Atea Pharmaceuticals, Inc. (AVIR) Earnings Call Transcript & Summary

Good morning, ladies and gentlemen, and welcome to the Atea Pharmaceuticals A2 -- AT-527 Clinical Update Conference Call. [Operator Instructions] I would now like to turn the call over to Jonae Barnes, Senior Vice President of Investor Relations and Corporate Communications at Atea Pharmaceuticals. Please proceed.

Thank you, operator. Good morning, everyone, and welcome to Atea Pharmaceuticals AT-527 Clinical Update Conference Call. This morning, we issued a press release which outlines the topics we plan to discuss. You can access the press release as well as the slides that we will be reviewing today by going to the Investors section of our website at iratteapharma.com. With me today from Atea are Dr. Jean-Pierre Sommadossi, Founder and CEO; Dr. Janet Hammond, Chief Development Officer; John Vavricka, Chief Commercial Officer; and Andrea Corcoran, CFO. They will all be available for the Q&A portion of today's call. Before we begin the call, and outlined on Slide 2, I would like to remind you that today's discussion will contain forward-looking statements that involve risks and uncertainties. These risks and uncertainties are outlined in today's press release and in the company's recent filings with the Securities and Exchange Commission, which we encourage you to read. Actual results may differ materially from what is disclosed on today's call. With that, I'll now turn the call over to Jean-Pierre.

Thank you, Jonae. Good morning, everyone, and thank you for joining us today. I will begin on Slide 3. Before diving into the MOONSONG data and our MORNINGSKY update, I would like to begin with a brief view of AT-527. AT-527, our orally administered product candidate for COVID-19, is a direct-acting antiviral agent derived from our purine nucleotide prodrug platform. It is designed to treat COVID-19 and prevent disease progression, viral transmission, and hopefully, the development of lung COVID complications. Given the evolving dynamics of COVID, which is becoming endemic, medical intervention, and oral therapeutics in particular, will be essential in stemming the tide of this pandemic and future surges. This drug targets viral RNA polymerase, a highly conserved enzyme critical to viral replication. Uniquely, AT-527 has a mechanism with dual targets, inhibiting both the NiRAN and the RdRp, which are the 2 functional domain of the RNA polymerase, and has the potential to create a high barrier to resistance and antiviral activity across variants. The Phase II MOONSONG viral kinetics study that we reported this morning did not meet the primary endpoint in the overall population. Based on what we know now, this actually is not surprising, as oral direct-acting antivirals, such as have shown similar outcomes in the mild or moderate outpatient population. What is reassuring is that the viral load reduction of approximately 0.5 log at day 7 as compared with placebo was observed in the high-risk patient with underlying health conditions. This is consistent with our data in the Phase II hospitalized patients with underlying conditions as well that we reported at the end of Q2. With that overview, I will now turn the call over to Janet for a detailed review of the MOONSONG Phase II interim results and an update on the Phase III MORNINGSKY trial. Janet?

Thank you, Jean-Pierre. I'll begin with Slide 5. It has become abundantly clear that COVID variants are creating new challenges for treatment and prevention. These variants can also make the evaluation of clinical trial results more challenging because of differences in the way the respective variants act and how patients respond to such variant. Ongoing SARS COVID-2 genomic surveillance worldwide has improved our ability to rapidly identify such variants. As you can see on this chart, almost 6,000 variants have been sequenced, which is an astounding number. This chart was recently published and it's already out of date. The evolution of the virus will make COVID-19 endemic, with the ability to continue circulating for years to come. SARS-CoV-2 variants interest have been associated with increased transmissibility, neutralization resistance and disease severity. Since the MOONSONG trial was designed, we all have a clearer understanding of patient risk factors for susceptibility to disease and an enhanced understanding of the immune response in different clinical setting. Furthermore, as the virus evolves with different variant strain, the viral kinetics have been shown to change as well as the disease profile and its clinical manifestation. This has presented unforeseen challenges to the MOONSONG study, which could not have been anticipated at the time of design and have impacted unfortunately what we've been able to demonstrate. Turning to Slide 6. As you can see here, our program includes 6 studies of AT-527 running in parallel. We're leveraging lessons learned from each study to better serve our patients and updating our program in real time. Some of the updates to the program include: firstly, reviewing potential modifications for MORNINGSKY based on the MOONSONG results and other recent results. We now anticipate the MORNINGSKY results in the second half of 2022. Amending the Phase II hospitalized study with the 1,100 milligram BID dose, we anticipate initiating the prophylaxis study during the first half of 2022. Turning to Slide 7. As Jean-Pierre mentioned, earlier today, we issued a press release announcing top line results from our MOONSONG study. MOONSONG is a randomized double-blind, multicenter placebo-controlled Phase II trial, evaluating AT-527 in mild or moderate COVID-19 patients in the outpatient setting. It is being conducted in collaboration with Roche to evaluate the antiviral activity, safety and pharmacokinetics of AT-527 in adult patients. The primary endpoint of this trial is the change from baseline in the amount of SARS-CoV-2 viral RNA as measured by RT-PCR at specified time points. A full analysis of the study is still ongoing. Moving on to Slide 8. Prior to reviewing the data, I'd like to take a moment to walk through key baseline characteristics and trial design factors that may have impacted the top line results reported today. The MOONSONG study was designed in late 2020 before the vast array of variants, including Delta, began to emerge, which they did earlier this year. Notably, a majority of patients were low risk and presented with mild symptoms, patients for an average 37 years old and with no underlying health conditions. Patients who've vaccinated have received varying doses and vaccine type. Having a healthier and partially vaccinated patient population with largely mild disease has important implications in determining the viral kinetics and may have made it more difficult to evaluate viral efficacy, particularly as owing to the circle of patients in the study, several variants of concerns with different characteristics were enrolled. Cohort A and Cohort B were enrolled at different time points during the pandemic, during which the predominant circulating variants and rates of vaccination were changing rapidly. Different variants emerging during the study, notably Delta, may have impacted results with viral load level and the time points at which they peak during illness differ greatly across variant strain. In addition, the Cohort B treatment group was compared to a pooled placebo patient population, including placebo patients from both cohorts A and B. The time difference between the cohort enrollment created a placebo group that included patients with different COVID-19 variant strains and vaccination safety. As such, these factors may have contributed to the top line results announced this morning, which I will now detail. Turning to Slide 9. Top line data from MOONSONG Cohort A, which included 30 patients on AT-527 and 30 patients on placebo, did not show a difference in the reduction in nasopharyngeal viral load in mild or moderate patients. We did however show a reduction in viral load of approximately 0.5 log at day 7 in the prespecified subgroup analysis of patients with underlying health conditions. This aligns with the interim results we previously reported in our Phase II study in the high-risk hospitalized patients, where treatment with AT-527 resulted in rapid and sustained reduction of viral replication. These data from the Phase II study in high-risk hospitalized patients will be presented tomorrow at the International Society for Influenza and Other Respiratory Virus Diseases, ISIRV, which is a World Health Organization virtual conference being held virtually this week. On Slide 10, MOONSONG Cohort B was designed as a 3 to 1 randomization to the active drug at the AT-527 dose of 1,100 milligrams BID. Protocol primary analysis for Cohort B was designed to pull placebo patients from cohorts A and B in order to increase the amount of exposure of number of patients exposed to higher doses in order to evaluate safety as you minimize the number of patients randomized to a placebo arm. The operational plan was to enroll cohorts A and B in parallel, with an almost synchronous randomization in both cohorts. However, operational delays resulted into the delay in completing the enrollment of Cohort B as compared to Cohort A and created a highly heterogenous placebo group that included patients with different COVID-19 variant strains and varying levels of vaccination, which confounded a comparison to the actively treated patients. Moving to Slide 11. When you compare exploratory Cohort B AT-527 1,100 milligrams BID in the overall population versus just the Cohort B placebo and not the placebo -- pooled placebo group that I just reviewed, at day 5, you start to see a difference, which becomes larger at day 7. The reduction in viral load in the exploratory prespecified subgroup analysis of patients with underlying health conditions was again confirmed in Cohort B patients administered AT-527 at 1,100 milligrams versus placebo. There was a viral load reduction of 0.5 log in the treatment group versus placebo. Moving to Slide 12. Safety evaluation was also a key objective for this Phase II study. Importantly, the safety profile in MOONSONG was consistent with previous studies and AT-527 was generally safe and well tolerated. In the MOONSONG study, the proportion of patients experiencing any adverse event, or AE, was 20% in the placebo group, 20% in the AT-527 550 milligram BID cohort group, Cohort A, and 27% in the AT-527 1,100 milligram BID group, Cohort B. There were 3 nondrug-related SAEs in each of the treatment groups, and all other AEs were grade 1 or 2. Gastrointestinal or GI-related AEs were the most commonly reported, 8% in the placebo group, 7% in the AT-527 550 milligram BID Cohort A group and 17% in the AT-527 1,100 milligram BID group, Cohort B, with mild-to-moderate nausea or vomiting resulting in premature study drug discontinuation in 3% of the placebo group, 0% in the AT-527 550 milligram cohort and 17% in the AT-527 1,100-milligrams BID Group or Cohort B. No clinically significant differences in laboratory abnormalities were observed in the treatment arm as compared with placebo. Atea has completed a comprehensive nonclinical program to characterize the safety profile of AT-527. Results from these nonclinical studies demonstrate that AT-527 is non-mutagenic and has no effects on fertility and reproduction. Turning to Slide 13. PK data from the Phase I trial in healthy volunteers clearly demonstrate that a dose of 1,100 milligrams BID achieved top levels well above the target concentration in the EC90 in the large majority of patients. These data will help to inform our assessment of modifications to the Phase III MORNINGSKY study. Turning to Slide 14. Although Phase II MOONSONG study did not meet the primary endpoint for either Cohort A or Cohort B in the overall population, based on what we know now, we are not surprised by these data since other antivirals in development have shown a similar outcome in the mild or moderate patient population. What is reassuring is that there was a viral load reduction as compared to placebo on day 7 in the prespecified subgroup analysis of high-risk patients in both cohorts with underlying health conditions. Importantly, this is a patient population where these changes in viral kinetics have also been associated with improved clinical outcomes. Based on the totality of these data in patients with mild-to-moderate COVID-19 and those hospitalized with the disease as well as our data in healthy volunteers, the Phase III MORNINGSKY study is being reviewed for potential modifications, which may include modifying the patient population and changing the primary end point. We are committed to advancing this important program as expeditiously as possible as we believe AT-527 will be a safe, effective non-mutagenic therapy in the treatment armamentarium for COVID-19 as it becomes endemic. With that, operator, we'll now open the call up to your questions. Thank you.

[Operator Instructions] Our first question comes from the line of Umer Raffat with Evercore.

I have a few today, if I may. First, could you expand on viral load difference in unvaccinated patients specifically? I wish there was a slide on that in the deck today, #1. #2, were there any hospitalizations or severe disease? And how did that look in active arm versus comparator arm? #3, I noticed on your Slide 13, it looks like you're second guessing on the dose, because I recall the EC90s on 0.5 micromoles we had seen and from the prior data disclosures it look like your modeling was tracking well, well ahead of that, but the confidence intervals do dip below that EC90 at least. And I'm curious your take is on that, because the equivalent of that PK graph for the 550 dose would look very different. I'm curious what that looks like. And then finally, JP, I know there's been a lot of investor questions on sort of overall communication at Atea of late. There was a lot of confusion on why there was no disclosure on why FDA is not letting you enroll in Phase III. Now we learned that an 1,100 BID arm was added, and nobody knew about it. We would really appreciate any clarity on all of this.

Sure. Janet, you can take the first 2 questions and I will go with the last 2.

So thank you. Thanks, Umer. So your first question was asking us to expand a little on the viral load differences between the unvaccinated and the vaccinated population. And I think some of the difficulty, as I tried to explain during the presentation, is really the heterogeneity of the patient population. The vast majority of patients in the study were enrolled in the U.K. about -- I think over 70% of them. And during the course of the study, patients started to be vaccinated. The vaccination practices, as you know, in the U.K. are different, and so we have patients who received their first shots and then didn't receive their second shots for up to 3 months afterwards. And so I think many patients are partially vaccinated. And so the background rate is pretty noisy. And really, it changes from the 2 cohorts. And so we're really unable to provide what you're asking for in terms of clear numbers because of these differences and the movement in the background of the data. But essentially, we have more patients, as you would anticipate, being vaccinated in the Cohort B than we had in Cohort A, but there are not enough patients there really to be able to see a clear signal between the 2 groups. And then with regard to your question around hospitalizations. There were 2 hospitalizations, both for progression of underlying COVID pneumonia. And both of these were in the active group. But again, with the small number of patients in the study, I think this is probably what you would expect, and particularly, I think in the patient population studies, the rates are pretty low because of the underlying health of the patient population that we studied. Jean-Pierre, over to you.

All right. Thank you. Okay. The second part on the doses, I'd like to remind you that we were -- we did not indicate the exact dose that we were evaluating in MOONSONG, but it was clear that this was a study on viral kinetics where we will evaluate several doses. Also, I'd like to remind everyone that the reason that we had to do a slightly different clinical development is that the FDA did not allow us to go in the outpatient, did not allow us to go in other dosing regimen except the 550 milligram BID because of the similar chemistry -- chemical structure than the purine nucleotide that - in Hepatitis and BMS were developing in Hepatitis C. So first, we had to demonstrate the safety, which we did. Safety is not a topic any longer, at least to our knowledge, with the regulatory authorities. By the way, I'd like to remind that there has never been an issue outside the U.S. And so this was the first time that we could evaluate higher doses. Also, we hope that we can have data with the in-vitro evaluation in the variants by the earning call next month. And we wanted to see -- basically, I think that's something that we had indicated also at the end of Q2, that we had about -- only about between 40% to 50% were achieving the target concentration in our hospitalized Phase II clinical trials, and that we were going to look if we could increase that percentage, which we did with the 1,100 milligram BID. So I'm a little bit surprised that it was unknown because we led to believe -- and let's say, we share with everyone that we were going to evaluate additional doses to see if we could increase the number of patients achieving the target concentrations. And then I'd like to remind you also that, as you can see, and this was done in no more volunteers, and what we have seen also in hospitalized patient is that we have quite a significant variability as well. So -- and then the last -- your last question, Umer. We try to be as transparent as with everybody, without, obviously, without selective disclosure. Since -- unlike any company, I believe, we do not discuss our -- any detail of our interaction with regulatory authorities. I think that this is standard, is nothing different than anyone else in the field. And we continue to work to open sites in the U.S. I think there were some questions about -- I can tell you since we are on the webcast now is that placebo patients in high-risk individual studies, the regulatory authorities in the U.S. are concerned about it. The -- it's clear that they want more of a hard point rather than -- hard endpoint rather than the soft endpoint. So this is all feedback that we take very seriously with our partner. And that's why that this is part of our evaluation and potential modification in the MORNINGSKY Phase III study.

And just to clarify, the -- Janet, I think what I meant was, viral load difference among unvaccinated only. So within unvaccinated, how is the viral load different between active arm versus comparator?

So Umer, I think it's more complicated than that really because of the differences in the variant strain that we most likely have during the study. And certainly, I think when you look at the variant -- the placebo rate, the placebo viral loads between the 2 cohorts, there is a difference. And we believe that this is a temporal phenomenon related to different strain. So -- and that has quite a significant impact. So the viral load in the vaccinated patient population doesn't appear to be very different, but the background rate of noise is such that it's really hard to know.

Our next question comes from the line of Tim Lugo with William Blair.

This is Lachlan on for Tim. So I think you said in the past that you have the option for an interim analysis in the protocol for MORNINGSKY, so does this change your thoughts around doing an interim analysis at some point on the patients that you have already enrolled? To that end, can you provide an update on the enrollment? Just a specific number roughly.

So Tim, yes, we believe that an interim analysis may be something that we'll need to do. What we'll do with the patients who have been enrolled in MORNINGSKY to this point is something that we're and still rapidly trying to assess as we work out what our next steps will be there and what we -- whether we'll -- what potential modifications we'll make to the protocol. But I think as we take the study forward, certainly, we've watched what others have done, and we think that an interim analysis at a certain point for efficacy maybe something that we would really want to do.

Okay. Great. And then if I can follow up with another one. You've mentioned the differences in variants obviously between cohorts could be impacting results. Do you have more granularity on that? Or when can we expect more granularity on that? And to -- in this point on the vaccination, the different vaccination status, how are you thinking about enrolling patients with or without vaccines moving forward? Would you be changing any criteria there or anything to try and control for some of this noise that you're seeing?

Yes. So I think that's one of the things that we've clearly seen here, is that we don't need that sort of background noise. And I think moving forward, one of the things that one would want to do would be to eliminate enrolling patients who have been vaccinated as we move forward with the study.

Got it. And on the variants more, when can we expect...

And on the variants, we'll have the sequencing data, we believe, hopefully within the next month or so, and then we'll be able to provide greater granularity around that.

But in the same time, Tim, I just want to emphasize that we believe that -- and based on the data here that we are definitely active against the variants, I don't know to the -- if there is maybe a variability. Those -- the challenge with those in-vitro assay also is look, it's a variability of 4 to fivefold. And what we see overall, when we have a 0.5, but the -- those assay varies between 0.2, 0.225 to 0.9 depending on the assay, which is nothing surprising. So as long as we have such kind of variability, we are still awaiting the result of the Delta. We -- and as Janet has indicated, one of the challenge was that the 2 placebo group and the Cohort A and Cohort B are either basically, it's heterogeneity. It's not just a variant. It's the variant. It's the vaccination. It's the high risk, low risk. It's the very high level of 0 positivity. But you see basically in excess of a log and almost a 1.5 log both in term of viral load at baseline and viral kinetics, the difference between the 2 placebo arm.

[Operator Instructions] Our next question comes from the line of Eric Joseph with JPMorgan.

I just wanted to follow up on the progress with MORNINGSKY. Can you just shed a little light on how many subjects have been enrolled in that so far? And whether any -- their data could be used going forward in the revised protocol for the Phase III, what you need to -- and would you need to upsize potentially better onto [ 6 feet ] on the revised primary endpoint.

Janet?

So patients are continuing to roll into the study at the moment. But we haven't provided comments I think or information previously as to the numbers. And I don't think we're in a position to be able to do that today. We are also looking really at ways to potentially modify the study. And we're going to have to determine what that will mean in terms of sample size and so forth as we make those decisions. And we will have to communicate that subsequently.

Okay. And just a follow-up, if I could. Previously, you've talked a little bit about the discussions with U.S. government and other health care authorities globally. Can you just talk about where those communications currently stand and how they may proceed going forward or whether they might be conditioned on Phase III results?

Sure. John, do you want to address that question?

Sure. So yes, our discussions with government are ongoing. And we have provided an update to the government agencies, so to the U.S., with regards to the MOONSONG results. And we continue with our positive discussions with them. So we -- our discussions are continuing and are positive.

Our last question comes from the line of Roanna Ruiz with SVB Leerink.

I just want to ask regarding the higher 1,100 milligram dose in MOONSONG. Could you elaborate a bit on how you selected that dose? And is it possible you could push the dose higher in MORNINGSKY with future amendments?

Look, I will answer part of the question and Janet will go from a clinical standpoint. As you have seen, Roanna, there is quite variability from the -- from a PK standpoint from patient to patient. And so we wanted to really see if it was dose proportional with the 550 milligram, actually, it is. And also in term of really what we achieved related to the EC90. And you see we are in the range of 290 gram paramount, which we think that it's a good achievement there. Related to a higher dose, I doubt it, but I'm going to let Janet also address that from a clinical standpoint. Janet?

No. I think we think probably 1,100 milligrams is as high as we might want to go. Secondly, I think as Jean-Pierre mentioned, there is a very nice dose proportionality. But I think at the dose of 1,100 milligrams, the vast majority of patients are achieving levels above the EC90 level that we believe will be necessary to target. And I think also one's starting to see a little bit of GI tolerance creeping in. And so I don't think I want to take it any higher because I think that might become more of a problem where we to go any further. So I think 1,100 milligrams is probably as high as I would want to see it go.

Maybe, Janet, we can take the opportunity here to mention that there was -- since we did not expect on the site at no view on those notes there and being -- in fact, you think that we are going to make some recommendation now. So go ahead, Janet.

Certainly. I'm sorry, I should have said that. I think I mean, Roanna, one of the things was that we were not anticipating the GI effect. And actually, they're pretty mild for the most part, but completely unanticipated. And we believe they will also be quite manageable now that we know to anticipate that there may be some nausea in particular. And so I think that can be managed with antiemetics and such therapies and also with counseling and advice to both physician and patients. So we anticipate that those rates will actually drop as we move forward with trials because we can provide that information to all participants. So we hope that those rates will not be worse than they have been in the study, but we believe it's a manageable adverse event.

Great. And last question from me. Was there anything else in the patient baseline characteristics that you might seek to adapt or explore any amendments for MORNINGSKY?

Janet?

So I think the things that we've seen which really correspond exactly with what others have observed in the space are really that younger patients and patients without underlying risk factors at -- and patients with low BMIs, et cetera, and vaccinated patients tend to do pretty well. And it's really the high-risk patients where one is going to have the greatest likelihood of seeing an impact clinically and probably also from a viral kinetic perspective. And I think, unquestionably, if we are seeing that benefit in the high-risk patients, it's likely that there is benefit that is being created in other subgroups of patients. But certainly, I think in terms of taking the drug across the finishing line quickly, this is the patient population where one is going to be most easily and quickly able to show that. And so I think that's really the big take home from this.

There are no further questions at this time. I will now turn the call back to Jean-Pierre Sommadossi for closing remarks.

Thank you again for joining us and for your continued support of Atea. Thank you again.

Ladies and gentlemen, this concludes today's conference call. Thank you for your participation. You may now disconnect.