Good afternoon. Thanks for joining us for another session at the 42nd J.P. Morgan Healthcare Conference. I'm Brian Cheng, one of the senior biotech analysts here at the firm. I'm joined by my associate, Sean Kim, who's also in the audience. On the stage, we have Intellia's management team. I'll pass the mic to John Leonard for a short presentation, followed by a live audience Q&A. John, the stage is yours.
Thank you, Brian. Can everybody hear me okay? Just want to make sure. So it's a real pleasure to be here to tell the story of Intellia, and I've got two of my colleagues, Glenn Goddard, our CFO, on the end there, and David Lebwohl, who's our Chief Medical Officer. If you have any questions, they'll participate in some of the things that may come up. I'll be making some forward-looking statements today. This is the disclaimer that we make available. The company is approaching its tenth year. It was formed shortly after some seminal work that was done in the CRISPR field.
What we've been setting off to do in this journey is to find ways of making what has become Nobel Prize-winning science in medicine, and I'll give you a progress report of what we've been doing and where we are and where we think we're going. At this point, we would say that we're poised to have the first-ever in vivo CRISPR therapy come to the market as we begin our first pivotal Phase III program for TTR amyloidosis. If we hit the timelines that we're working for, we expect to have later this year even a second in vivo Phase III program for HAE, and we'll talk a little bit about some of the data that lies behind that. Yeah, right here. So is that what you're looking for?
Yeah.
Okay.
Yeah, we're showing other slides.
Oh, okay. Are you seeing something here? Okay, anyway, sorry. One of the things to keep in mind is that, we're getting a very clear notion of safety and efficacy of these products. We've now dosed in excess of 100 patients from the... Oh, boy! Was it something I said?
Oh. Whoa! You're back.
You're back, John.
Okay, so as I said, we're some of the safety and efficacy profiles are coming into focus. We now have over 100 patients that we've dosed across the two indications. We're building a pipeline that goes beyond these programs in both the in vivo and ex vivo space, and we'll say a word or two about the platform that we have, which we think is the most comprehensive gene editing toolbox in the industry. Just by way of introduction, for people who may be new to the story, I think it's really important to understand just a you know, fundamental aspect of why people are so excited about CRISPR, and CRISPR in whatever derivative form exists. And it really comes down to the ways depicted on this slide. It's a very simple, modular, two-part system.
In that complex of two parts, there's the guide RNA, as depicted here, and Cas protein. A simple way of thinking about this is the guide RNA is essentially a GPS system that confers incredible selectivity, and brings with the potency of how the complex will behave. It guides that Cas protein, which in its native form, acts as a nuclease to cut DNA. But importantly, that protein also serves as a platform for the various derivative forms of genome editing that have come to the fore in the last several years. What you see on the right is how that protein complex behaves, which goes to the nucleus in a very top of the cascade of genetic information, modifies genes, and now we're able to modify genes in any number of different ways.
That's captured here, and I think it's a really important notion to understand because there's different names and different capabilities that have been developed. All of them, and I mean all of them, begin with CRISPR-Cas. It's this notion of the selectivity that comes with the guide RNA and that, that platform protein, Cas, either in its nuclease form or as the platform to add other enzymatic activities. So whether you're doing base editing, whether you're doing DNA writing, and there's different forms of that, or other forms of technologies to bring genetic change into the nucleus, all of them rely on CRISPR-Cas, they all rely on guide RNAs, and it's impossible to do those well without being proficient in understanding this protein guide RNA.
We have all of those different capabilities and proprietary forms in our toolbox, and we're at a point where we can choose the best tool for the particular genetic challenge that we're looking to pursue. You can think of the archetypal sorts of edits, which are depicted on the right, whether it's the knocking out of a gene, you can search genetic material, you can repair at the level of codon, and you can combine and mix and match this, especially in the ex vivo setting, where taking cells outside the body, we can make any number of different combinations of that. We've made a lot of progress with that. Again, the idea is that having all of these tools in the company, we're able to choose what we need for whatever might happen. ... All right, back up. All right, back up.
Oh, where were we? Try to do this fast before something happens. Anyway, I mean, just to, you know, this is where we left off, with different tools in the toolbox and the archetypal sorts of edits. You're only as good as the delivery you have. And it's, you know, absolutely necessary to have a variety of different delivery tools, which we've also been working to collect at the company. First and foremost, we use a chemical approach in the form of lipid nanoparticles, some of which can be functionalized to go to places beyond the liver, in particular, the bone marrow, et cetera. We also have biological approaches in the form of AAV, and this can be modified with other technologies as well.
Together, what we've done is put ourselves in a position on the in vivo side of things, where now accessible to us are the liver, bone marrow, and with collaborators working to central nervous system, eye, and muscle. And all of these things figure in our priorities for the next several years. As we look back over the last year, in terms of the progress that we've made as we've been building our company and moving towards these pivotal trials, it's been a very, very successful year with especially moving our lead programs forward. NTLA-2001 is targeted at TTR amyloidosis and NTLA-2002 at hereditary angioedema. And what we were able to do in the last year was clear the first in vivo CRISPR IND, and in fact, the second CRISPR IND for both 2001 and 2002.
We've advanced the 2001 program for an amyloidosis to phase III, and we'll talk a little bit about the MAGNITUDE trial, which is the pivotal program for TTR cardiomyopathy. In addition, last year, we began and completed enrollment for the phase II program for the HAE study on 2002, and along the year, we presented interim data for both of those programs that has been remarkably consistent and remains highly positive. We've also moved beyond knockouts to the liver, to knock-ins in the liver. It's this idea of reconstituting genes and have submitted a CTA for NTLA-3001 for alpha-1 antitrypsin deficiency associated lung disease. And as part of that platform of different approaches, we've continued to make progress with the various tools that we have, especially with DNA writing technology. We begin the year with over...
with approximately $1 billion in cash, which we expect that to carry us well into the middle of 2026, and that will take these clinical programs either to completion in the case of 2002 or well beyond enrollment in the case of 2001. The strategic priorities for the company broadly, as we think over the next 2-3 years, really fall into 3 broad categories: executing those pivotal trials, launching the next wave of in vivo and ex vivo clinical programs as we expand beyond knockouts in the liver, and continuing to build out the platform. So going a little bit more deeply into the pivotal trials, and as I said, we expect to complete patient enrollment in this time period. And we anticipate that we'll be submitting a BLA for hereditary angioedema sometime in 2026.
The next wave of in vivo and ex vivo clinical programs are looking for proof of concept for that gene insertion, as well as initiating clinical development on the ex vivo side for what we think is a breakthrough allogeneic approach that, if successful, should enable solid tumor, cell-based engineering. And then finally, we'll continue to build out the editing and delivery modalities by moving those different delivery approaches beyond the liver, as I mentioned, and advancing the DNA writing technology to our platform. So key milestones for the upcoming year, by program fall into dosing their first patients for the phase III trial for cardiomyopathy, continuing to expand the enrollment of that program as expeditiously as possible, and moving the companion indication of polyneuropathy forward as we prepare for a phase III program.
And again, along the way, we'll present data as the cohorts originally treated continue to mature. For 2002, the hereditary angioedema program, we anticipate that we'll be initiating a phase III study in the second half of this year, subject to final agreement with the FDA and other regulatory agencies. And we will continue to expand on the phase I data that we've. I will review with you very quickly, as well as presenting the phase II data that will be completed here in time this year. And then finally, as I said, we anticipate beginning dosing for the gene insertion program for AATD. So let me hit just a couple of highlights for the clinical pipeline with those two programs, and we begin with just a summary safety statement.
With the 75 patients that we've reported on for phase I studies, 2001 and 2002, we've been struck by what is a very favorable safety profile. We've now observed patients in the case of 2001 for up to 2 years and for 2002, it should be HAE program for over 1 year, and that continues to mature. When adverse events occur, they tend to be very mild in severity and typically consist of infusion-related reactions right around the time of dosing.... The effect of editing has been followed by following surrogate markers in the form of the proteins that are encoded by the targeted genes. And here what you're seeing is the effect of editing the kallikrein gene, which is related to attacks of hereditary angioedema.
What you're seeing on the Y-axis is the decline in the circulating levels of that protein from baseline in a dose escalation study. So starting at the top, you're seeing the low dose, then the intermediate dose, and the highest dose. What you see is that relative to that dashed line, which constitutes benchmark levels of inhibition, typically can, conforming with, Takhzyro sorts of levels of inhibition. All of those levels, but especially, beyond the initial dose, are, are beyond the, the benchmark approach. So we're very, enthusiastic and excited about the level of editing that we've been able to achieve. That reduction in kallikrein protein can be, assessed by attack rates, corresponding to those same patients.
What you're seeing here are those three groups, and I just call your attention that it's the lowest group, then the highest group, and then the middle group, and that's a function of the chronology of the study. What you're looking at is the attack rate during the screening period, the administration of the drug, several-week observation period, and then the continued follow-up of these patients. So we count in that observation period, the number of attacks. But what I think is important to note is that every single patient achieved a no-attack status. In those two instances where patients had, you know, attacks essentially every other day, it took a little while longer, and we don't really understand why that's the case. But all of those patients achieved an attack-free status.
I call your attention to that top patient, which is in the lowest dose group, far out on the right, you see an attack. It was actually a sports injury, where individual had abandoned all prophylaxis, was playing soccer, had a sports injury, some swelling of his hand. Whether it was HAE or getting hit with a soccer ball, we don't know, so it was called an attack. But I think the remarkable finding here is that all of these patients have reached a point where they can abandon the prophylaxis they had and maintain an attack-free status. We will update these efficacy data as we go forward, but this is the basis for a lot of the excitement that we have.
As we think about 2002, as it goes forward, that was phase 1, where we've chosen two doses from that work. The low and the intermediate dose, 25 milligrams and 50 milligrams. And we're completing that phase 2 study, which will serve as the basis for a dose selection for the ultimate pivotal trial that we expect to begin the second half of this year. And as I said, we will present those data when we have them and continue presenting the follow-up to the patients. As we turn to 2001, which is the program to knock down the pathogenic protein that causes amyloidosis of the TTR type. What you're seeing here are the results of a similar sort of design, which is dose escalation for cardiomyopathy indication and the polyneuropathy indication on the right.
What you can see is with the various doses, in all cases, as you get to the higher doses, we've been able to knock down TTR levels to, you know, extremely low levels as a function of change from baseline. These are in the order of 90%-95%. And what's, I think, important to realize is that we're at the point where we're measuring the absolute levels of the protein. And what we found is that, the residual absolute TTR concentration at day 28 is about 17 micrograms. To our knowledge, this is the lowest that's been reported and seems to be about a half or a third of what's found with silencers, where we've been able to see those kinds of numbers.
I do think as this program goes forward, as people think about TTR and the proper treatment of that disease, these numbers and these measurements will be very, very important. One thing to keep in mind, just to remind you, is that these patients have been treated a single time, at time zero. What you're seeing are the results that are remarkably consistent over the observation period, where it's 2 years on the right, as we begin the polyneuropathy, and it's 1 year in cardiomyopathy patients. Again, these data will be updated as the year goes on. That information goes into the dose chosen. It's 55 mg fixed dose for a phase III program, which is now called MAGNITUDE.
This is a randomized, double-blind, placebo-controlled study, which is essentially adding NTLA-2001 or not to the standard of care, including tafamidis. You see, it's 765 patients that we expect. It's randomized 2 to 1, drug to placebo. I think that's attractive to patients and should drive the rapid approval of it. And the primary endpoint consists solely of cardiovascular-type endpoints, whether cardiovascular mortality or a variety of different cardiovascular related consequences. There are secondary endpoints, as shown here, including TTR and quality of life. And, in contrast with some of the other studies done, what we're doing is allowing somewhat sicker patients in with higher levels of NT-proBNP at baseline, which we think will add to the rapid completion of the study.
They'll be stratified by disease severity, wild type versus mutant of TTR, and then again, with amyloidosis. We estimate that the average duration of patients on the trial will be about 30 months. All patients will need to be followed for at least 18 months over the course of the trial as we count the events. So, as we see in NTLA-2001 going forward, we're imminently to dose patients. Much of the activity this year will be opening sites and enrolling patients as quickly as we know how to do, and we will progress the companion indication polyneuropathy as we do that work, and over the course of the year, present some of the data. We do think both of these programs, when successful, have tremendous commercial potential.
If you look at what is happening for the amyloidosis market, we're now on our way to a projected $11 billion+ marketplace. There's over 500,000 patients estimated now. Diagnostic tools have come to the fore that allow us to find these patients. And when we think of what the current cost of treatment is for the existing therapies, you see in excess of $450,000 a patient. Similarly, for NTLA-2002, which is a less prevalent condition, about 20,000 patients worldwide. It's a very, very expensive treatment paradigm, and with that, a global market that is estimated to be on its way in excess of $6 billion.
And again, these are very, very expensive patients, and we think that we'll be able to price our product in a way that will be health resource sparing for the system and very, very competitive here. Finally, as we think about going forward, you know, I think simple ways of thinking about the progression of the program and the company beyond these pivotal trials are the waves of activity. So look for data with respect to in vivo gene insertion, and importantly, how we take ex vivo allogeneic cell therapies into the clinic, and then as we progress the platform of DNA writing and these different delivery modalities outside the liver. A quick look at the pipeline of the company.
You see, later stage programs with NTLA-2001 and NTLA-2002, alpha-1 lung disease here, and then a variety of different research programs that we expect to play out here over the next two to three years, as well as work on the ex vivo side. Finally, just back to where we began, which is the mission of the company, coming up with approaches to solve and create curative medicines for gene-based diseases. Innovation is at the core of the company as we think about having, you know, the incredible tools that CRISPR makes available to us. And from our standpoint, you know, the CRISPR revolution is here to stay, and it's only going to get bigger and more comprehensive as time goes on.
So with that, I'm happy to turn to Brian and take whatever questions you want to address.
Great. Let's start off with the Q&A. For those of you who are in the audience, if you have any questions, you can raise your hand, we have runner on the floor. For those of you who are joining us virtually, you can submit questions on our conference portal.
Do you want to jump in?
No. Oh, okay. Sure. I think last year when we, when we spoke, you know, we hosted a dinner with you, your, you and your team. You haven't started a pivotal study, lots of questions about how the regulatory agency thinks about gene editors, as a whole. When you look at your portfolio today, you, you have taken a pretty big leap across the board, right? You're now in pivotal stage in ATTR-CM. Looking back, what, what have been the biggest challenge to get to the pivotal stage for ATTR-CM? And, you know, as you think about your earlier stage, programs and, and even your peers, you know, that are focusing on base editing, what, what is the potential, you know, uphill challenges that you think you and or maybe even other peers may see?
Well, I'm not going to be able to comment on what goes on in other companies than how they do their work. But for us, you know, I think it's important to remember that we started literally from scratch, and what we did was identify delivery modalities that had been developed for a different approach, siRNA. And so we spent a lot of time adapting that to, you know, the coding sequences and RNA for CRISPR-Cas. A lot of work with basic chemistry, et cetera, for making the LNP delivery be possible to be effective. And that took time because that was working out a lot of basic systems.
The other part that was, I think, really essential, but has served us well, is coming up with comprehensive ways to understand where the enzyme goes and how it behaves when it gets there. And so we've spent a tremendous amount of effort thinking about how to characterize off-target profiles of any of these different guides and industrializing that process. So I think as a testament to how that's played out, of the four INDs that we submitted, one collaboratively with Novartis, every single one of those has used that same off-target methodology, and every single one of those INDs has been cleared in the requisite 30 days.
I think, yeah, that was probably the biggest hurdle, building out that foundational stuff, but it puts us in a really good position to go and do some of the other things that we want to do now.
Great, maybe just one more from me before I turn to audience. I want to start off with HAE. Usually, I think we kick off with ATTR, but now that you have planned for BLA in 2026, and, you know, potential pivotal start also, in the back half, potentially this year, what's your latest view on the on a pivotal design, trial? And, how confident are you that, you know, you can meet that 2026 goal and, and, you know, besides the pivotal study, what, what else do you need to kind of line up to get there?
Well, I am very confident we'll hit the 2026 goal, and David is going to tell us how.
Yeah. First, let's talk about what we've seen in phase one. What you saw is the event, the attacks going to zero. Of course, having that kind of result makes your design of any future study much easier. You're going to have a very strong treatment effect, and you can have a very small trial because of that strong treatment effect. We will have the results of the phase two. The patients have been enrolled. They'll be followed for 16 weeks, so we'll have those results towards the middle of this year. We'll have-- That will help us to choose a dose. We looked at two doses, 25 and 50. Should say, if you look at the data, we think 50 is more likely.
There's a more consistent effect and a deeper effect, but obviously, we want to do the larger group of patients. It involves 25 patients, so now, that will be also supportive eventually of our BLA. The trial we're going to do itself is going to be similar to other trials in this disease, pivotal trials. They tend to be very small because of the strong treatment effect, as I mentioned, probably less, can be significantly less than 100 patients. We saw that the phase 2 enrolled very quickly, in 9 months, at a small number of sites.
We do think this trial can enroll very quickly to phase 3, and that, hearing the investigator enthusiasm, hearing from the patients themselves, we do think that we can enroll it quickly as the phase 2 did, and be able to have the BLA in 2026. The other pieces that are important here that are really all set really are the drug manufacturing is going to be submitted as part of the phase 3 study. This will be the commercial drug manufacturing. The 2001, as you know, went through quickly through the IND. Methodology is very, very similar to what we did for 2001, so we could we also think the manufacturing is in good shape for the pivotal study.
What about off-target, you know, studies and also the preclinical work that you have to do for the reproductive side as well? You know, how confident that you can finish that before, you know, asking for the FDA to see if the phase 3 plan is the right one to move forward?
So looking at those two pieces, the off-target work has already been presented as part of the submission for the phase 1/2. So they, you know, we've been able to have discussions with them because we have RMAT designation. This allows a more free communication with the FDA. We feel that the off-target work we have done is satisfying them as it did for 2001 to go into phase 3. For the women of childbearing potential, we had presented data showing breeding study showing that we're not passing this on through the germline. Some localization studies showing that the editing is not in the ova. So between those two, they did want us to go one step further. It's a study slightly different from the breeding study.
It's a reproductive toxicity study, where you look at the developing embryos. Based on the fact that in the breeding study, the animals were completely healthy. There were no unexpected deaths in that group. We fully expect that to be straightforward and show a good result. And that will be done well before we start the phase 3.
Any questions from the audience?
Keep going.
In terms of, I guess, overall data flow, when we think about ATTR-CM, I think part of us also want to see any functional benefits as well. How soon can we see any potential functional measures? And, you know, like, I think you're tracking the MR of the heart, proBNP levels. Are those mature enough today to for us to get a sense, this year?
Yeah, we think that these things evolve pretty slowly, the MRI, and you see that, for example, in the APOLLO-B data. You saw that there's not much functional change over the first year on a drug. So it will take longer to see the maturity. We haven't guided yet to when this will be mature, but as we have in the past, when we have a body of data that we feel is interpretable, we will bring that forward and talk about the data.
Okay. And where are you now in the MAGNITUDE-2 trial? And yeah, we'll start with that first.
So we announced we've initiated the first site. So at this point, we can start to enroll patients. Not gonna talk about, you know, patient by patient. We'll probably announce when the first patient gets treated after they've been screened and actually treated. But. Then periodically, we'll be giving updates about the enrollment and looking to see that we're keeping that on track.
I think one question about the study is always about the trial size of the study. I guess one question I always have is: How do you think about the powering of the study? And, you know, when you compare it to some of the competitors, you know, when you look at Ionis study, it's also much bigger than yours, right? So how do you think about the powering, and how confident are you around the trial design? And lastly, you know, the potential interim analysis, right? What would unlock that piece as well?
Okay, and-
Go for it.
Yeah. So the trial size, I think you've seen, is similar to HELIOS-B. I think our assumptions are close to HELIOS-B, then, and we can talk about where Ionis is going with a very large trial. It's large enough at 765 patients to see a very important treatment effect for patients in terms of reducing cardiovascular events and cardiovascular mortality. So, we feel confident in terms of the size of the trial. We're gonna have a mix of patients who have mutations and don't have mutations, patients who have more advanced disease and less advanced disease, patients who are on tafamidis, and patients who aren't on tafamidis. It ends up being a similar design to the other style of studies. We have about 50% of the patients on tafamidis. We expect to have about 50%.
We've looked to patients who are a little bit higher standard in terms of having greater medical need. They have a NT-proBNP of greater than 1,000. So this will be a group that we really feel we can demonstrate an important benefit for them in terms of their medical needs. I'm trying to think of the other question.
On the interim analysis.
Yeah, yeah.
Yeah. So the interim analysis is built into the study. The idea there is, there's a possibility if we feel we're gonna have a-- we do feel we can. The hypothesis will have a better treatment effect than the silencers because instead of average of 80% reduction, we're getting a very consistent 90% or greater reduction, and that difference is, as you can imagine, getting, as mentioned, getting two-thirds or one-half lower in terms of the TTR. That can make a big benefit in terms of the heart function. At the interim analysis, this could be a positive trial, and that's really the idea of an interim analysis, to try to pull in the results by about a year and hopefully the registration by about a year.
We have a question from the audience.
Thank you. Thanks for the presentation. You mentioned about this, delivery, AAV delivery or other technologies as well suited tissues like, CNS or eye. Given the size of the AAV packaging limit is 4.7 kb, and, how do you tackle that to really deliver the CRISPR-Cas, you know, to the target tissue like CNS, eye with AAV?
Yeah. Thanks for the question. We, as I said at the beginning, have a variety of different Cas, SpyCas9, Nme as well, and it turns out that the Nme is a shorter coding sequence. And so we're able to insert that into AAV, along with the guides, and have that all fit in a way that is an all-in-one kind of format. In addition, we've built out self-inactivating technology of that virus. So as it gets into a cell, it can carry out the editing function and then self-destruct, if you will. So I would anticipate that most of the early indications we would go after would be knockouts, because that's the format that's ideally suited for that. Inserting genetic material or, you know, other sorts of approaches will require different sorts of delivery methodologies, which we're working on as well.
So, I just point out that the AAV approach we're taking is done collaboratively with Regeneron, at least for the brain and for muscle. And they're experts in antibody technology, as I'm sure the audience is well aware. And so this is an approach that uses an antibody-mediated way of getting across the blood-brain barrier or into muscles, and is detuned so that the liver, which typically takes up a lot of the virus, doesn't do that. So this has been demonstrated already, and our contribution to the program is those innards, if you will, the all-in-one format and self-inactivation.
Any other questions? When we think about the potential functional benefits, the difference specifically, you know, since half of the patients are gonna be on Tafamidis and the other half not, do you have some insights about just how adding NTLA-2001 could... You know, what the delta is, right? In terms of your primary endpoint benefit.
Yeah, the way we're thinking about it is, of course, some of the patients will not be on tafamidis, and the benefit is it's just gonna come from the great reduction of TTR that we're getting to very low levels. In the patients on tafamidis, this also can be potentially synergistic because you're getting the TTR to very low levels and then stabilizing the rest of the small amount of TTR that's left with the tafamidis. The increment, obviously, we don't know until we study that. We do think the increment will be similar in the two groups of patients, and that's what we're looking to prove in the phase 3 trial.
Returning to AATD lung, we were planning to file the CTA sometime this quarter. It's earlier than you have planned even. So can you remind us what you're looking for in terms of the AAT level that you're trying to normalize? Are you still shooting for... 'Cause I think heterozygotes already are—it's already viewed as good enough, right? So, is that where you're aiming for, or are you shooting for higher?
... Yeah, we so what we've shown preclinically in the non-human primates is that we can achieve what would be called normal levels, which is probably more than heterozygous or the upper limit of heterozygous. What we understand, talking to pulmonary experts, is the higher, the better in this disease to protect the lung better. It's certainly an important piece of this. So we do think getting to that type of level, which is 2,000, would be valuable to patients, and it's our target. You're right that we don't know that we need to get that high. We'll obviously be looking at the data and deciding, are we getting to high enough levels to push the program forward?
Okay. And then, in terms of just the registration path, conceptually, should we think of it as very similar in how, as how you approach it with ATTR and HAE? You get some biomarker POC, then move to a larger set of data. How, how different is the path forward for AATD lung compared to what you have done so far?
So there've been much fewer approvals in AATD lung. There's not a real well, well, way forward, the way there has been for ATTR and HAE. What we think the FDA is looking for is to... If we could achieve normal levels of alpha-1, probably also, of course, show good safety, maybe show some early effects on, on lung markers like, FEV1, that a single-arm trial would be a way to get approved. It may require randomized trial and follow-up to that. You know, I don't think they, they really know at this point in this disease, but that, that's what we expect. The most important thing being to achieve normal levels of, of alpha-1 antitrypsin.
Great! We look forward to a year of a lot of catalysts coming.
Mm-hmm.
Thank you so much for your time today. Thanks for joining us. Thank you.
Thank you.
Thank you.