Roivant Sciences Ltd. (ROIV) Earnings Call Transcript & Summary

Good afternoon, everyone. My name is Yatin Suneja. I'm one of the biotech analysts here at Guggenheim. It is my pleasure to welcome all of you to our 5th Annual I&I Conference. This is day 2. And our next presenting company is Roivant. From the company -- we have a couple of executives from the company here in our room, but I'm going to be chatting here with the Chief Executive Officer, Matthew Gline. Matthew or Matt, why don't you maybe kick us off, make some opening comments, tell us what's going on at Roivant, what are some of the up-and-coming milestone that we should be focusing on? And then we'll go into a little bit more of a fireside session after that. Matt?

Perfect. Yes. Thank you. Thanks for having us. Thanks, Guggenheim. It's a great privilege to participate in the conference. And obviously, thanks for all the great coverage as well. It's exciting to be here. It's a pretty wild moment for us as a company. So it's been a crazy 2023. Over the course of the last year, we've generated an enormous amount of clinical data. We've generated data on atopic dermatitis for our commercial program, VTAMA. We've generated data in our anti-FcRn antibody that supports a potential best-in-class profile for that drug. I'm sure we'll talk much more about that, that one and Immunovant. And then we've also generated quite a lot of data over the last 12 months in our anti-TL1A antibody, RVT-3101, for the treatment of ulcerative colitis. And the reason I say it's a wild moment is because just about 2 weeks ago, we announced that we were selling that program to Roche for $7.1 billion in total, of which about $5.3 billion or so comes to Roivant. So a really exciting year. It feels wild that we're sort of coming to the end of it. We still have data to come this year, believe it or not. We've got data coming this quarter in brepocitinib, our dual inhibitor of TYK2 and JAK1 in SLE. That's coming pretty soon. And then we've also got data -- 2 different pieces of data coming in our FcRn franchise. One is the sort of final piece of the multiple ascending dose data for IMVT-1402, our next-generation potential best-in-class FcRn. And then we've got data coming in Graves' disease in our original bato in our first-generation FcRn antibody. It would be the first time anyone has generated data in that disease for an FcRn. So it's a really exciting set of data.

Got it. Very good, Matt. So let's dive into the program. So maybe starting with the TL1A first. Obviously, you were able to transact on that recently. It created a lot of value in a very short amount of time. Help me understand the reason to out-license that. And the reason I ask you is that because I have gotten mixed feedback from some investors, but I would love to hear from you how -- what you were thinking? How you were thinking? And what are you going to do with the cash that you have now been able to generate?

That second one is the question of the hour, as it were. So I'm happy to take a crack at answering it, although we've said we expect to be patient. In terms of the why, look, the TL1A is a great program, there's no question, and I'm -- a part of me is sad to hand it over to Roche who, I think, are going to do a phenomenal job with it, but it's a potentially transformational kind of program. And the truth is -- for those who don't know the history, so we in-licensed this drug from Pfizer less than a year ago actually, and we paid $0 up front. It was a sort of profit-sharing kind of arrangement where Pfizer kept 25% of the program, and they kept European rights, and we had U.S. and Japan. And things just moved really quickly for the class. We and Pfizer together generated some great Phase IIb data. Prometheus, as you're already familiar with, generated some great Phase II data. They got acquired by Merck for $10 billion. And just the world moved so quickly here that -- we had been talking to lots of big pharma companies because they were interested in the class, and Roche, in particular, but several of them were very interested. And when push came shove, we kept that discussion open for as long as we possibly could to see about our other programs, to understand what we had in our FcRn franchise, to understand the opportunity there, to understand what strategic options might be there. But the truth is we ran out of time. Roche basically said it's now or never. And the value that they were offering and the transformative potential meaning for the capital associated with it, it just felt like a deal it was very hard to pass out for a program that was so recently in our hands. And so we took it. Again, we think Roche is going to do a phenomenal job with that program. And this amount of capital will really allow us to pursue our other programs, including the FcRn, fully in a way that, given the capital markets and given everything that's happening in I&I, we are not confident we would have been able to do otherwise.

Got it. I think some of the -- I mean, again, I'm acknowledging you created a lot of value for it. I think the conversation that we have with some investor is that, hey, look, we were at a point in Roivant where we could see it becoming a $20 billion, $30 billion company, and this would be an anchor asset. Now you gave that away. So what are the asset within Roivant that can make it a $20 billion, $30 billion company over the next, let's say, 5 years or so?

Yes. So our view, to be clear, is that doing this transaction dramatically increases the probability that we will be a $20 billion to $30 billion company or more. And I think the reason is we, outside of that program, had a phenomenal portfolio. We know that an anti-FcRn antibody can support a $30 billion company, there is one, and we think we have a program that is better. 1402 suppresses IgG more deeply than argenx' efgartigimod, for example. And so we know that alone is probably sufficient to be the basis of a tens of billions of dollar company. And on top of that, we have a dual inhibitor of JAK1 and TYK2, which, I think, if this week has taught us anything, it's, a, that TYK2 is an interesting mechanism, but may be even more interesting when paired with JAK1 as an important anti-inflammatory target. And we know that drugs in that sort of category generally can deliver important anti-inflammatory effect and that brepocitinib is a very potent molecule. I think like even if the only programs in our entire portfolio were those 2, there was no question that, that could be the basis of one of the largest I&I biotech companies out there. And on top of that, we have a portfolio beyond that, that includes earlier-stage programs. And one of the features of Roivant is that we are constantly hunting for new programs. Again, the TL1A program was not even a part of our portfolio a year ago, and this is one of the best environments for that activity that we've ever been in.

Yes. Talk about the timing of the TL1A deal. You could have -- because I think there were -- there was -- in the media, a while back earlier in the year, you could have done it. Now what were, like, driving factor into it? Was Immunovant profile anyways to get into the picture when you were deciding the time?

I think we really -- there was no benefit to transacting earlier at some level. Like the more information we could have in our hands at the time that we made a final decision, the better. And so we kept it open for as long as possible. Probably the single most important thing to me was knowing what we had at Immunovant before making a final decision on the TL1A, right? Like if the Immunovant program were not best in class, in some ways, we would have had a different decision to make because the TL1A program was sort of unique in our portfolio. And also, the capital needs would have been more tractable, right? We would only have been funding the TL1A program alone and not that, plus the FcRN program. But as it currently stands, knowing that Immunovant was best in class, knowing, frankly, that we had an asset that was going to carry us to being a $20 billion or $30 billion company, if things break our way, it made the decision to sell the TL1A easier. It gave us the capital we need to pursue with the FcRn program to full opportunity. And as in between the 2 programs, there are puts and takes. We've talked about some of them before. But one of the things I like about FcRn is the work ahead, while difficult for sure, is clinical development work at its core. It's about choosing the right indications, running the right studies, running them in parallel, picking the right end points. And bluntly, that is what Roivant has been good at for our entire history, is running clinical trials. We have a great team at Immunovant. They're set up to do a good job at this. Roivant has given a lot of thought to the indication strategy there in partnership with Pete and the Immunovant team. And I think we are set up to win. With the capital that we now have access to and the quality of the programs there, I think we're set up to really deliver a great outcome at FcRn.

Yes. Okay. So for Immunovant perspective, like how much more de-risking needs to happen to make sure that you have the best in class? Like how important are the 600-milligram dose data that's going to come -- become available in the next few weeks or so?

Yes. I mean I think I'll share with the Roivant view. Obviously, you talked to Pete and the Immunovant team as well.

They didn't come here because they're in a quiet period.

They're in a quiet period. I think, like, from a Roivant perspective, what are we hoping to see? I think the 600-milligram MAD dose is largely de-risked at this point. We saw quite a lot of data in the SAD and the MAD data back in September that supported the sort of profile of the class at least insofar as IgG suppression and albumin and LDL impact are concerned. I think we're looking to see data that is consistent with that. Bluntly, I think the biggest risk to the 600-milligram MAD dose at this point isn't that the profile of the product is impaired in some way. It's that we already know that LDL is a noisy parameter. And I feel like -- I hope that the data is clean enough in the balance of factors such that people can get comfortable with the profile in aggregate. I think there's obviously some chance that LDL is up a little bit or up, whatever, 5% or 10% as a function of noise. And hopefully, if albumin is clean and if IgG suppression is where we needed to be, then that kind of noise is understood to be what it is, but that's probably, to me, the biggest risk in a 600-milligram MAD data. So that, to me, is the most important sort of establishing parameter. Then the other question is, okay, so once we know that we have the best in class or at least deepest IgG suppression and once we know we have a clean profile from an analytic perspective, no LDL impact, no albumin impact, then where are we? So at that point, I think most of the FcRn companies, including argenx and others, have acknowledged at various times the deeper IgG supression generally yields better clinical benefit. But I think it's incumbent on us to show that as much as we can, and we have a lot of opportunities to do that over the next 12 months. We have the Graves' data, which is obviously run principally at 680 and bato. We have the induction period of CIDP data. We have the MG data coming in the second half of next year where we will be able to at least cross-trial, compare our data to some of the other anti-FcRn antibodies in MG. So I think that's sort of important to support the profile of the program.

Got it. And then going back just on Immunovant, I think there was a 1,200-milligram dose in the SAD. I think that should have given you enough clarity, right, because that's a much higher dose. And if you don't see anything there, you should probably not see anything at the 600-milligram dose.

Agree. All I'll say is we're not particularly worried about the 600-milligram dose.

Very good. Okay. So you mentioned that obviously, FcRn could be a big class, but you only own half of it, right? So how should we think about the rest of it?

Yes, we own about 60% -- 55%, 60%, depending on how you count it. So how should we think about the rest of it? Right now, it's a nice setup for us, to be honest. It's a great program. I think it's got public shareholders who are really excited about it. And for the moment, we get to share the cost with that group. So when Immunovant needs capital, the public shareholders get to set the price, and we put in 55-or-so percent of it. And that's important because it's a potentially relatively spending program given the breadth of indications. So I think that's a good spot for it to be right now. We -- again, we love the program. I think it makes all the sense in the world for the moment for the status quo to continue for it to be an independent public company, but that could change depending on data, et cetera. I think one of the things is -- my expectation is that we will want to see just what strategic alternatives wind up being available for Immunovant as a part of this process. But we're approaching that from a position of absolute strength right now because we know that no matter what happens, we have the capital to run the program through the completion successfully. Frankly, I suspect through the profitability. And so we can make a decision only to entertain options that get full value for the program.

I see. So I mean -- so there are 2 thoughts, right? You can bring them inside Roivant by acquiring it, but right now, independent because you have the funding.

Yes.

And as it gets more de-risked, the more value it creates.

That's absolutely right.

Okay. So that's on Immunovant. The cash that you're going to get from the Roivant transaction, are there any tax implication for that?

We don't think so. We think we will qualify for a tax exemption.

Okay. So you would. Now let's spend some time on brepo. So you will have the data in SLE. Talk to us how the studies are -- or study is set up. What would you like to show? I think there is a competitive dynamic with RINVOQ and TYK2 also in that context?

Yes, perfect. So look, what do we know? We know that JAK inhibition works in SLE. We know that from baricitinib studies way back when. We know it more recently from the RINVOQ data that came out early this year and AbbVie's decision to continue to pursue RINVOQ in SLE. We know that TYK2 inhibition matters in SLE. We know that principally from the deucravacitinib data, which was fairly compelling. And we know that for the most part, in indications where cytokine signal by either TYK2, JAK1 or both are relevant that brepocitinib, which has quite a lot of data at this point in multiple indications, delivers very compelling, generally best in category, best-in-class clinical benefit in those areas. And we know that SLE is one of those indications. We know that it's mediated by a combination of interferons and other cytokines that are signaled by each of TYK2, JAK1 and both. So that's all sort of disease biology. We have -- I'd say RINVOQ data, I'd characterize, from an SRI-4 placebo-adjusted delta perspective is like low teens basically, which, I think, is like good data. It's in the same sort of general vicinity as the 2 approved products, Benlysta and anifrolumab. And I think that's a reasonable place to be. We're not AbbVie. So that data is probably not quite good enough for us. And brepocitinib is not approved in all the other indications that RINVOQ is. Then we have the competitive bar set by deucravacitinib. So our view, if you look at the deucravacitinib data, the 3-milligram dose had, like, a 23% SRI-4 placebo-adjusted delta, but the other 2 doses were kind of mid-teens or lower. Our view, and I know there's some discussion about this fact, is that there is a little bit of an imbalance in severity across those 3 arms and that the higher response rate in the 3-mg arm is explainable, at least in significant part, by the dose imbalance. So our view is that the competitive bar set by deucra, all in, is probably a mid-teens SRI-4 placebo-adjusted delta. In our view, because we're not BMS either, is that in order for brepocitinib to be competitively successful, that we'd like to see better than that. We'd like to see something that beats a bar of mid-teens SRI-4. So ideally, like high teens or better would be ideal. And that's kind of the basis on which we're going to make the decision on whether to run another -- so this is set up to be one of 2 pivotal studies. If successful, we would need to run one more. And we'll make that decision based on the balance of the data, but really focused on are we going to get to better than mid-teens SRI-4 delta?

Okay. What are the time lines? So let's say the -- so these data are going to come in Q4, right?

That's right. Yes.

Relatively soon. What are the time lines for it to be on to the market, right? Like how long it's going to take another SLE study to run?

So let's look at the data first, and then we can share some guidelines on the design of that next study. I think some of that depends a little bit on how good the data is and what exactly we decide from a study design perspective, but I think you can figure -- these are likely to be, but not certain to be 52-week studies. And so you can figure it's probably a few years to run the study once we get started, and we've moved to get it started quickly if the data supported it.

Okay. What about other indications? Like dermatomyositis is another indication. It's small, but I think it's a pretty niche indication.

Yes. In fact, we don't think it's particularly small. We think DM could easily be a potentially blockbuster indication. When we acquired brepocitinib from Pfizer, I'll say like SLE was not the heart of the thesis actually. The heart of the thesis was orphan rheumatology. It was dermatomyositis. It was noninfectious uveitis. It was maybe HS, which is obviously an area where there's been enormous amount of development lately. And SLE was kind of a bonus. The biology looked good. It was a slightly bigger indication. Pfizer was already running when at the time was really a Phase II study with an interim. And basically, we worked with Pfizer. We redesigned that study mid-flight to get rid of the interim and turn it into one of 2 potentially pivotals and gave ourselves basically the option at an SLE indication because we thought the biology made sense. But actually, DM and HS and NIU are really the heart of the investment thesis. And as far as, like, DM is concerned, for example, DM is one of these great remaining orphan indications. It's got a significant number of patients, maybe high tens of thousands of patients. It's a patient population with a high unmet need. Basically, the only approved therapy other than glucocorticoids is IVIG, which is a pretty challenging treatment regimen. There are a couple of other things in development. Pfizer has an anti-interferon antibody in development. And then there's also -- the FcRns are in development. argenx is developing efgartigimod in myositis. But our view is that it's the kind of indication where an oral therapy that sort of looks and rhymes with a very potent JAK inhibitor, particularly with TYK2 activity as well, could be a very compelling treatment option for patients, and these are patients without a lot of options.

Okay. When are we going to see the data? That's 2025?

It's 2025. That's right. So that study is enrolling now. It's enrolling well. It's again a 52-week study, and it reads out in 2025.

Okay. Very good. Then we move on to VTAMA, right? I think the launch has been pretty successful. Just curious like how much now weight you put on that asset given that it's commercial, you're going to have a label update on AD the end of next year. So how are the dynamics there and the launch process?

Yes. I mean, look, so I'm a true believer in VTAMA. I think it's a great product. The patient -- I use the product. I think the launch certainly has been successful relative to any reasonable expectation for it. I think if we're to be honest, 2 things. One is I think it has been deeply overshadowed by the rest of our pipeline for a lot of reasons. Most of them correct, right? Like I think both the risks associated with VTAMA's path through a large outcome and also like the possibility of that outcome versus what a TL1A or an FcRn or even brepocitinib could be, it's just a different kind of opportunity. And so I think that's just like a truth. It's been living in the shadows of these other programs. I think that's fine. I also think it's -- like to be blunt, it's clear that the pace of the launch in psoriasis has been good, but you can always hope for better and -- especially over the past, call it, 6 to 8 months, I think it's been like slow and steady growth, but not like blowing out of the water growth. There are reasons for that, that we're coming to understand. It just takes time to convince patients who've been writing -- sorry, patients -- physicians who have been writing steroids for so long to switch to writing a novel topical, and we're working on solutions to that, but I think it's just taking time. So I think like that's another reason why I suspect it's been overshadowed. I think our view at this point is that the "downside case" for VTAMA most likely is that it's on a path to being a profitable product. How exciting a profitable product? Depends on a number of factors, including how well we can convince people to switch behavior and on how much we're willing to resource? And how much capital we're willing to spend on continued investment in DTC and in other forms of marketing? And then also, look, we have extraordinary data in AD that we generated earlier this year, and that's a significantly larger indication than psoriasis. And I think the opportunity also depends a little bit on how that launch in AD look. So we're actively evaluating exactly how to think about the future for VTAMA and how to turn it into that profitable product that it will become, but I think we're sort of considering all of those factors. And one of the things that I'm determined is we have all this capital coming in. Just because we have the capital, I don't necessarily want to, like, spend it by default. So we want to be very rigorous in our capital allocation decisions, including to VTAMA, which, again, I fundamentally believe in as a potentially great product here.

What are you hearing from physician on the competition on ZORYVE, right? I think the script for them also seem to be trending in the right direction.

Yes. Look, ZORYVE's script volume has been solid. We've always thought ZORYVE was a pretty good product. To be honest, like I don't really view ZORYVE as our principal competition. Our principal competition is really trying to take script share from steroids. And both we and ZORYVE are tiny, tiny, tiny portions of the total market now relative to topical corticosteroids. So I think we sometimes hear about them in a sort of "competitive dynamic." But mostly, I think the idea that patients and physicians now have access to novel topicals is great. And the same thing in AD with OPZELURA, to be honest. It's also a good product that I think physicians are happy to have access to. Feedback from physicians, in general, is very positive. I think physicians like having a novel topical option. They like something that's nonsteroidal. They like something that can be used anywhere on the body. They like not having duration of use limits. I think patients are generally very happy with the product. We have at least one -- there's one investor in the company who we meet with us at conferences like this, a public investor who's on the product and who occasionally just like mentioned to us they think the product works too well. That's like -- they don't need refills as often as they think it would be optimal for us. Like I think, like, in general, feedback on the product is quite positive. I think empathy is a real issue. Docs are used to steroids and write them a lot. And steroids are pretty good products. So the fact that we're better is something we need to continue to take advantage of and develop. But in general, I'd say feedback has been good.

I mean how should we think about the AD? It comes on to the label next year.

Yes. So AD is a huge market, first of all. It's about -- so this is an unfathomable number. Every week in atopic dermatitis, there are over 300,000 topical prescriptions written. So for context, like in psoriasis right now, we're doing 4,000 to 5,000 scripts in a normal week. There are 300,000 scripts written a week in atopic dermatitis. It is a very large market. That's the first thing. The second thing is it's a pretty heavily pediatric market. Somewhere between a quarter and a half of the scripts are pediatric, depending on how you count it. And several of our competitive products either have meaningfully worse efficacy in pediatric settings or in the case of OPZELURA, for example, historically haven't been approved in pediatric settings. They've read out a pediatric study. I think they're looking to get younger patients added to the label. But remember that OPZELURA is a JAK inhibitor, has a black box warning, which probably not particularly important medically, but my guess is it just may matter to parents as they think about application of the product to a young child. So our Phase III study went all the way down to age 2. We studied the pediatric population. We delivered great data that's, I'd say, comparable numerically to OPZELURA and general comparable numerically to some of the systemic therapies, maybe a hair lower than OPZELURA, but not significant, and significantly better on a cross-trial comparison basis than, I'd say, most of our competitor topicals. So I think the opportunity there is very large, and I think some of the same dynamics will apply. Steroids have been in wide use, and docs are very comfortable with them. But I think there's even more desire for nonsteroidal topicals. We saw it with all of the promise around EUCRISA a few years ago, although that product had liabilities that price [indiscernible] launched to some degree, but I think there's just a huge opportunity in atopic dermatitis as well.

How about -- how should we think about the inflection, right? Is it that's the one that's going to drive the inflection or is it [indiscernible]?

That is certainly the biggest, most visible sort of opportunity for an inflection in the product, is the AD launch and the associated uptake. It's certainly possible that we could see an inflection before that if we start to make a difference in some of these messaging to docs around the product. But I think, like, the biggest, most obvious inflection from here is the AD launch.

Got it. Okay. That's good. Now in terms of other pipeline asset, like what are you highlighting to investors? Obviously, Immunovant we discussed, brepo we discussed.

So to be totally honest, my suspicion is that most investors are still focused largely on the things that we've just talked about, but there is -- there's a bunch of other exciting stuff in our portfolio. We have -- this isn't the pipeline asset, but we have an -- so we have a great IP estate around lipid nanoparticles for the use in mRNA delivery. And we have some litigation ongoing to enforce that IP. We believe that both the Pfizer and the Moderna COVID-19 vaccines infringe on our IP. So those are sources of value that I think many investors are paying some attention to and some investors are paying a lot of attention to. So that's one thing. There's some important milestones in that litigation this coming year, including a claim construction hearing in February that'll lay out the court's interpretation of what the language and the patent claims mean. So that's a relevant milestone. We've got 2 other datasets outside of FcRn and brepo and VTAMA coming. One is some relatively small amount of open-label data from a small molecule, SF3B1 modulator, in low-risk myelodysplastic syndrome. This is for patients who are dependent on transfusions to treat anemia in those patients. That's a small amount of data in a relatively risky program, but like it'll be interesting to see that data coming potentially as soon as the end of this year. And then next year, we have data coming from a study of -- I think I mentioned this before, an anti-GM-CSF antibody called namilumab in sarcoidosis, which again -- so anti-GM-CSF antibodies have had a little bit of a tough development history. It's sort of a target in search of an indication. We think there's a pretty good chance that sarcoid is that indication. Sarcoid is another one of these sort of great orphan markets where there's a high unmet need, not a lot of other therapies that have been successfully developed and where the field has learned a fair amount about clinical development. And a lot of the disease progression is a function of the formation of these granulomas. And GM-CSF is known to be involved in that sort of disease pathophysiology. So we are hopeful that we have some opportunity there. That study reads out next year as well. Neither of those opportunities, the SF3B1 or the GM-CSF, I think are, like, on investors' radars at all, and that's fine. Those are relatively high SKU, relatively risky opportunities. But if they work, they will quickly elevate to relevant programs in our pipeline.

Maybe a final question. So the cash that you're going to get, I mean any consideration of returning it back? Time frame within reach?

Yes. Look, I think it's a great question. It's one that we get all the time. We've said over and over again, and I think this is, like, just an important point, we intend to be patient and thoughtful in making decisions around that capital. It is a unique asset to have that kind of cash to deploy, especially in this market. So that's a -- it's a rare opportunity. And I don't think there's a benefit to rushing. I'm sure the market would, like, be short-term happier if we made a decision quickly, but I don't think there's, like, a much of a benefit to rushing to make a decision there. I think funding the existing pipeline is incredibly important to us. I think thinking critically about what the BD opportunities might be, and we see many, some of which are absolutely as exciting to us as the TL1A we just sold. Seeing what those opportunities look like is important in seeing them through. If there is excess capital beyond that, I absolutely think using some of it to return to shareholders, to clean up our cap table, maybe we've got some relatively concentrated holdings -- holders, I think, are all sort of on the table, but we're not going to make that decision today. And frankly, the deal hasn't even closed yet, so we don't have the cash yet.

So in terms of the size of the deals, right, I mean, in the past, you've stayed in these very creative transactions. Like now that you have more cash, like does that increase -- I mean how are you thinking? Like could you go bigger on the types of deals that you could do?

Yes. Look, it's not lost on us that we have more flexibility. But that said, we are fundamentally value hunters. And even in the relatively strong capital position we've been in over time, we have not found a lot of, like, public M&A or very large up-front transactions meet our own threshold. So it's possible that could change. The capital markets are weak right now, and that creates opportunity. So that's something we're watching closely. But my suspicion is that a lot of the deals we do, if not all of them, will look pretty similar to the kind that we've done historically. Now notably, even though the up-front cost of many of the deals we've done has been low, we have not shied away from classes like TL1A and FcRn where the actual cost of developing the program is high. And certainly, the capital position we're in allows us to think expansively around what that could look like and puts us in a position to be -- like, look, if a big pharma company is looking to out-license a program, and it's a great program, but the development cost is 9 figures, bluntly, there are not that many other people that can go after this program. So we are uniquely well positioned. And some of those programs are really exciting. That's why we were there at the front of the line for the TL1A.

Very good. Very good, Matt. I think that's all I had for you.

Great. Well, thank you very much for having me again.

Thank you so much.

This was a great day and great day to catch up.

Thank you. Thank you.