Sage Therapeutics, Inc. (SAGE) Earnings Call Transcript & Summary

Great. Thanks, everybody, for continuing to join on. It's my pleasure to be hosting a discussion with Barry Greene, CEO of Sage, who I and probably everyone listening have known for a long time since his Alnylam days. Barry, thanks again. Maybe you want to just do the sort of obligatory 3- to 4-minute kick off and give an update on zuranolone, SAGE-324, SAGE-718 and then we can get into more specifics.

Yes, absolutely, Paul. Thanks, and I'd like to thank Stifel for having us here today. As we set out in January and established the goals for this year, we have executed on the expand and accelerate strategy we set out. We've seen, as you know, positive readouts for zuranolone, SAGE-324 and SAGE-718. It's quite remarkable. And we've gotten clarity with the agency for our filing path along with filing path for zuranolone in the second half of 2022 with the rolling submission starting early in 2022, and an additional associated submission in PPD. So we feel like we've gotten real good strategic alignment with the agency and real good clarity on moving forward. And of course, we're on track to initiate the Phase II study for 324. In addition to the Huntington's study, the Phase II placebo-controlled study for SAGE-718 Huntington's, we've articulated very recently that we plan on initiating a Phase II study in Parkinson. So we're advancing the expand and accelerate strategy with a very strong balance sheet. We're making smart decisions. Behind the lead programs following the Sage methodology is a pipeline of Sage advented organic programs that we think have to a higher probability of success than most because of how much has been elucidated in brain health and driven by Sage for NMDA and GABA. So setting what we believe is in pretty good shape right now and looking forward to finalizing the year with some data readouts and then coming back in January and articulate what 2022 and beyond look like.

Yes, yes. Okay. Awesome. Well, maybe let's talk a little bit about CORAL and the recent change that you made at day 3. What was the thought process in changing the endpoint? And how does this change primary endpoint kind of impacts the importance and how we should interpret data days 15 and 42?

Right. So if we take a step back, Paul, we articulated in January that given the current data set at the end of 2020 that we set a path to motion with landscape in that. So 3 different Phase III studies, 2 in MDD, 1 in PPD that we believed any one of which a positive gave us the filing package. And I'll remind you that WATERFALL was studying zuranolone on top of either nothing or a stabilized antidepressant. CORAL, as you know, is concomitant use with an antidepressant. So de novo use or patients washed out. And then SKYLARK is asking basically the same question as Robin did, but we wanted another PPD study. But with WATERFALL hitting, we confirm with the agency that we have a filing package. We've articulated that we're going to file first in MDD and then a supplemental filing in PPD that based upon the time lines, we have an opportunity of getting both indications approved at the same time. So we feel like we're in really good shape. Given that and given the fact that other than safety, CORAL wasn't necessarily required for the filing, but it was sitting out there. We decided to really move CORAL to what its original intention was, which is, can you jump start someone getting well that has depression. As you know, standard of care, if it works, takes 6 to 8 weeks. Many people don't make it that far because they suffer things like weight gain, ***ual dysfunction along the way. So by giving zuranolone with a standard antidepressant, can we accelerate someone's path to wellness, as early as 2 or 3 days. And that's what we're really trying to look at there. So that will give us data that should help prescribers out there with zuranolone as a monotherapy, zuranolone on top of the stable antidepressant or if someone feels like they want to get both at the same time because they believe that antidepressants might prevent the next depressive event, but they want someone to get better faster, we'll have data for that as well. Strong -- we believe a strong data package.

Yes. Yes. Okay. Okay. Great. All right. The obligatory question on durability, day 42, how to think about those data, regulatory significance. I guess we've talked about how for ZULRESSO, right, the FDA was pretty agnostic to the reality that in one of those trials at day 30, actually placebo did better than drug. And I think a lot of investors, you can forget about that, right, that, that is -- the lines crossed, right? And that theoretically sets a great precedent for zuranolone as it relates to the importance of how to interpret those data. That said, MDD has a greater chronicity to it than PPD. So I guess from your perspective, how confident are you from a regulatory perspective that some of the Wall Street folks who say you need a p-value or you need better separation there are just kind of dead wrong in terms of what the agency cares about?

I mean, Paul, I think you got it right. What we're looking at and what medical folks believe and the agency believes is that if someone can get better and stay better, that's what we're looking for. Having a robust placebo effect, which we're seeing across all depression trials is a part of running depression trials. It's inappropriate to use the words, the drug loses effect. In fact, at day 42, we saw 86% of the day 15 effects. The drug is not losing effect, having low dose placebo narrows that. Now what we don't want to see and we're not seeing this is that someone gets better, they stay better at 15. And then as soon as they're off the drug, they quickly rebound back to baseline. If we saw that in a majority of patients, that would be the opposite of what we're looking for. We're looking for many, many patients, most patients to get better, fast and stay better longer. And the SHORELINE data that we've articulated, the fact that 70% of patients in the 30-milligram dose only required 1 or 2 -- 2-week course of treatment with a 70% response rate for 50 of that data are coming out later this year but we saw an 80% response. So hopefully, we'll continue to see that hold up. But there's evidence that these patients get better and stay better. And if you think about it, it's almost a reverse placebo. They know they're not on drug, and they're not seeking to get back on drug or feeling poorly enough to get back on drug. They are reporting that they're staying in their kind of normative range, which is profound.

Yes. Yes, yes. Okay. Fair enough. So I guess, as it relates to SHORELINE, can you talk a little bit more about -- I guess it's kind of hard to talk about out of panel without really looking at data, but I guess -- I think one of the interesting things that we've tried to think about, right, is do patients who get better or do patients who get worse, stay in that trial over the long term? What can you say about the retention rate of the study and the quality of the data as you go out and your ability to kind of make conclusions after a year or so of follow-up?

Yes. I mean when we come out with a 50-milligram data later this year, to your point, we can be a lot clear about the percent of the patients that stayed the full year on the trial. But it's very, very high. We've seen that in the Phase III studies as well, almost placebo-like facts is not seen large numbers of dropouts. Now it's possible, Paul, that someone that took their 2-week dose and has stayed better is feeling better and doesn't necessarily believe that they want the overhead of a trial. We might lose some of those, but the retention rates are quite high.

Okay. Okay. Great. And in your regulatory discussions, have you gotten feedback on anything related to how many patients that the FDA wants at a certain number of redoses to kind of have a broad label that doesn't have limitations on it?

Well, so what we're looking for is -- it's too early to talk about specifics of the label. We're looking for the treatment of MDD. We are looking not to have huge numbers of limited doses. So we know that the overall safety database is large enough and then how much data we'll have at each dose retreatment is a matter of how much data we'll have at that time following its review discussion and a review issue. We don't think we'll be terribly limited. Although the good news here, Paul, is that even if there was a scenario that said we were limited on a number of doses, getting to fourth and fifth dose is very, very rare. And those are data that we could supplement later if that, in fact, was a limitation. We think most patients in our kind of commercial estimates, if you will, most patients will be at that around 2 doses, 2 -- 2-week dose per year based on the data we have today.

Yes. Okay. Okay. Very good. And I guess what are the other studies that you're doing, especially for 50 mg that are gated to an NDA filing? And how might the outcome of those trials be articulated?

Yes. So we chose to repeat a few of the studies of 50 milligrams that were done at 30 milligrams, some of the pharmacology studies. But we think we're in really good shape. I mean this is not an area we're going to argue to remove a warning that says, don't operate heavy machinery or drive the next day until you understand the effect of the drug. It will be a scheduled drug. So these are almost check box type studies that we need to repeat because we're not doing them to remove any warnings from the label. In fact, we want those warnings and label are the right things to have in the label for a drug like this.

Okay. Makes sense. I think that, that machinery commentary might even be on the label of SSRIs, too, not just benzos.

And even Benadryl.

Yes. Fair enough. So then, I guess the other sort of question right, is -- and you and I have talked about this, right? It seems like there's a lot of investors, not just in depression and most categories of those generic drugs and the new branded drugs sort of just by default model the use of a drug in the treatment failure population, the people who tried cheap long-standing generics. And for SAGE-217 or zuranolone, right, you guys have studied the drug and early lines of depression. And you, Barry, right, with a lot of commercial expertise and experience have talked about this drug being used even in primary care. And again, I think a lot of investors and analysts just assume this is a fourth-line therapy. What's the disconnect here? And what makes you confident that this can be an early line drug that ensures okay with utilization after maybe failing an SSRI, for example?

Yes or not, maybe even as monotherapy when newly diagnosed, which is a good place to use a drug like this. Well, Paul, if we step back, let's look at the evidence. While some might believe this is a well-served patient population that can be treated with generics, the facts just don't bear that out. Depression is rising at an ever-increasing rate. It was rising quickly with the COVID pandemic into a fourfold increase from pre-pandemic level. So if we had drugs that worked effectively, that would not be happening. We would not see dramatic rising of cases. We know that the average patient takes the drug for about 7 weeks. We know that the leading cause of discontinuations are weight gain or ***ual dysfunction depending on gender. And we know that these drugs take 6 to 8 weeks to work. So if it's you, someone in your family, people that have kids out there, particularly someone that might be off in college, who wouldn't want them better and the chance to get better in 2 or 3 days. It just kind of makes -- it just makes sense. The idea of they're cheap, but let's treat the fail is actually economically not viable because, as you know, depression that either gets diagnosed late or treated late is a leading comorbidity for other downstream illnesses like diabetes, cardiovascular disease, even infectious disease as COVID, for example. People with depression have higher case of COVID because it messes with their immune system. So medically, getting someone better, fast and keeping them better, medically is the right thing to do. And we think as we educate on the totality of the data, that's incredibly -- that becomes a very convincing argument.

Yes. Is the VBA approach you did viable here? And if so, what would it take to actually implement and manage that?

Yes. I mean we've actually stated this that our intent along with Biogen is to lean in, if you will, and proactively launch with value-based agreements. So essentially, that allows or that has us talking to payers about what issues or risks they might have in allowing a prescription for zuranolone. We start value-based agreements, eliminating those. So for example, if they're worried about the patients needing 4, 5 or 6 doses at a population level, we can protect them about against that. If they're worried that they require zuranolone and then other antidepressants or high expensive other therapies, we can protect them against that. And some of the preliminary discussions we've had are very positive in terms of payers acknowledging of VBA for a large category actually desiring it. And then many have systems or capabilities on their own. Many might need us to do the analysis, but it can be done with ICD 9 and 10 codes at a population level. So it will take some capability to implement, but this is a classic area where a value-based agreement makes lots of sense for both sides. And you've talked to payers. They don't want sick people in their plans. They want people to get better. They just don't want to pay for an expensive drug that doesn't work. And by taking risk, we're protecting them against that.

Yes. Okay. Okay. Well, could you give an example of what that might look like, like agreeing -- like could it be something like agreeing to an insurer that he will pay for it, if a patient needs a third or fourth course of therapy, like what are some examples?

Yes, there's an example. I mean, in general, if you think about someone having a certain number of covered lives, we can agree based on epidemiology, how many patients might require a drug like zuranolone. So there's a volume of patients. There's a number of doses, and then there's the outcomes. Do they take this, but need other things that the payer has to protect them on. We can put that in a value-based agreement so that they get rebates based upon outcomes rather than predefined rebates just for access, which makes no sense at all.

Yes. Okay. Okay. From a pricing perspective, to accomplish this goal where whatever the prior auth requirements are easy and not cumbersome enough that a primary care physician would really be willing to do it. Do you need to price below specialty tier on an annualized basis?

It's really too early to talk about the specifics of pricing yet, Paul. But we will be conscious about ease of prescribing, numbers of patients as we think about price.

Okay. Okay. Okay. Great. Anything else on zuranolone that we didn't cover?

No, look, we are thrilled that we have a very clear path for filing with a rolling submission being acknowledged and asked for starting early next year. And having clarity around CORAL gives us great enthusiasm for zuranolone moving forward. With the filing underway or the NDA team, obviously, we and Biogen are now moving towards the commercialization efforts and how to make sure the markets are prepared, payers are prepared, patient advocacy are prepared. So there's a heavy lift in front of us, but it's a lift that we're capable of doing with Biogen.

Okay. So 50 mg SHORELINE a day by any year and then CORAL early '22?

Correct.

And have you said of CORAL, I mean I would assume you said it completed -- have you said it completed enrollment?

Yes. It completed enrollment and all sites were closed to future enrollment. So when you close the site down, there's obviously some patients in the queue that we have to factor in. We don't want to shut the site down and then have patients sort of hanging out there. It's not the right thing to do for them or the sites. So that's what pushed us into early next year.

Yes. Okay. Okay. Okay. Sounds good. Switching gears, should we talk about SAGE-718?

Absolutely.

So maybe just to set the stage there, can you just walk through the data you generated so far? And the evidence that this drug may have a procognitive effects?

Yes. So I mean, you'll remember that Sage discovered that in certain Huntington's patients, there was a lack of 24 hydroxycholesterol as cognitive decline came on. That gave us kind of a natural metabolite to understand if increasing that through positive allosteric modulation of NMDA could have benefit in executive function, learning and memory without the stimulatory effects that you see with other kinds of medicines. So that was the biologic hypothesis. What's exciting is that all the preclinical work, the early clinical work and the ketamine challenge studies as well as both Huntington's and Parkinson's, now demonstrates that people with 718 improve executive function, learning and memory and then don't have sort of the stimulatory effects, as I mentioned. That -- those data sets have continued out. We read out Parkinson's that read the same way. We'll have Alzheimer's at the end of the year. And the data has given us confidence to start a well-controlled, placebo-controlled study in Huntington's as well as recently announced a well-controlled study in Parkinson's disease. So we have at least 2 paths there for initial indications with Huntington's being first. We'll look at the Alzheimer's data later in the year. And if that gives us hope, then we'll likely start a study there as well.

Yes. So this Alzheimer's data is the next event for this drug, and it's an open-label study. So what would you be -- and I guess in the short-term trial with Alzheimer's right, patients tend to not get worse. So what are you kind of looking for to be confident that the hypothesis could extend to [ ADL ].

Right. So if you think about what we saw with Huntington's disease, not only did they not get worse, they got better in 2 weeks, which is quite remarkable. We saw that repeating in Parkinson's. And Huntington's is a fairly homogeneous disease relative to cognition. Parkinson's is a little bit less in both executive function, learning and memory, whereas Alzheimer's more of the learning and memory upside of the equation with a much more heterogeneous population. So what we're looking for is do we see in some a majority of the patients actually improvements in learning and memory, not just lack of decline but true improvements. If we see true improvements and even in a subset of patients, that gives us a lot of confidence to think about what the right patient group to study is because, as you know, in a couple of weeks, memory improvement just doesn't happen with anything on the market or in development today.

Yes. Yes. Yes. Okay. And in terms of Huntington's and Parkinson's, what percent of these patients present with cognitive symptoms early in the disease? I think investors listening it might be used to kind of hearing, obviously, in the Parkinson's type drugs that largely target the motor symptoms or in Huntington's even also sort of may look at like chorea or try to get genetic-targeted therapies to the deep brain. So are these -- I guess, maybe really what I'm trying to ask is would this drug be for the whole populations? Or would it be some sort of subset?

Yes. Let's start with Huntington's. If you talk to neurological Huntington's experts, they will tell you that the biggest issue they have with their patients early in disease is cognition. That's what goes first before some of the other symptomology sets in. And people can be in their 30s, late 20s, 30s, early 40s, after being diagnosed where cognition starts to slip. And that's important, Paul, because even before some of the motor issues step in, people can lose independence because they can't maintain activities of daily living. Parkinson is a little bit more split in terms of executive function, learning and memory. But there's lots of stuff for Parkinson's right now that can work fairly well in terms of dopamine replacement to stop some of those other symptoms. But really, for a subset of patients that start losing the executive function or learning or memory, there's nothing to help that part of Parkinson's. And obviously, we all know about what happens in Alzheimer's disease.

Yes. Okay. Okay. Very good. Should we talk briefly about 324?

Sure.

So I think the controversy around that compound, right, as it relates to the initial data you saw a nice signal in tremor, but maybe semi high rates of sedation and things like dizziness, and I think right or wrong, right. Some investors and analysts are concerned about therapeutic index in an elderly population that might be more vulnerable to the consequences of adverse events like this. So can you sort of speak to why you think that might have been an artifact of just this study and the way you dose the drug? And what your goal would be for a subsequent trial in terms of mitigating these types of things?

Yes. So let's talk about the subsequent trial. So right now, we're initiating a Phase II study testing various doses and frequency so that we believe that in addition to that 35% to 45% decrease in essential tremor, which also led to an improvement of activity with daily living that we have a drug that people can take chronic. So that's what we're studying now. We're very confident based upon our read of the data that we'll get there. They'll have both efficacy and then the drug that someone can take chronically because it is a condition that requires chronic administration. The KINETIC study was a study done to pick the highest dose that we felt possible because we were looking to see was reduction in essential tremor over 30 days. Prior to that, we had sort of a single dose and reduction of essential tremor that tied to serum levels over a period of time, but as serum wade, tremors came back. So this gave us a view that we can dose for 30 days, we saw no tachyphylaxis at the highest dose. And I'll remind you, Paul, that to get really good pharmacokinetic data tied with the PD data, we dosed during the day. So if you give something that makes someone sleepy during the day, they're going to be sleeping. But the risk of running a study like that at night is that you have to wake people up to get blood and they're going to report that they're sleepy when you wake them up. So it's one of those Phase IIs that no matter what, you're going to have a high level of [indiscernible]. We are exploring evening dosing. And as we did with zuranolone, we've seen that when you move to evening doses, people take their drugs during the evening, they go to sleep, and they're usually pretty good. So we think that dose frequency and a move to evening dose will be part of the solution set here to give us an efficacious drug that can be dosed chronically.

Yes. Okay. Great. And so what's the timing there?

We're going to start that study before the end of the year. So you'll hear about that shortly.

Okay. And in your last earnings release, I think I remember you mentioning epilepsy as well for this molecule. Is that something you could be moving forward soon?

Yes. We really wanted to lock in on essential tremor with Biogen. And once we're sort of set there and we have that dosing frequency, epilepsy and other movements disorders start opening up to a drug like this. So that's in the future, but not in the near term. What I'd like to do is demonstrate that dose and frequency, which I'm confident we'll have. And then that opens up a number of other potential indications for 324.

Yes. Yes. Okay. Okay. One question that I got from related to NMDA, I guess, Barry, how do you make sense of a number of the setbacks for things like D-cycloserine, the NERC and Takeda compound. I think there's this sort of view out there, right, that NMDA-targeting medications have had this great mechanistic rationale for a while, but it hasn't really panned out all that well. What do you think and how SAGE-718 different?

Yes. I guess the way I'd put it, Paul, maybe it would be a metaphor and if you -- it will be winter in New York. So if we drive down the streets of New York in February, we'll see apartment buildings with their windows open because they've got those old -- heat on or heat off. So the approaches of NMDA have been hit it hard or don't hit it. What we've done is put in a thermostat. So we are dialing in a very specific allosteric upregulation of NMDA very precisely with SAGE-718. It's just a fundamentally different approach. It's rather on and off. It's -- I'm going to set the room temperature at 72, the room temperature is 72.

Got it. Okay. Okay. Great. What else, Barry? Anything else that you would like to highlight in the last minute or so?

Thanks for all the questions, Bob. I'm pretty thrilled with the execution we've had this year, our ability to continue to execute. And we've got a pipeline of Sage invented organic programs, all of that have very different features that we're moving forward on. So obviously, we've got to get zuranolone right in 324, right? I believe, and you and I've talked about this, that SAGE-718 is fundamentally overlooked in terms of the massive value it can create. And if you think about a drug that can give people months to years of independence relative to the natural history for other drugs on the disease, that's huge. And then we have a whole pipeline behind that with the balance sheet to execute.

Great. Awesome. Well, thanks a lot. Always a pleasure talking to you and look forward to the next couple of events in December and January.

I appreciate it. Thanks for the interest. Take care. Bye.

Thanks.