Everyone, and thank you for joining us on the third day of the TD Cowen 45th Annual Healthcare Conference. It is my pleasure to have with us today for a fireside chat the team from Alkermes. We have Chairman and CEO Rich Pops and EVP and COO Blair Jackson. Maybe if you guys want to just start off with a quick state of the business, obviously a lot of progress on the commercial front and through the pipeline, and maybe what should investors expect over the next 12 months, and then we can kind of dive into some of the specifics.
Thanks, Joe. I'll be brief. I think the gratifying thing is that the business is evolving exactly as we have intended it to evolve over the last several years. A couple of years ago, we decided that because what was happening both commercially and in the development pipeline, that focusing the company as a pure-play neuroscience company made a lot of sense because it would have two really distinctive elements.
One would be a profitable and growing commercial business focused on the neuroscience with addiction, serious mental illness, with drugs and schizophrenia and addiction. And then also this orexin pipeline that we were seeing scientifically was maturing really nicely. And our hunch was that that was going to mature into what would be a really exciting lead product in ALKS 2680, but also pipeline products behind that.
So as we roll now into 2025, we've simplified the business in a great way. The cost structure is in great shape. The commercial products are driving a billion-dollar top line, profitable company, and that funds the expansion of this orexin pipeline. So it's a non-dilutive self-funding way of expanding and elaborating this pipeline that we think has the potential to be one of the really hot new areas in neuroscience and in pharmaceuticals in general.
This orexin pathway is foundational to wakefulness. So it's called the master controller of wakefulness. And I think you all are beginning to see from us and others the data emerging that if you can develop small molecules that can mimic the action of the natural neuropeptide, you can drive significant wakefulness.
And the first demonstration of that will be in nature's knockout model, which is narcolepsy type 1, which is a deficiency of these neurons in this neurotransmitter. So in the first embodiment, it's replacing a deficient natural neurotransmitter and seeing whether we can recapitulate or approximate the natural sleep-wake cycle. But the pharmacology extends obviously well, well beyond that.
If you can do that safely and in a well-tolerated way, you start narcolepsy type 1, the penumbra extends to narcolepsy type 2, to idiopathic hypersomnia, and probably beyond that into other wakefulness or attention or vigilance indications where that type of stimulus of that circuitry in the brain could be useful. So we feel like we've been doing this for a long time. We've evolved this business in multiple ways in response to different environmental conditions. It started as a drug delivery company focused on delivering drugs into the brain.
Then we elaborated into a very, very advanced drug delivery company with medicines that we developed, big pharmaceutical companies that drove a significant top line business of royalties. We segued completely from that business now where the top line is all proprietary products that we invented ourselves, developed, and got FDA approval.
So we have a lot of experience and a lot of scars and a lot of, I think, real-world sensibility that we apply to the development program now. And I think you can see that in the way we're developing ALKS 2680. So maybe that is introduction, Joe. Actually, I'll say one other thing, which is the macro environment is actually central for what you guys do and for what we do too. And it's something we're very, very plugged into.
I'll just say I feel like Alkermes is extremely well-positioned in this complicated environment for the following reasons. We are developing drugs that are ones that we feel like payers won't have an opportunity to really control the tempo of their use. If you're developing new medicines that are disease-modifying to replace existing medicines or improve on what precedes them, it's very difficult for payers to have the upper hand in restricting access to them.
That's why we think this orexin profile portfolio is so important. We manufacture our drugs in the United States. We actually manufacture them in the state of Ohio, which is not irrelevant. We don't sell our drugs in Europe at lower prices. We don't for that reason, because of lower prices.
And I think that we've got a business now with its profitability driven by a diverse top line, the ability to self-fund, not have to rely on the vicissitudes of the equity markets to fund our program. We are not capital constrained. And I think that we've got the ability to develop medicines that can thrive, actually, in a price-sensitive market. So with that, I'll stop, Joe.
Perfect. And if anyone's obviously done their KOL checks on the mechanism, it looks like these can shape up to be pretty massive drugs. Maybe just to set the stage, how big of an opportunity is there in narcolepsy and idiopathic hypersomnia? Maybe what's the unmet need from a bedtime oxybate standpoint, those that aren't controlled on stimulants? Kind of where would these kind of fit in?
Blair, you start and I'll have some color.
Yeah, I mean, I think this is a really interesting marketplace because there's a surprising number of narcolepsy patients around the world and in the United States. And if you look at the current market right now, it's about a $2.7 billion market. But that's really only driven by about a quarter of the patients that are being treated. So about 16,000 people are receiving the oxybates, which are really the leading drug in the space right now.
The other 75% of the people are being managed by things like antidepressants and stimulants to try to deal with their symptoms. So it's a huge opportunity for drugs that can fundamentally change the disease profile. And what's really interesting is the diagnostic criteria between NT1 and NT2 can somewhat be blurry.
And then when you get into further elaboration of the disease into something like idiopathic hypersomnia, it gets even a little more nebulous. And I think what's super interesting about this space is despite the fact that there hasn't been drugs that have fundamentally changed the disease, you're still getting a lot of patients being managed by these physicians.
And one of the things that's striking is there's actually 200,000 NT1 and NT2 patients who have the disease, but only about 100,000 of them actually get the diagnosis because it's just not driven by the therapies that are there. So as this new class of compounds come into the space, there's opportunity to penetrate into the currently treated class, but also to start to shift that diagnostic criteria towards more utilization of medicines in this patient subtype.
Perfect. And across the NT1, NT2, and IH subsets that you've demonstrated efficacy in so far, you are almost maxing out that MWT efficacy signal. I guess what is your expectation for phase two? Do you need to max that out? Is there an ideal kind of level of benefit of wakefulness out of there?
Well, I think the most important point to realize is that this is a new therapeutic category. And I think it's incumbent upon us, and I think our program is doing this, to define the dose ranges and the degree of efficacy driven by those various doses against the backdrop of the variability of the disease. NT1 you almost model as a more kind of mono-dispersed phenotype. But NT2, IH, there's a broad amount of variability.
Our view is that you want to have a drug that can address all the variability in the patient population as well as the variability of individual patients' desires and needs. So in so doing, then you need a drug with a fairly wide therapeutic index, and you want to test it across all those doses in phase one and phase two. And that's exactly what we're doing.
We're actually building a profile of a medicine that we're building for the long term because we see, as Blair said, the differential diagnosis between the various disorders is blurry, and also different patients have different desires for their treatment outcomes, so what we're hoping is that we can build a molecule that has the tolerability profile, and we define it clinically in such a way that physicians will know how to use it across this entire spectrum.
Maybe related to that, we have heard from our KOL checks that it would be ideal to have a range of doses to kind of tailor the orexin need to the specific patient. I guess in the phase two, you're obviously studying several doses, but how will you be able to interrogate sort of that dosing flexibility structure? And what's the ideal setup to kind of enter a phase three? Would you take for multiple doses per indication?
I'm sorry, I really didn't answer that question. Which is typical. You're asking sort of what looks like good efficacy. So what the orexin class has shown in general right now, ourselves and others, is the ability to almost peg that MWT test as far out as you want. And right now, so the construct that people use in the clinic is this 40-minute wakefulness test.
It's an artificial setup, right? You put people in a dark room, you measure how long they stay awake. And if you have NT1, you're going to fall asleep in under five minutes. And if you take an oxybate or the current drugs, that five minutes might become 10 or 11, 12 minutes, not very much. And the orexins have shown the ability to drive that to the limit of the test itself.
I don't think you need to go all that way, and I think different patients may or may not want to go all that way. That's why the range of doses, that's what got me on this diatribe about the doses. You're going to want to have a range of doses because if you have dose-proportional responses, you'll be able to dial in duration with dose as well as the magnitude of effect.
So there should be no, I think, the first principles. You want to have a significant space between these drugs and the drugs that precede them, i.e., modafinil, oxybates. If they're well tolerated and they're driving significant wakefulness beyond that, that's table stakes. That's going to get you in the game, and then I think it's the qualities of the drug from a commercial perspective that separate between the various competitors.
To your question on the clinical trial, I think as we look at the phase two program, one of the things is we have three doses in the NT1 population of four, six, and eight, and then we have three different doses in 10, 14, and 18 in NT2 and IH, so it allows us to interrogate a wide range of doses across the various
treatment profiles, and our goal in the phase two is really to elaborate the range of dosing both on the low end and on the high end to characterize what best works for this treatment population because we expect to see a little more variability in NT2 and IH than in perhaps NT1, just given the nature of the disease, so then as we move into phase three, we'll set in on the doses that we move forward.
It might be three in each cohort, or it might be maybe two. We'll see what the best range is, and that's what we'll move forward with. And can you talk a little bit about what you see as the ideal length of a clinical study for narcolepsy and idiopathic hypersomnia? The three kind of leaders in the space are maybe taking slightly different approaches. And when you look at the Takeda phase three, you look at the Centessa phase two, and then your program, what is sort of the ideal?
The glib answer is the best program is what FDA tells you to do because they're the ones who are going to be ultimately reviewing the application. And I think that's why I think you see a couple of us have very identical phase two programs. So back to the point about an early-stage development in a new therapeutic category, what does FDA want to see?
They'd like to see randomized classic parallel design studies with multiple doses being interrogated for a multi-week period. So what you see from us and from Takeda has been parallel design studies, multi-week, double-blind, where patients you don't move around on the dose. And so in our phase two, it's designed with three doses, six-week double-blind period where people stay in their dose for that six-week period.
At the end of the six-week period, patients will all go into an open label period where they center on the middle dose for two weeks, and then they get to choose their dose. That's never been done before because no one else has had the flexibility on dosing, so I've said before, I think our 160 patients that we'll have data on at the end of this year in the second half from NT1, NT2 for the whole field will be a
huge increment of data because it's the first time we're going to see a range of doses across NT1, NT2 for that extended period of time, six weeks plus seven weeks, looking at efficacy, its profile over time, side effect profile over time, and the patient-reported outcomes associated with being on these therapies for this extended period of time. I think it's going to be really definitive in many ways for our understanding of the medical and commercial potential of these medicines.
I guess a question on how do you choose the dose range that you want to do. I think one of the things that we think about when we look at the clinical trials. Take the NT1 patient population, for example. When they come into our trial, they need to wash out of their existing medications. And that takes a period of time. It's a few weeks.
They feel awful after that. And then they get randomized into the trial. And on day one of the trial, they know whether they're in an active or whether they're in the placebo because just the response is so strong with the orexin programs. So with an NT1 patient group, you want the trial to be as short as possible. It's not really great for them over that period of time.
And so if you look at our NT1 portion of our program, we have a shorter period of time in life versus the NT2 where it's a little longer because the NT2 patients are maybe a little less severe when it comes to that. So as we look at the programs as we move into phase 3, to Rich's point, it's going to come down to what does the FDA want to see.
And with Takeda leading the way from a time perspective, they asked for kind of a three-month period of time to test those patients. But I think as you talk to them about the NT1, asking them whether or not they want to see that level of efficacy or that level of time in those patients, I think that's a discussion to have with the FDA. If there's opportunities to shorten that, we would definitely try to do that.
Can you talk a little bit about the ideal dosing frequency and PK of these drugs? Obviously, for programs that are BID, they say it mimics the natural sleep-wake cycle and kind of pattern of orexin exposure during the day, or once a day. What do you think related to that? What are the pros and cons?
Just to be clear, I think people with BID medicines aren't saying that that's what mimics the natural sleep-wake cycle. They're saying that by dosing it that way, they can more closely approximate the natural sleep-wake cycle, which is not bimodal. It's a natural diurnal pattern of it. So from our perspective, from the design intention at the outset, it was to mimic the natural waveform by having elevated levels above threshold concentrations diminishing into the evening so patients could go at night.
That'd be a classic QD or once daily dosing. The issue with twice daily or three times daily or four times daily is not necessarily the fact that you have to take them multiple times a day, but by definition, it implies that there's a diminution of activity that needs to be augmented.
And the trouble with something like this is that you can't dose too late in the day because then you're going to worry about insomnia. So it's better if the natural PK of the molecule is consistent with that sleep-wake cycle. So far, that's what we've seen in our study.
And that was. It's not by accident. Our background, we think about PK. We have so much background in developing controlled-release medicines for injection or for oral that the original design intention was not just potency and selectivity and oral availability. It's PK because we think that's the way that we can differentiate.
Maybe let's talk a little bit about tolerability. What is your expectation for maybe what you are going to see from a tolerability standpoint in the phase 2? We've obviously seen some insomnia, which maybe we've already touched on. Obviously, there's the urinary urgency, AEs, and then also some visual disturbances in the initial experience. Which of these are on target? And can you kind of classify them in terms of ones that you're maybe concerned about or ones that maybe patients can kind of adjust to over time?
I think we all benefited from the Takeda phase two data, the eight-week data. It was really the first data set that showed the profile of continuous administration of these in the outpatient setting, and I think their data and our data taken together, it's remarkable how well tolerated this class of medicines is. The effects tend to be mild to moderate, transient, and you've highlighted the hallmark on-target side effects that you see are insomnia and what we call polyuria, others might call urinary urgency, just that you feel this desire to pee.
And it doesn't correlate actually with more urinary output. It's just a sensation, and that appears to be dose-dependent and also appears to be transient. The previous data set indicated that, like the insomnia, it largely presents early in the dosing, first week or so, and then diminishes without a concomitant diminishment in the efficacy.
We expect to see that in our study as well. But again, that theory plus data will answer the question. There's been a lot of talk about visual disturbances. And a lot of that has to do, I think, with the ominous nature of the terminology. In our phase 1b study, we saw one case of a visual disturbance in patients with NT2. That visual disturbance was noticing that the lights looked brighter when they were turned on. That's graded as a mild AE.
Mild means it's noted by the patient, but doesn't affect them. In the IH patients, we had a similar phenomenon where instead of light sensitivity, it was blurriness of vision. Again, mild and transient. At the 25-milligram dose, which we're not testing in phase 2, with NT1, we don't know whether it's on target or what, but that's what this data set will answer.
In contrast to the 1B, now that we've identified that and any other side effect, you look for it in phase two. So for example, now in phase two, all patients will have a baseline ophthalmic exam upon randomization. If you get a visual AE of any degree, we'll work it up from a licensed ophthalmologist workup because then, is it pupil dilation? Is it something we don't know now?
But our expectation is that if it's on target, it should be consistent with what we've seen before, which is mild to moderate and transient. But we'll know with greater certainty, and that's the virtue of this idea of testing multiple doses for an extended period of time. We'll be able to see if there's dose response and the transient nature and dose response nature of that and any other AE.
Can you talk a little bit maybe about just the overall enrollment process or progress in the phase for the NT1s and the NT2s? Obviously, the NT2 kicked off later, but it seems like due to less competition in NT1, maybe those data will come around the same time. How are you thinking about that?
Yeah, they'll both finish up in the second half. Any of you who've developed drugs, if you look at enrollment curves, projected enrollment curves, they're very nonlinear because the beginning of it, it's all about site initiation.
And we're initiating sites in Europe and the U.S. and Australia. And some of the bureaucratic processes are massive. It takes months and months to get some of these sites up. So you say, why do you do it? By the time they're up and running, the study will be almost over. You do it for phase three. So our curve looks just like that. It's very nonlinear. Coming out of the holidays, all of a sudden, that's why I think on our Q1 call, we should be able to give you a great degree of precision about where we are. But right now, it's all still settling in that second half.
Can you talk about the progress on the phase two for idiopathic hypersomnia? What are the different sort of endpoints you look at in idiopathic hypersomnia versus narcolepsy and kind of how is the progress on that study?
It's an interesting disease. Remember, in the beginning, we thought we might not develop this drug for idiopathic hypersomnia, but the data from the phase 1b was so clear, and we went and met with the KOLs and the patient groups, and there's just a huge need for new medicines for idiopathic hypersomnia. Despite the fact that it's characterized by excessive daytime sleepiness, it is a very different disease.
Patients sleep a lot. They just don't feel awake. They have a lot of sleep inertia. They have a lot of just different presentation, and you can confirm that differential diagnosis in a sleep lab, but what's interesting about it is its adjacency to patients without these diseases. It's a more intact orexin system in the brain, and so the fact that we're seeing wakefulness changes in those patients despite probably not an obvious orexin deficit is interesting.
So it's actually regulated by a different part of FDA. Narcolepsy is regulated by the Division of Psychiatry, and IH is the Division of Neurology. So we're actually filing a separate IND with Neurology. That's why we haven't started the studies yet. That's all well underway. So we expect to start the IH study in the next few weeks.
Its endpoint, you don't use the Maintenance Wakefulness Test. We did in the one B to show the effect, but that's not the coin of the realm. For idiopathic hypersomnia, you use certain patient-reported scales, the Idiopathic Hypersomnia Severity Scale, IHSS, and I forget what the primary is. Epworth is the primary.
And Blair, you kind of mentioned this at the beginning, that there's almost a range between NT2 and IH. And what do we know about the size of the IH population right now? Are these patients able to be identified easily? And are you easily to be able to tell who is an NT2 and who is an IH patient for trial enrollment?
You know, it's a great question, and the answer is no, you can't really tell them very easily, and in fact, the easiest, the brightest line is between NT1 and NT2 because it's sort of, I mean, you could do it if you were to look at orexin tone and you're able to do a spinal tap, but you don't do that in the United States. So what it really is, is the presence or absence of cataplexy that kind of gets you into that.
As you're looking at NT2 and IH, as to Rich's point, the disease is a little different, but it's almost IH is a diagnosis by exclusion, and so really, it's people who are struggling with wakefulness who didn't really fit into the previous two categories that fill that bucket.
I think as we think of developing these drugs moving forward and providing efficacious drugs that can now be used, I think you're going to see that population start to expand. Right now, what we know is that there's 80,000 treated narcolepsy Type 1 and Type 2 patients in the United States. That fits what I talked about earlier. There's about 40,000 idiopathic hypersomnia patients who are being treated right now. How big that total population is, though, is anyone's best guess, but it could be quite large.
You do have additional orexin candidates that are entering the clinic this year. Can you talk a little bit about the differentiation of those versus 2680 and kind of where else can you take this in terms of a therapeutic indication?
I just want to make one other analogy that we think is relevant. And that's what we've seen in the treatment of tardive dyskinesia, which really wasn't a diagnosis until you have drugs. And so there's two good drugs there now. And because physicians have an instrument to treat this condition, the diagnosis rates go up.
I see a direct analogy into these diseases of hypersomnolence. So our team just got back from Italy. I was talking to them yesterday, for example. IH isn't a thing in Italy. It doesn't exist. So it doesn't get diagnosed until you have a drug for IH. And then so it's the fluid nature of some of these poorly treated, less prevalent diseases.
And then, yeah, so then maybe some of the follow-on compounds, how are they different from 2680? And where can you take that from an indication standpoint?
So the sine qua non, to play in any manifestation of these, you need highly selective, highly potent, orally bioavailable blood-brain barrier penetrating compounds. So that limits the medicinal chemistry space at the outset to structures that can satisfy those goals. Between the various indications, we have not been specific and won't be specific about the particular differences of them.
But you can imagine how you might have PK differences or you might have absolute dose differences because what you really want to avoid is substitutability between various indications. That also could involve additional pharmacology. So look what's happened in the GLP-1 space, right? They start with pure GLP-1 drugs.
Then you add GIP, then you add glucagon. There's ways of dimensionalizing the pharmacology based on other known mechanisms that could be complementary to the indication that you're looking at. So we've been doing this for a couple of years. We have some very good ideas on it, and we're not super keen on sharing them with others at this point.
Maybe on the overall business, you did announce the EBITDA guidance of 215-245 for the year on a GAAP basis. Can you kind of go through what goes into that and maybe some of the changes that we should have been forecasting in 2024 versus what we're going to see in 2025?
Sure. I mean, we're coming off a big period of transition for the company, as Rich mentioned in his opening statement. And I think we really kind of completed a lot of that activity last year. So as we look at moving into 2025, we see a couple of things. First and foremost, we have a highly profitable commercial organization, three products.
We expect to see consistent growth of those products over into 2025, Lybalvi being the fastest growing of the assets playing in that oral antipsychotic space. And we have expansion of the sales force and continued DTC activity with that drug moving forward. I think when it comes to the cost structure of the company, we're going to maintain a strong discipline on our OpEx spend. We're allowing for expansion, both in sales and marketing for the sales force expansion in our psych franchise.
We're adding about another 25% of the reps, 33% of the reps in the space, and we're allowing for expansion to increase spend in the orexin program to allow for movement into the idiopathic hypersomnia indication to get ready for phase 3 and to move into some of those expansion programs with the new molecules.
That's really all that we're moving up on the spend side. I think the major difference between last year and this year has to do with that transition I mentioned earlier as it pertains to our manufacturing and royalty revenues. We did two things last year. One, we sold the Athlone facility to Novo Nordisk. So we're transitioning out of some of that manufacturing and royalty revenue, and then number two, we have an expiry of one of our long-term contracts with J&J around Invega Sustenna.
That has a net of about a $200 million impact on the manufacturing and royalty revenue that we won't see moving forward. But we've been planning for our business to transition for that for a number of years, and that's what we're going to see going forward. We are relying on our proprietary products for growth.
Perfect. And it does seem like the expansion of the neuropsychiatry force is going to put you in a good position to maintain and grow your share in schizophrenia and bipolar, even though there are new market entrants kind of entering here. Can you talk about, is there an ideal patient for Lybalvi? Is there an ideal patient for Cobenfy? How do you see the market kind of evolving there?
Lybalvi's place in the market is driven by efficacy. That's its raison d'être, and it has the benefit of the vast clinical experience physicians have with olanzapine. And now with Lybalvi, we've recently shown four-year data showing the maintenance of the weight and this ongoing efficacy.
So I think that, unfortunately, for patients, this is a market that's characterized by frequent switches. So the average length of therapy on an oral antipsychotic is six months. So when we see something like a new entrant come into the market, from our perspective, operationally, it's much more about maintaining share of voice than patients don't get displaced off medicines that they're on. That happens naturally through the course of the disease. So the sales force expansion benefits both Aristada and Lybalvi and just keeps Alkermes' share of voice at competitive levels.
Perfect. And with that, we are at time. So congrats on all the great progress and more to come this year.