Vertex Pharmaceuticals Incorporated (VRTX) Earnings Call Transcript & Summary

Good day, and thank you for standing by. Welcome to the Vertex Pharmaceuticals VX-864 Phase II Results Conference Call. [Operator Instructions] I would now like to hand the conference over to your speaker today, Michael Partridge. Please go ahead. .

Thank you. Good afternoon. This is Michael Partridge of Vertex. On today's call, we will discuss results from the Phase II clinical study with our investigational Z-AAT corrector, VX-864, for the treatment of alpha-1 antitrypsin deficiency or AATD. Dr. Reshma Kewalramani, Vertex' CEO and President, will provide prepared remarks on the call today. Dr. David Altshuler, Vertex' Head of Global Research and Chief Scientific Officer, will join for Q&A. We expect today's call to run for approximately 30 minutes, and we intend to conclude the call at 5 p.m. We recommend that you access the webcast slides as you listen to this call. This call is being recorded. A replay will be available on the Vertex website. We will make forward-looking statements on this call that are subject to the risks and uncertainties discussed in detail in today's press release and in our filings with the Securities and Exchange Commission. These statements, including, without limitation, those regarding Vertex' AAT program and Vertex' pipeline are based on management's current assumptions. Actual outcomes and events could differ materially. I will now turn the call over to Dr. Reshma Kewalramani.

Thanks, Michael. I expect by now you've seen the press release we issued today in which we announced the achievement of the primary end point in the Phase II study of VX-864. To recap, we saw rapid and statistically significant increases in functional and antigenic AAT levels across all VX-864 dose groups, with clear separation from placebo. With these results, we have proof of mechanism for a small molecule corrector in people with AATD for the first time ever. VX-864 was generally well tolerated, and we saw no evidence of on-mechanism toxicity. The magnitude of AAT increase did not reach levels expected to drive therapeutic benefit. And as a result, we will not advance VX-864 into late-stage development. However, based on these findings, we are one step closer to meeting our goal of developing small molecule correctors of Z-AAT protein that could transform the lives of people with AATD. We've gained important insights from the 864 Phase II trial, and we will apply these insights to our ongoing AAT research effort to advance new, more clinically efficacious molecules with the potential of achieving therapeutic benefit and to accelerate future clinical studies. We expect the first of the next set of small molecule correctors to enter the clinic in 2022. I'll review the results of the study in more detail next and close with key learnings that we believe position us to successfully develop future medicines targeting AATD. First, the study design. This was a randomized placebo-controlled trial that assessed 28 days of treatment with 1 of 3 doses of VX-864 or placebo, followed by a 28-day safety follow-up period. The primary end points were changed in plasma, functional AAT levels from baseline at day 28, and safety and tolerability. The secondary end points were changed in plasma antigenic AAT levels and pharmacokinetics. Overall, we enrolled 44 subjects across the 3 dose groups and placebo. We enrolled adults with PiZZ genotype and severe AAT deficiency across 30 clinical sites in the U.S., Europe and Canada. The demographics of the treatment groups were generally comparable, although the baseline functional and antigenic AAT levels were slightly higher in the placebo group. On to efficacy. We saw a highly statistically significant increase in both functional and antigenic AAT levels at day 28 compared to placebo across all 3 dose groups. Mean functional AAT levels increased from baseline between 2.1 and 2.3 micromolar, reaching an average of approximately 6.1 to 6.3 micromolar. And the mean antigenic AAT levels increased between 2.6 and 3.4 micromolar reaching approximately 7.5 to 7.9 micromolar. The fact that functional and antigenic AAT levels are similar indicates that the properly folded Z-AAT being secreted into the blood is functional and furthers our conviction in the mechanism of action of our small molecule correctors. Efficacy data plotted over the duration of the study provide further insight and confidence in the robustness of using a corrector approach. We see a rapid increase in functional AAT by day 7, which was sustained through the 28 days of treatment. This is clearly distinguished from the placebo arm, which is absolutely flat over the 28-day course. We also observed a slow and steady decrease in AAT levels when treatment was completed at day 28. This is consistent with the expected half-life of native AAT protein and further confirms the biological activity of VX-864. The ability of VX-864 to drive increases in AAT is clear and unambiguous. Every patient in all VX-864 dose groups showed an increase in functional and antigenic levels at day 28. Patients receiving mixed 864 went from an average of around 4 micromolar of functional AAT at baseline to an average of more than 6 micromolar at day 28, representing an approximate 1.5-fold increase. These levels represent clear evidence of biological activity but not yet to levels of functional AAT that are associated with therapeutic benefit in people with this disease. Moving on to safety. VX-864 was well tolerated across all groups. There were no discontinuations due to adverse events and no serious adverse events considered related to that drug. The majority of AEs were mild or moderate in severity. The most common AEs in patients treated with VX-864 were diarrhea and nausea and were not treatment limiting. Liver function tests were similar between placebo and VX-864-treated groups, and overall, there was no evidence of any impact on LFTs with VX-864. We're pleased by the overall favorable safety profile, which supports our continued investment in small molecule correctors of Z-AAT protein. With that, let me summarize the key takeaways and the next steps for our program. First, we've established that the corrector mechanism of action translates to the clinic. The increase in functional and antigenic AAT were rapid and consistent and properly folded Z-AAT chose biological dynamics consistent with properly folded native AAT protein. Second, we see no evidence of on-mechanism toxicity, and VX-864 was generally well tolerated. Third, With these clinical data in hand, we can now work to optimize our small molecules and bring forward those molecules that can deliver a greater magnitude of benefit. To be clear, our focus for the next generation of molecules is to raise functional AAT levels into the therapeutic range. That is to say greater than or equal to 11 micromolar. We have discovered and are working with renewed urgency to advance additional molecules to the clinic in 2022. Fourth, We have gained clinical experience in the AATD trials that will enable us to move faster in the future. From here, we'll be able to design smaller and faster trials with the goal of accelerating the overall development time lines. As we did in CF, we're now in the position to advance multiple molecules, and we could do so simultaneously in order to then advance the best molecule or molecules into pivotal development. In sum, I'm encouraged by the efficacy and pleased with the favorable safety results from the VX-864 trial. With these results, we're one step closer to reaching our goal of developing small molecule correctors to treat the underlying cause of AATD, and we're committed to doing so. I look forward to sharing progress on our growing CF business and the advancements of our pipeline, including our AATD small molecule portfolio in future updates. Before we get to questions, let me extend our deep gratitude to patients with AATD who participated in this trial and to the trial investigators. Thank you for your time, effort and commitment to the study. Let's now open the line for questions.

[Operator Instructions] Our first question is from Liisa Bayko with Evercore.

Yes. Sorry about the outcome, but hopefully, you've learned a lot. It seems like that's the case. Can you maybe talk about, sort of going forward, what would be some of the minimums you'd be looking for in terms of kind of changes to be able to take the compound forward? And then maybe you can talk a little bit about what are some of the differences in the next molecules you have coming up. And are you still thinking -- I know you had one that was going to enter Phase I. Will that still enter Phase I based on what you're targeting?

Yes. Liisa, thanks very much for the questions. There's a few different questions in there. But if I try to elevate and address all of them, there's really a question in here about what are we looking for and why do we have confidence that the next wave of molecules that we bring forward are going to be able to hit our target. And you asked specifically about our targets. So let me address that, and I'm going to ask David to walk you through some details of what we need to do because the path forward is very clear. It's not easy, but it's very clear and tell you about the scaffolds, the characteristics of our next wave of molecules. So to be clear, what we're looking for in our next wave of molecules, the first of which will enter the clinic in 2022, is to hit the protective threshold of 11 micromolar. So here we are today at about 6, 6.2. With the next molecule, we're looking to get to greater than or equal to 11. David, can you walk us through the details for how we are thinking about what we learned and how we translate that from the lab to the clinic?

Absolutely. And thank you for the question. In translating any new program for the first time from the laboratory to the clinic, one of the key features is obviously to build a PK/PD model to understand dosing and response. This includes the usual allometric scaling of PK, so you can predict exposure from preclinical species to humans, but also based on laboratory data it's predicting the exposures needed to achieve clinical response. Now in the case of AATD, one of the key models we use is a mouse that carries the human Z-AAT gene. And we know that the mouse model involves a whole host of factors that different -- differ between human and the model. For example, these include the mouse model has multiple copies of the human Z-AAT transgene, and that has different levels and regulation in patients. The half-life of human Z-AAT when expressed in the mouse is different than in patients. The metabolism differs. Transporters differ among other things. So using the best data from the mouse in each of these factors, we built a model of dose response. And based on this model, we picked doses to ask 3 questions in the Phase II trials. First, what are the exposures achieved in the blood? Second, is there a biological effect? And if so, what is the magnitude of that effect? So what did we see in the Phase II study? Well, with regard to exposures in the blood, the levels achieved were exactly as our model predicted and scaled with dose. With regard to the biological effect, the study was unequivocally positive, with a highly statistically significant response at each dose, every patient responded. There was a clear on effect and an off effect after stopping dosing. The AAT-produced is functional and had the expected half-life. With regard to the magnitude of effect, it was lower than predicted, and it was insufficient to reach the desired transformative effect in patients. So why is that? What we've learned is that some of the parameters in the model were off. The clinical data focuses us on the relationship between the exposure of the drug in the blood and the exposure at sites of action. And with the human data that we now have in hand, we can retrain our predictive models based on that and allow us to create improved predictive models to select and optimize from our portfolio of scaffolds and molecules already in our portfolio to move into clinical development.

So Liisa, just to make a long story short, there are a lot of predictions we make when we go for the first time from the preclinical space into the clinic. David's outlined what some of those are. Now that we have this human clinical data, we can optimize the model, but more importantly, optimize our molecules to get us to where we need to be. And that's really what we're focusing on, different scaffolds, the characteristics that overcome these parameters that we need to optimize so that we could get to this level of 11 micromolar or higher.

Okay. That was a really robust answer. So it seems like you've kind of identified what you need to chisel away at next.

Yes, very much so, Liisa. I think that, that is the important point. I don't want to breeze past the most important point. This is the first time a small molecule corrector for this disease has been put into a Phase II trial of the patients. And this is the first time that we've shown proof of biological activity, not we, Vertex, but the field. That's an important part, and we know exactly what we need to do now to optimize that. Again, it's not easy, but we know what to do to get the next set of molecules to get us to the levels we want.

Our next question comes from Michael Yee with Jefferies.

Maybe for Reshma or David, mechanistically, what is your hypothesis as to what you think is going on there? Is it a potency corrector shop alone issue. Actually, I'm never clear if you've actually thought -- plan to shop around or when it's binding. Or is it more drug exposure that you need to get into the liver and the correlation from exposure to animal to humans is not correct? So I would think that, that's sort of the equation there in the liver. Maybe you could comment on what you think the issue is there.

Yes. Michael, I think you've hit on it and maybe I would summarize it in the following way. There are a number of parameters for which we make assumptions before we have clinical data. So clearly, our model predicted correctly that this mechanism can lead to elevations in functional AAT. But the magnitude of the treatment effect is off. If I summarize what David said, really, what we're looking at is getting exposures higher in the target tissue, and you're very correct, that means liver. There are these other parameters that we're working on that David mentioned, but you're very correct about the liver. And just to be clear, our mechanism is catalytic. That is how this molecule works.

Our next question comes from Salveen Richter with Goldman Sachs.

This is [ Andrea ] on for Salveen. Maybe, Reshma, if you could speculate here on the lack of observed dose response and what that might imply for the mechanism of action here that you were just speaking about?

Yes. It's a really important question that you're asking, and let me take a minute to walk you through this so that we all understand it that this is a really key question. When you see a flat dose response, which is what we're seeing here, there are a few things to think about. I'm going to go through all of those, but I'm going to speed forward to the end and give you our conclusion. And then I'll walk you through it. We believe we're at the low, shallow part of the PK/PD curve, okay? And I'll come back to this. But what you should think about when your flat dose response is from the very top, maybe there's no real response there to be seen. Maybe it's a spurious result, and that's why you don't have a dose response. That is clearly not the case here. We have a highly statistical result in terms of functional and antigenic AAT. It's clearly separated from placebo. There's a clear drug on effect and a clear drug off effect. This drug works. There is proof of biological activity. That's not the answer. The second thing we think about is whether we're exposure limited. That's to say could we get to the drug concentration levels in the blood. In this instance, we could. And as we increase the dose, we get to greater exposures. That's not the answer here. The third thing you could think about is, all right, well, maybe you're at a ceiling. Maybe this is the very best you can do. Maybe the very best you can do is move from an average of 4 micromolar to 6 micromolar. But I don't think that's the case here either. And the reason for that is the following. If you think that this is the most we can get, it has to be for 1 or 2 reasons, either there is a limitation to the small molecule correctors, and it could only correct so much protein or you think that patients with AATD simply don't make protein for our small molecule correctors to act on. Okay? So let's take the first one first. Across all of our assays, when we evaluate our small molecule correctors over a broad range of Z-AAT amount, whether it's in pure protein or in cells or in our mice. The small molecule correctors can promote the proper folding of nearly all the protein. So that's not it. And if you think about the AATD patients, they have a functional AAT level of, let's say, 4, which was what happened in this case. In [ you and I ], that number is north of 20 micromolar, so there's plenty of headroom there. And when you walk through all of those potential alternative explanations for the dose response, which you really get to is what we're seeing is that we're on the low, shallow part of the dose response curve. And what we need to do is get to molecules that are more potent, more efficacious at the right tissue deliver for us to be able to get to therapeutic levels of -- that we want to get to, greater than 11 micromolar. I hope that's helpful.

Our next question comes from Mohit Bansal with Citi.

Great. And maybe just driving a little bit more on the previous question here. So given the doses, you said that you had the shallow end, how much more incremental exposure was required to get to 11 micromolar? And the other question is do you think the [indiscernible] could be more [indiscernible]?

Yes. Mohit, these are really good -- really, really good questions. So what we know now is that we're at the low, shallow end of the curve, right? And what we now need to do to get to exactly your question is to take these clinical data, and you can imagine, we are doubling down on our work in the lab. We are translating the human data to make the appropriate adjustments in our models, but more importantly, in our molecules to get us to those levels. So it's a little bit about potency. It's a little bit about exposure and it's a lot about getting the right exposure in the right tissues. And one important thing to tell you, Mohit, is we're not going back to square one here. We already have molecules in the lab that come from different scaffolds that have different characteristics that allow us now to get to where we need to be. And when I say that, what I mean is we needed these clinical results so that we can get to molecules that we can predict get to these -- the level of 11 or greater. And that's exactly what we have now.

Got it. And the combo, do you think...

Yes. Sorry, yes, yes, the combo, really good question. So Mohit, I won't exclude that as a possibility. But from all of our preclinical data and very importantly, across these assays that look at pure protein or protein in cells or in our mouse model, with a single corrector, we can get to and we do get to levels that are above the protective threshold. So I'm not going to say no never, but I think that we're going to get there with a single corrector and the wave of molecules that we have in our preclinical development, I think, will get us there as a single corrector.

Our next question comes from Phil Nadeau with Cowen and Company.

Kind of a two-part question from us. One is how consistent was the response among patients? Was everyone tightly correlated? Or were there actually some patients who got closer to 11? And then second, at one point, you had been suggesting that you could move a clinic -- another molecule in the clinic in 2021. Why is it now 2022? Is it based on the learnings from this study? Or was there some other issue with the lead preclinical candidates that you had identified previously?

Yes. Phil, this is Reshma. Let me take your first question, and I'll ask David to tell you about the status of our molecules in the preclinical space. Phil, the results are remarkably consistent. Every single patient responded, and obviously, we look at individual patient responses. Every single patient in each of the VX-864 groups responded. I would say that there are -- I'd say the mean represents what we see across the patients. There are some patients who are a little bit higher. But I think the mean that we described does indeed reflect the patients when we look at them individually. David, do you want to just talk quickly about where we are with our molecules and the next wave?

Absolutely. No, thank you for the question, Phil. We have a portfolio of molecules from different scaffolds with excellent and improved potency and efficacy. And what we're doing is taking the data we now have from this clinical study. We're retraining our translational models, and we'll use those to select those that are in preclinical development already and also to optimize future molecules so that we can bring forward ones with the greatest chance of success in achieving greater than 11 as the protective threshold in the next wave. And I think the other point that I would make is just that, as Reshma mentioned, we really have learned from this study that we can accelerate clinical development. There's really 2 things we learned that I think are just worth noting. One is that because the effect is very rapid, the studies can be short. You'll see that at 7 days, we achieved the maximal effects seen. So the studies are short. Second of all, there's very limited variability, limited variability both intra and inter individual. So small studies can give very tight error bars. If you look at the data, you can see the studies can be small. That means that we can take an approach similar to what we took in CF. We'll be able to bring forward multiple molecules once we retrain our translational models and be able to rapidly assess that. I think we'll be able to move more quickly, certainly than we did in the past than you might expect.

Phil, with regard to the specific question around 2021, we want to take what we've learned in this clinical trial. We need to translate it into our model so that when we do this with the next wave of molecules, we actually have a medicine that can get us to greater than or equal to 11. Hence, my suggestion is that we're going to have the next one in '22.

Our next question comes from Brian Abrahams with RBC Capital Markets.

This is David sitting on for Brian. Definitely sorry to hear about the outcome as well. I guess I realize the question might be a little early, but -- and I realize you didn't do biopsies in this study, obviously. But are there any ways to sort of gauge what's going on in the liver maybe? Maybe some sort of biomarkers to gauge what's happening with polymers in the liver given that the functional and antigenic levels of AAT sort of just tell us a little bit about what's going on in the outside?

Yes, yes. Let me start with that one. Scientifically and in terms of innovating in this disease area, the results today are actually really important. It's a really important day because it's the first evidence of biological activity for small molecule corrector in this disease. And I'll just remind you that the only other disease and the only other class of molecules that work by properly folding proteins is the CFTR modulators for CF, obviously, something that we discovered. So I want to make sure that we recognize the importance of what we've done here today, de-risked this mechanism in terms of efficacy, de-risked this mechanism in terms of safety. With regard to your question about the liver, it's an excellent question. I'm not going to go into terrible detail here because the insights that we've generated from this study are very important, and it can be enabling for competitors. What I will tell you is the exposure in the liver, the target tissue is very important, and the results that we got from these Phase II studies tell us what we need to do with the next set of molecules and the properties to look for. So I feel really good that we understand that and that we're going to translate that with the next set of molecules. Obviously, you can't know that with the first molecule that you bring, right? You make the assumptions, and when you go to humans, sometimes you know that you were exactly right. And sometimes you know you get the data and you know that you have to optimize those models. That's where we are today.

We have a question from Matthew Harrison with Morgan Stanley.

Great. I guess what would be helpful to learn for us is can you just talk about what's different that you've learned here versus the data you already had from 814 that's allowing you to optimize your PK/PD models and have confidence that you now know what the rate parameters are.

Yes. Yes. Great question, Matthew. So let me just take the biggest part of it first. You'll remember 814 was a prematurely terminated study, so we never got to the finish. And the reason we terminated it early is because we had liver toxicity, right? So one of the key, key questions, we always believe, and you'll remember we talked about this, whether that liver toxicity was specific to VX-814, or it was on mechanism. Clearly, if it was is on mechanism, that's very, very concerning because that doesn't have a path forward. What I can tell you with the results today is our belief that it was molecule specific has been confirmed. There is no on-mechanism toxicity. That's really important. That is a big takeaway. Second, with VX-814, we didn't get to the finish line, so we weren't able to assess whether we could move functional AAT levels. We couldn't assess the time course, and we couldn't assess whether the AAT we produced actually had the characteristics of native AAT. That's everything that we've been able to do here. And maybe most importantly, for the future, we have the parameters now from the human trial so that we can refine our models to get to the right exposures in the liver to optimize some of the other parameters that David talked about so that the molecules that come from the lab can get us to those levels of greater than or equal to 11. I'll just draw a quick analogy for you because all of us are pretty familiar with our CF experience. The reason we have such high confidence with our CF molecules when they come from the lab into the clinic is because of the HPE assay that has been trained over the years with human clinical data. That's the place we are at now with AATD. The last thing I'll tell you about the CF experience is, if some of you go back into maybe 2015 or so and think about the ORKAMBI experience, a lot of people thought that with ORKAMBI for the F508 delta homozygote, we would never be able to get above 2.5 for ppFEV1, right? But look at where we are now for the Delta F508 homozygote with the triple combination, we have far more than a 2.5 for ppFEV1. We're all the way up to 14 or so. I feel like today is akin to where we were with or ORKAMBI for the Delta 508 homozygote. And now we can go optimize and the next wave of molecules in this disease can get the patients to the kind of therapeutic benefits that the triple combination have gotten for patients in CF. I know that we're out of time. Thank you all for joining. In closing, let me just step back and just go to 50,000 feet for a moment. Where we are with our AATD program with VX-864, we have clear proof of biological activity. We have no on-mechanism toxicity and that means we've de-risked this mechanism both from a safety and an efficacy perspective. We have what we need in terms of the learnings to retrain our models and maybe more importantly, get more optimized molecules into the clinic. And I expect the first of those to be in the clinic in '22. I do think we're going to be able to go much faster, smaller clinical trials, faster clinical trials to get us to pivotal development based on what we've learned in this trial. We're going to redouble our efforts in the lab for AATD, while we also work on the 6 other disease areas that are in the clinic and grow our CF business. We obviously have a lot going on. With that, I'll say thank you, and again, I'll turn it over to Michael.

Thanks, everybody, for joining us on short notice, especially. I know there were a couple of other folks in the queue to ask a question. The IR team and Reshma and David will be in the office tonight, and we will look forward to following up with you. Thanks again.

Thank you. And this concludes today's conference call. Thank you for participating, and you may now disconnect.