Okay, good afternoon, everyone. David Amsellem here from the Piper Sandler Biopharma Research Team, and this is day three of the Piper Sandler Healthcare Conference. We're delighted to have Alkermes and CEO Richard Pops. Thanks, Rich, for joining us, so we got a lot to talk about, so let's dive into Vibrance-2, shall we? But maybe before we go into the nuts and bolts of the study, I'll start with sort of a high-level question. I think a lot of the investor community, maybe we're looking for a little more potency, if you will, on maintenance, away from this, and Epworth. But we also know that NT2 and IH are not the same as NT1. So I'd like for you to contextualize the efficacy results that came in NT2 for Alixorexton.
So first of all, good to see you.
Good to see you too.
The Piper Sandler Research Team. It's good to have everybody here in the room together. Before I plunge into the minutia, I think it's important to plunge out of the minutia because I think investors, in particular, certain investors, are just wallowing in these fine details. I think first principle question I can ask you, based on the Vibrance-1 data, what's your assessment of the probability of Alixorexton being approved for NT1?
Quite high.
Yeah, I agree with that.
Yeah.
Based on the Vibrance-2 data, showing that we could show statistically significant and clinically significant changes in our first study, what's the probability of our being able to get labeling that also extends to NT2, do you think?
Quite high.
Okay, so stop right there and say, therefore, we have a drug that's going to be a transformational drug for narcolepsy.
I agree.
So there you go. And if that were accommodated in our evaluation and people began to realize what the economic implications and the clinical implications, the medical implications for patients and for our company, then I'd be happy.
Yeah.
Because right away we plunge into what this looks at week four compared to week three, compared to this, compared to that, compared to another company who doesn't have data compared to this. It's actually silliness.
Yeah.
And so, from our perspective, developing the drug over a long wavelength, and I was thinking about it as I saw you. I think this is like the third year we've done this.
Yes.
And if you put those three dots on a graph of what's, excuse me, transpired in those three years, it's this amazing ascending arc of conviction and data and rigor that leads to our belief that we've got really important medicine.
There's been a lot of de-risking.
Exactly.
I've said it.
Exactly, so I think from our perspective, we are preparing to launch a drug into the hypersomnolence space, which is entirely why we announced the acquisition of Avadel, because we're all in. We're thrilled with the data in Vibrance-1 and Vibrance-2, and ultimately, getting to the question you asked, Vibrance-2 was de novo clinical research.
Yeah.
No one's ever run an experiment with an orexin 2 receptor agonist. At doses that were deemed to be therapeutic doses across this variable patient population over an eight-week period of time with a five-week follow-up, looking at a range of doses and fixed dose arms for multiple weeks, and what's amazing about that is you learn new things.
Yeah.
And so first, your first question, potency. First of all, let's get the semantics right.
Sure.
It really has nothing to do with potency. You're really talking about the degree of efficacy. And what you're talking about is in one assay, the Maintenance of Wakefulness Test, where we saw a number which represented a mean number, which is adding all the observations together and dividing by the denominator. In contrast to NT1, where that mean number is actually a measurement of central tendency, in NT2, because there's so much variability, that it's just a number. We have people who respond really, really high on the MWT. We have people who don't respond so much on the MWT. It may be related to their baseline levels of orexin in their brain compared to the dose they got. This is all new area to be. But overall, across that variability, the signal powered right through with a p-value and a numerical difference between the groups.
When you look at the individual plots within that study, you see amazing new information. You see that a lot of patients don't respond on MWT, but respond on ESS.
Yeah.
70% of the patients at the 18-milligram dose achieved normalcy on the ESS score, which is their assessment of how they feel. Yet the MWT is much more variable. What does that mean? We don't know.
Yeah.
There's also site-specific effects that happen in various countries and in various sites. But the overall tendency is through all that complexity, if you look at the two groups, you see a statistically significant and numerically significant difference in the two groups. Now, with 93 patients of data over eight weeks with follow-up data, we actually have more information now to tune the phase III data study to get the outcome that we're looking for. So long answer to a simple question, we're thrilled with the NT2 data.
I think there's something interesting that you brought up when you announced the results and this concept of disease heterogeneity in NT2, and I almost think of NT2 as a different disease than NT1, and I think we in the financial community need to think about it that way, but talk about disease heterogeneity, because I think it is really important here. I think you mentioned Orexin CSF levels in these patients are variable per literature, so what does that heterogeneity mean, not just for ALKS 2680, but the Orexin 2 category more broadly.
I think that in the fullness of time.
Yep.
The orexin 2 receptor agonist will actually be, we'll use them as a probe for a more accurate differential diagnosis. Because as you know, NT2 is a bit of a catch-all.
Yeah.
People who present with hypersomnolence, excessive daytime sleepiness without cataplexy are deemed to be NT2. And NT2 in the IH designation can flip-flop depending on the assay and the day that they're tested. So it's an imprecise definition. So we knew going in that we would expect a higher degree of variability. And we knew from previous work that these patients would therefore require a larger dose in order to drive efficacy across this variability.
Yeah.
So, I think that what we're learning, when we sat down with the KOLs when we got the Vibrance-2 data, I think, A, the response was overwhelmingly excited because no one's actually never run a study just enrolling pure NT2 patients in this way with these doses. And so to see the signal emerge, but also I think it underscores our fundamental belief that in the commercial setting, what you're going to want is a range of doses.
Yep.
In a phase II setting, what you do is you keep people in a dose lane for an extended period of time to learn. But you take those learnings and you start designing your dose symmetry for phase III in commercial, and you realize what's going to happen, like many variable diseases, there'll be a recommended starting dose likely for NT1 patients. There'll be a recommended starting dose for NT2 patients.
Yeah.
But there'll be a range of doses to vary around.
Yeah.
That will include the new knowledge from the study, the split dose.
Yep.
Because that's the other finding from the clinical research was the difference between the NT1 patients and NT2 patients with respect to this tolerance that builds over time.
Yeah. And I wanted to delve more into that. What's interesting in talking with key opinion leaders about dosing, you hear a lot from KOLs about dosing flexibility. And the idea that, hey, split dosing might actually be a good thing in practice. Patients like to have control in calibrating their daytime wakefulness. So I wanted to get your thoughts about that idea. And what I'm saying here is, what I'm asking is, to the extent that Alixorexton is a split dose product in NT2, maybe even in practice in NT1, how do you think about that commercially?
That's why we still believe that a QD dose, a once-a-day dose, is a really strong anchor dose.
Sure.
Many patients will love that convenience. Take one dose a day, you're good. Through the course of the clinical trials, particularly with our European sites, they were the ones in the beginning, even NT1, said, we have different patients with different needs.
Yep.
In some countries, people have dinner at 10:00 P.M. and stay up till midnight, and if they're getting up at 8:00 A.M., no one dose is going to be appropriate for that person compared to the person who wants to go to bed at 8:00 P.M. every night, so even from the earliest days of the NT1 study, where we have a really clear once-a-day product, the idea of a top-up dose or a PRN dose was emerging, because what was happening is these patients were experiencing a degree of wakefulness they've never had before, and they want to keep it, and so it made sense to us.
Then when the NT2 data came in, it just became explicitly obvious because in that study, in contrast to the NT1s, when you looked at the individual plots of the patient's response over the eight-hour time period of MWT, the signal was driven by hour two and hour four.
Yep.
Then it became chaotic thereafter. So we could address that with higher dose. The most parsimonious way, the simplest way, is so if you increase the dose, you're going to increase all the AUC in order to get that extended tail. The more parsimonious way is just split the dose.
Sure.
If somebody wants to have six- and eight-hour time points higher, split the dose. And I think that'll turn out to be in the commercial setting part of this beautiful flexibility because we haven't run into, in the NT2 patients, any dose-limiting toxicity.
Right.
At the doses we've tested.
So got a few follow-ups here. So just regarding that variability in six and eight hours, does this go back to the discussion of disease heterogeneity, or is it something specific really to the orexin category as it relates to NT2, or is it just something you think that might be specific to the molecule here?
I think it's the disease. And I think it's.
Yeah.
Our hypothesis is that it's like other. Think about other pharmacological interventions that. So take testosterone or dopamine. If you administer exogenous agent to somebody, what happens? Your normal homeostasis, you make less of it inside. So when we started administering an orexin 2 receptor agonist to patients with a orexin tone, unlike the NT1s, people have orexin in their brains. Our observation is that over the first four weeks or so, there's a reequilibration of the dose that's necessary to drive wakefulness.
Yeah.
Because, and why do we know that? We know that because at our time zero time point in our phase I-B study, we saw no diminution at the six-hour and eight-hour time point.
Yeah.
But at week four, by then, it had eroded. And so the answer to that is just to raise the dose or split the dose.
Yeah.
We have the capacity to do both given the tolerability profile.
Do you know, have you decided how far apart you would administer Alixorexton in phase III?
Absolutely.
Okay.
Yeah.
But that's not happening. We're not going to know here.
The thing is, that's the virtue. If you do, if you spend the money and the time to do a proper study, there's nuggets in there of gold, and they're worth what you pay for. So both in terms of the dose, the way to model. So think about it this way. Now we have 93 patients of variability. And we can calculate and model now the appropriate doses that'll capture, let's say, 90% of that variability. And without having, there was a circumstance, let's say we had run into a dose-limiting toxicity above 18.
Yep.
Then we would have had to figure out how to stay within that window.
But you didn't.
But we didn't.
Right.
We have enormous flexibility, and we have the data to model it. That's proprietary, and we're going to keep it that way.
Okay. So, but here's, I'll ask the question a different way. And if the response is consistent at four hours post-dose, doesn't it stand to reason you're going to administer dosing more than four hours apart?
Maybe.
Yeah.
Yeah. But it actually, it doesn't require a supercomputer to figure this stuff out.
Sure.
When you look at it, it's pretty straightforward.
You certainly haven't run into a dose-limiting toxicity, so theoretically, you can push the dose.
Correct.
The total daily dose and split it.
That's right.
That is something that we more likely than not will see in the phase III.
Yeah. I think you can expect a couple of things. There'll be an anchor QD dose.
Okay.
Because the 18-milligram dose is a beautiful dose.
Yep.
We can get the drug approved, we believe, on the 18-milligram dose QD.
Okay.
But why not engineer some of this more flexibility into the program? So at first approval, we'll have that in the label.
Yeah. Maybe you could do 12 and 12.
Maybe.
Maybe. Okay.
Maybe.
You know, what's interesting, and this is a sort of an observation that a lot of KOLs make in all their years of using modafinil. Modafinil was approved as a once-daily drug. Mostly in practice, though, it's used twice daily.
Yeah.
So the dose is split. So what's interesting is that, and I'm not comparing the orexins to modafinil. This is more potent, obviously. But we see split dosing in practice a lot. Is that, does that sort of square with what you're hearing from KOLs?
Yeah. We actually know from a competitor's study that was testing a twice-a-day dose in phase III.
Yep.
That in the course of the multi-week study, patients would time their second dose depending on what they wanted to do that night.
Right.
So I think that stands to reason. You see that in ADHD as well, where you see split doses.
Sure.
People tune the pharmacology to suit their lifestyle. That just makes sense.
Why not just spend some time on MWT versus Epworth? Some KOLs really don't like Epworth. But looking more at Epworth in the course of the Vibrance-2, it seemed like there might have been a bit of a waning over time, particularly at the 14-milligram dose. So I just wanted to level set how you're thinking about the potential for tachyphylaxis.
We think that the tachyphylaxis phenomenon occurs between day one and week four.
Yep. Okay.
Our analysis of the data from week four to week eight, steady. We see no diminution. If you look at Epworth and into the extension, we haven't presented those data yet.
Yep.
It stays level. It doesn't go away.
Okay.
Now, Epworth, to your question, Epworth is a very rudimentary instrument.
Mm-hmm.
But it is widely used. It's eight questions. You score from one to three of how likely you are to fall asleep in various settings: driving a car, watching a TV, things like that. But it's been used as a regulatory tool many times. But we also, we felt like it was a necessary prerequisite for running a registration-enabling study. But also, we were interested in more specific domains, like particularly cognition and fatigue. Because when you talk to patients with narcolepsy, what do you hear? NT1 patients, you hear about excessive daytime sleepiness, cataplexy, brain fog, and fatigue.
Yeah.
Fatigue different than sleepiness. We wanted to use instruments that hadn't been really used in these studies before, but to pick up these domains. That's what was so exciting about the Vibrance-1 data. That signal came through immediately. Much like the Epworth, first observation, week two, sustained, people go from being highly symptomatic to normal. Self-assessment as normal. We've run similar instruments in the NT2 study. We haven't presented them yet, but we expect to see a similar phenomenon of an effect on cognition and fatigue.
Yep.
In the NT2 patients as well. So I think Epworth is almost coin of the realm in a way.
Yep.
So you use it, but we want to augment it with more clinical information.
Sure. Sure. That makes sense. I wanted to toggle over to safety and tolerability in Vibrance-2. I mean, it essentially, it did confirm your safety hypothesis in the NT2 setting. So I guess the question here is, knowing what we know about AEs in Vibrance-2 and the fact that it did confirm your hypothesis here, how much higher could you push the dose? I mean, how much freedom do you think you have to aggressively push the dose? I know in the phase I, I think you went up to 50 milligrams. I'm not saying that that's going to happen. But talk generally, if you can, about that.
So first of all, let's get to the underlying premise of the question, which is critical.
Sure.
Because the premise is that we need to push the dose to drive MWT higher. I'm not sure that's true.
Yep.
For the reasons I said before. If you, what we saw in the NT2 population, there were clear non-responders on MWT. So that, if you're thinking about making an average MWT that includes a bunch of zeros in it, you're never going to, you see what I mean? You're never going to push that average of, as I said, because it doesn't necessarily represent the central tendency.
Mm-hmm.
Also, when you're normalizing 70% of your patients on ESS, you're in a highly effective dose range. Now, with that said, we've tested in the MAD studies up to 25 milligrams.
Right.
So we clearly feel like we could move to 25 milligrams without any problem.
Yeah. Okay. So let's talk a little bit about competitive considerations within the orexin category. So I think there seems to be an emerging sentiment among KOLs that, look, they're all good. But we know that there could be potential incremental points of differentiation or more. So how do you size up your NT1 data versus, say, Orexin agent or ORX750? And same question around NT2, not Orexin agent, but maybe ORX750. I mean, I believe there is room for multiple orexins. I think this market can accommodate that. However, there are certainly competitive considerations. So how do you compare and contrast your NT1 data and your NT2 data versus the competition?
I think your first observation is the right one. In a new therapeutic category that's making such a huge advance potentially for patients, there will be multiple drugs.
Yep.
And there should be multiple drugs. And that'll drive the size of the category and the utilization of the medicine. So let's start there.
Mm-hmm.
The only data sets you can compare are Takeda's and ours.
Yep.
We're the only ones who presented data.
Yep.
And I think we're big fans of the way Takeda has conducted their program. Because they've run large randomized studies, they've published their data, they follow the sequence that one would follow normally that the FDA has prescribed. And we're doing the same thing.
Yep.
And we presented our data on NT1 at World Sleep, and we presented the NT2 data without a scientific meeting, but in great visibility for you all so you could learn from it. So in those two data sets, the commonality between the two data sets is that orexin 2 receptor agonists are generally safe and well tolerated with mild to moderate side effects, many of which attenuate over time. And the quality, the efficacy, and the degree of efficacy is much greater than the previous drugs.
Yep.
So it's confirming that original hypothesis.
Yep.
And I think both we and Takeda are, whereas the efficacy, I think, was self-evident a couple of years ago even, now I think we're building a database of safety and tolerability that is well understood.
Yep.
Our drugs, Takeda's and our drugs, are different. Theirs is given twice a day at one dose for the NT1 population.
Mm-hmm.
We expect to come to market with a range of doses for NT1 and NT2. So a little bit more dosing flexibility and a little bit more ability to explore the dose-response curve.
Sure.
Anybody who comes behind us has got to replicate those data sets and see how they stack up.
Yep. Let's talk about the phase IIIs for NT1 and NT2. I'm not going to ask about specifics, but just to be clear, have you decided on the actual dosing for your phase IIIs for both NT1 and NT2? I know we're not going to get details, but have you actually made those decisions?
We have. I think our briefing book for the end of phase II meeting went in this week.
Okay.
So, yeah, we're good to go.
How many patients per arm in the phase IIIs for both NT1 and NT2?
I think FDA is going to recommend somewhere between 40 to 50 patients per arm.
Okay. In both NT1 and NT2. Okay.
Yeah.
Have you said, or can you say, within the NT2 study, how many different active dosing arms you're going to have?
Not yet.
Okay.
I think the reasonable expectation would be either three or four.
Okay. So I want to spend a little time on IH, which I'm excited about because there's a lot of white space here. There's only one drug approved in IH.
Yeah.
Let's talk about Vibrance-3. First, what do you think we can read into that study? Different population, but based on what we learned from Vibrance-2, do you think that orexins are going to behave in IH similarly to how they're behaving in NT2?
I do, and why do we believe that? We have some data from our 1b study in IH patients where at the same doses, we saw a similar response, but I think we just know so much more after the NT2 experiment. First principle, NT2 experiment, can you drive wakefulness and other measures in patients with orexin tone in their brain?
Yep.
Answer is yes.
Yep.
We talked about the variability of that and the flexibility of that diagnosis between IH and NT2. So I think we'll have a lot of the same patients in an IH study that we might have had in an NT2 study. So my belief is that, and again, I could be wrong, but my belief is that the efficacy, the circuitry in the brain that we're activating with these orexin 2 receptor agonists is profound. It drives wakefulness.
Mm-hmm.
I believe that signal will come through in the IH patients in the same way. Now, what did we learn in the NT2 study?
Yep.
We applied a phase III level statistical hierarchical analysis to a phase II study without any ability to estimate effect size.
Yep.
We did that because it's just best practice. Because then when you go to FDA later during the NDA, they can look at that study differently than if you had just.
Sure.
Done it without any pre-specified analysis plan. So we actually tend to look at the unadjusted P values as more important than the adjusted P. Because the adjustment comes from the hierarchical testing of various tiers that you establish a priori. So in the IH study, what I'm hoping to see is that we see just the signal of wakefulness come through the noise.
Mm-hmm.
Remember, in that study, we're not using MWT as a primary endpoint.
That's right.
We're using IHSS and ESS.
As co-primaries.
Yeah.
Yep.
No, I yes. No, I don't know. Well, I think ESS is the primary endpoint and IHSS is the key secondary.
Is the key secondary.
Yeah.
Got it. Okay.
So we won't have that same statistical alpha spend in the same way.
Got it.
But my point is, that's all sort of phase III stuff.
Sure.
What we're looking for is the signal. So across this range of doses, do we see that signal emerge? And our hunch is that it will.
Yep, so we got a minute left. Got to spend a little time on oxybates.
Good.
And I know you can't really talk much about it, but this is more of a question about just the overall role of oxybates in the world with orexin agonists. And wanted to get your thoughts on how you see the role of oxybate products evolving, both in NT1 and NT2, and we can talk about IH in a bit, with an environment where multiple orexins are available. I think I know how you're going to answer it because of your investment here, but love to hear your thoughts.
I went from being completely naive about oxybates when we got into this area of pharmacology.
Yep.
To becoming quite a supporter of them. And that's from talking to patients.
Yep.
Even through the context of the course of our phase II program and Takeda's phase III program, talking to patients who've been in the studies who are overwhelmed by the quality of the efficacy they're getting with the oxybates, with the orexins, still certain patients saying, "I still like that nighttime consolidation.
Yes.
And I think that will remain.
Yep.
We obviously think that there's a big opportunity to continue to grow the overall diagnosis in this market, and therefore the wedge that will be dedicated to oxybates, which is always going to be a minority, will continue to be robust as the overall diagnosis and treatment of this condition continues.
One question, in IH though, one question I hear from KOLs, or one issue I hear with KOLs is that sleep fragmentation in IH is not really an issue. Oxybates are really good for sleep fragmentation. The issue in IH is sleep inertia.
Inertia.
So in IH, I see a real role for the orexins. But I'm not so sure about the oxybates in IH going forward. What do you think?
I think that, again, I think some of these pharmacological tools will help refine the differential diagnosis.
Sure.
And so I don't think all IH patients are going to want or need oxybate. But I think some of them will. And you can see the growth from Jazz now.
Sure.
Is in IH. So it tells you something is there.
Yep. All right. I'll leave it there. We're out of time. Thanks for.
Thank you.
Thanks to everyone in the audience.