I think that's all the time that we have today. Thank you, everyone, for joining us. Thank you, Kirk and Bret, for going over the extrahepatic advances.

All right. Thank you, Vasant. And I'll echo my thanks to Kirk and Bret as well. And thanks to all of our listeners. That does it for today's RNAi Roundtable. And I would like to remind you that the replay and slides from this event as well as all the other roundtables we've posted in this series can be accessed on the Capella section of Alnylam's website. I'd also like to remind you that Alnylam will be hosting its 2021 R&D Day on the morning of Friday, November 19. And there, we will review progress across our pipeline of investigational RNAi therapeutics. This will be a virtual event. We hope that you can join us on Friday, November 19. So thanks, everyone, and have a great day.

Thank you. This concludes today's conference call. Thank you for participating. You may now disconnect. Goodbye.