Good afternoon, and thank you for joining Guggenheim's second SMID Cap conference. I am Debjit, and joining me from Centessa is Mario Accardi, the President of the Orexin Program. Did I get that right?
You did, yes, Debjit. Hi. Great to be here with you.
It's been a very transformative year for Centessa, at least 2024. It's become more morphed into an orexin story. I'll let you lead off with a brief introduction, and where do you think you're going from here?
Yeah, thank you, Debjit. It's been a transformative year for Centessa 2024. Incredibly exciting. We've taken ORX750 through a phase I study and demonstrated, quite frankly, and very clearly, best-in-class profile with ORX750 in the, I would say, highest bar study, which is the acutely sleep-deprived healthy volunteer study, hitting 38 minutes on the MWT that, as you know, is the key registration endpoint used for narcolepsy and these indications, and 23 minutes placebo-adjusted, and quite frankly, with a safety tolerability profile that I think speaks to itself. W e've been extremely pleased with the momentum that we've generated and the profile that we've shown with 750. We also, in 2024, back in November, opened up and started our phase II study in narcolepsy type 1, narcolepsy type 2, and idiopathic hypersomnia.
It really helps set the context and the scene around 2025 and what to expect, which, as you know, is data across three indications: NT1, NT2, and IH. We have also data coming up with ORX142, our follow-up molecule in sleep-deprived healthy volunteers. W e're really laying the ground here at Centessa for the creation of a multi-asset orexin agonist franchise that has the potential for us to really achieve a best-in-class profile across NT1, NT2, and idiopathic hypersomnia, get us potentially first-in-class in NT2 and IH, and at the same time have follow-up molecules, additional orexin agonists, highly potent, highly selective molecules that will enable us to enter these very large indications that in the neurodegenerative psychiatric and neurological conditions.
A s you know, we've also recently hired Steve Kanes, our Chief Medical Officer, that we brought on with his incredible amount of experience in the CNS space, was former CMO of Sage. So we've really helped to cement things, and incredibly excited about what's coming next in 2025.
Thanks for the intro. G iven that these are small molecules, how comfortable are you and your peers on the safety profile to date?
First of all, with ORX750, we've designed a molecule, again, thanks to this unique structural biology capabilities that is really at the core of the program. But back when I started the company in early 2019, Orexia Therapeutics then got acquired by Centessa, the fundamental goal was to use a structural biology platform to help design these exquisitely potent and selective small molecules. W e did that thanks to a really deep understanding of the orthosteric binding site that has allowed ORX750 and a number of other clinical candidates to be designed. W e feel extremely confident in terms of where we are. We have a molecule that is very potent and is translating to very low doses in the clinic, right, as we've seen in the acutely sleep-deprived healthy volunteer study. We have a molecule that is extremely selective, 10,000-fold more selective at OX2R compared to OX1R.
We're really seeing no relevant pharmacology across any other GPCRs, right? T his really builds confidence in terms of how we've designed this small molecule and the data that we're seeing in the clinic, Debjit. We're seeing clinical data that is essentially very similar to placebo across all doses tested if you look at our safety tolerability table. Q uite frankly, the way in which we've designed 750 is essentially trying to be as close as possible to how the native peptide interacts with the binding with a receptor by essentially being almost equipotent to the native peptide and by activating the orexin-2 receptor with so-called unbiased agonism. N either skews towards beta-arrestin or G-protein, having something that mimics what nature has designed as closely as possible.
A lot is to be said also around the way in which we've designed the molecule in light of the PK profile, right? We have a relatively flat Cmax to AUC while enabling single-dose , QD dosing. We did hypothesize that many, a lot of these, maybe a lot of these on-target AEs that have been frequently reported by other orexin agonists could indeed be Cmax mediated. It was a very nice surprise for us to see such an incredible tolerability profile, which is significantly differentiated from other molecules.
With the recent update in January, even at the five -milligram dose, you guys did not capture any insomnia. Do you sort of have a theory why you're not seeing that? Can you keep that profile at the, let's say, one and two-milligram starting dose in the phase II study in narcolepsy patients?
Yeah, I would mention two things there. The first one is that I think we've hit the sweet spot in terms of duration of action with 750. We are at 38 minutes on the MWT, which means that we have significant efficacy and normative wakefulness clearly across all time points of the maintenance wakefulness test, right? That's roughly eight hours. W ith a single dose, right? We're keeping healthy volunteers up at night, even through to the earlier parts of the morning. I t really bodes very well for what we're going to be looking at in our patient studies. As you know, healthy volunteer data translates extremely well to patients, both from a PKPD efficacy perspective, right? The acutely sleep-deprived study is a model of non-orexin loss pharmacodynamics, PKPD.
At the same time, if you look at the data that's been generated with clinical biology, by clinical precedent, you can really draw a line on this essentially wakefulness threshold, which is the minimum brain exposures that you really want to reach in order to drive efficacy and promote wakefulness. Putting everything together, right, we feel very, very confident around those selected for our phase I studies that, as you know, is one milligram in NT1 and two milligrams in NT2 and idiopathic hypersomnia.
A lot has been sort of discussed about once daily versus the flexibility or apparent flexibility of a BID dosing schedule. Where do you guys stand on that? Do you think you can mimic the circadian rhythm with once-daily dosing?
The goal, Debjit, is to have a PK profile that allows you to be above this wakefulness threshold throughout the duration of the day and essentially promote wakefulness and reduce cataplexy, of course, in NT1, and then for this to come down into the later parts of the evening and allow patients to go back to sleep, right? U ltimately, by having a low dose and a single dose, it just gives us tremendous flexibility for us to fine-tune that duration of action, right? The fact that we don't need multiple doses, right, to reach almost the ceiling of the MWT, which is 40 minutes in an acutely sleep-deprived study, right, gives us, again, tremendous flexibility to dose up or down to further adjust that duration of action.
A BID drug, if you need to achieve the same with a BID drug, that essentially means that your drug is cleared too quickly. I need to compensate for that and add multiple doses to be above that wakefulness threshold during the day. We can achieve that with a single drug, with a single dose.
The phase II design, it's kind of unique. What led you to that design? And given that patients are not going to be on placebo for more than two weeks, how is enrollment coming along?
The ultimate goal of a phase II study is to help you pick the right dose for registration studies. Working backwards from that, we asked ourselves, what is the best possible study design that would allow us to pick the optimal dose that will give us optimal efficacy and tolerability profile? Y ou're right, it's very different compared to a conventional phase II study where you're pre-defining the dose levels ahead of time, and then you're hoping for the best that by the end of that study, you will have hit all the right dose levels that you need for the registration studies. Here, we have split the study in different dose cohorts. We first enroll in one dose cohort, and we then do an unblinding at that specific dose, which, by the way, can be different according to indication, NT1, NT2, and IH.
By the way, every indication can progress at their own pace depending on enrollment speed, enrollment rate. Once we're doing the unblinding for that specific cohort, we then peg safety, tolerability, PK, and a large number of efficacy endpoints well beyond the MWT and ESS. We're also looking at cognition, attention. We're looking at patient-reported outcomes. We're also looking at PSG, right, the overnight nighttime sleep. We put all of that together, and that forms the basis for helping us select the dose for the next cohort, which can be up, which we could do is escalate up or dose de-escalate. Ultimately, with the goal of helping us find that perfect dose for the registration studies.
While neither, I mean, you're still doing the phase II study, but your competitor, Takeda, they have not seen any diminution effect in NT1 patients as yet. Do you think down the road these patients get used to the dose they are getting and then they might need to be escalated higher? In which case, how do you stack up?
Yeah, you're right. The data that we've seen in the clinic thus far from Takeda, I think it's, first of all, it's great data. It's really great to see sustained efficacy over time, right? I believe all the way up to six months they've shown in their data set with TAK-861. R eally, really great data on a completely transformational treatment for narcolepsy type 1 individuals. What I would say is that we expect to reach steady state in terms of effect size somewhere within the four weeks, right? That's what's been shown through clinical precedent, which is why our study design includes four weeks of continuous dosing. U ltimately, that is really going to help pave the way for the dose selection for registration studies. There is no reason to believe why longer-term dosing would cause diminishing efficacy.
We've seen no cases of tachyphylaxis internally, no receptor internalization, have since sustained efficacy in preclinical studies. And that, I think, question has very much been, I would say, answered by Takeda.
Given your starting dose is one milligram in NT1, you're normally expected to be about threefold higher in NT2 and IH patients?
Roughly, I would say 2 x to 3x, right? NT1, we don't really know why, but they're just more sensitive to the mechanism. Probably we hypothesize maybe because of some form of orexin 2 overexpression. And that's why you do need slightly higher doses in NT2 and IH, which, by the way, are very much the same doses that you need to keep healthy volunteers, sleep-deprived healthy volunteers awake, which is why our phase I study was so instrumental in de-risking 750 and de-risking that therapeutic window across indications beyond NT1. The goal for us at Centessa with 750 was really to have a single molecule that could help us move into NT1, NT2, idiopathic hypersomnia, and to be on any indication where there isn't necessarily a loss of orexin tone.
Given that you're starting at two milligrams in the NT2 and IH subsets, you don't expect to go meaningfully above three. Is that a right way to think about it?
We're starting off with clearly therapeutic doses, right? I think we've proven that in our phase I study based on mouse and NHP data. We went from an open IND to efficacy data in three months and a half. With this mechanism, you can really home in on the PKPD and really fundamentally understand what is the wakefulness threshold that you need to be above in order to keep patients awake. A lot of the learnings that we've used for the phase I study, we're definitely bringing to our phase II study. We're starting off with therapeutic doses. The goal will be to, of course, generate data from multiple doses, right, ahead of then moving towards registration studies.
Is the goal to maximize the MWT in this subset or get to a normative or wakefulness kind of situation?
We have a very clear TPP internally, Debjit, that we want to hit with these studies, and MWT is definitely an important metric, as is the more subjective endpoints like the ESS. T o my point earlier, we are also looking at other exploratory endpoints that go well beyond just simple wake promotion, and I think we'll be looking at everything in a holistic manner to ultimately guide us into that optimal dose for patients. T he fact that we're dosing at such low dose levels, we've cleared four milligram in the MAD. It's giving us 38 minutes. The five -milligram is giving us 38 minutes on the MWT, and that has comparable exposures to the four milligram MAD because of some slight accumulation of our ORX750, right around 20% or so. We feel very, very confident around meeting this target product profile.
Moving forward, then eventually, I think all of these drugs at some point will most likely be titrated or you may need more than one dose in registration studies to really help you fine-tune that duration of action that is probably going to be different from patient to patient. A lot of these stimulants, a lot of the sodium oxybates are titrated as well, as you know.
From a cadence of data readout from the phase II study, let's say the NT1 cohort completes first or whatever, IH completes first, do you release that subset and choose the data so you can sort of start prepping for the registration study or everything needs to come together?
It's a great question, and I wish I had an answer for you right here, but I don't think we've completely decided on this yet. I think we're going to be unblinding the cohort and seeing the data in order to be able to then decide the dose for the next cohort, which can be up or down, and I think we have definitely the option to release data on a cohort-by-cohort basis, or we have the option to wait until all cohorts are enrolled or maybe wait for all indications to enroll. We have full flexibility to do that. It'll definitely depend on a lot of things. I think we need to align with IR, Kristen, on this, but we definitely retain full flexibility.
Has any cohort been unblinded yet?
Pardon?
Have you unblinded any cohorts yet?
They're very much underway, but I wouldn't say that we've unblinded just yet.
Okay. Moving on to the platform.
Yeah.
Where do you see 142 or what's the optimal path forward for 142?
I'm incredibly excited about the pipeline here. I think we started off as orexin agonists are going to be a functional cure for NT1. I think quickly realized that these have the potential to go across any other indication where there is an excessive daytime sleepiness component. NT2, idiopathic hypersomnia, which are north of a $7.5 billion market alone. T here is a number of other indications outside of the rare hypersomnias where excessive daytime sleepiness is a huge part of the disease. Parkinson's disease, for example, 90% of Parkinson's disease patients, mild to moderate, experience excessive daytime sleepiness, right? EDS, another key symptom, major depressive disorder. Residual EDS in OSA, right, another key symptom of the disease. L et's also not forget about fatigue that, as you know, is closely intertwined to wakefulness. Fatigue and multiple sclerosis stimulants have been used there for a number of years.
What is really great to see is a lot of these indications. First of all, they have a partial orexin loss themselves, right? T here's an actual causal link, mechanistic link of orexin biology to the disease. A t the same time, stimulants and wake-promoting agents have been used across these other disorders for many, many years, yet with subpar efficacy. W ith an orexin agonist, which is again why it's so important to have de-risked 750 in the acutely sleep-deprived subjects, and we'll aim to do the same with our ORX142 this year, is to de-risk that early on and provide us, essentially then enable us to move into these large indications, right? And by having multiple molecules, it really enables us to have completely different commercial dynamics at play. They're distinct chemical entities, all very potent and selective molecules.
G iven the broader neuropsych indications, first of all, how big are these studies going to be, even in a phase II setting? Are you thinking a basket study across all these indications you mentioned, or it's going to be a separate Parkinson's study or a separate, just pick a neuropsych indication? Is that individual phase II studies for each?
We have, as you know, we have a cash runway went into 2027 that allows us to prosecute multiple clinical studies with our multiple assets, so that's definitely part of the plan, and I think there's a lot of clinical precedent there. There's studies in EDS and major depressive disorders, studies on EDS and Parkinson's disease, studies looking at cognition in dementia with Lewy bodies, for example, so our aim will be fundamentally to help generate the signal-seeking studies in neurodegenerative and psychiatric and neurological conditions, right, to really de-risk these indications, and we're going into indications where we already think there's going to be high POS because of the excessive daytime sleepiness component, but there are definitely studies that can be done in a relatively short period of time and really help answer the key questions early on.
Y ou're not ruling out partnering the larger indications and keeping the core narcolepsy within Centessa?
We haven't said anything about partnering, but this is definitely something that we can build ourselves. I mean, the rare hypersomnias, as you know, there've been a number of companies out there that have built very successful stories on the rare hypersomnias. T hat can definitely be achieved with, I think, a relatively low number of sales efforts, sales force. O n the larger indications, again, we retain full flexibility there. We're going to prosecute a number of clinical studies. Steve Kanes, our CMO, will help us execute on these studies. W e have the opportunity to build something transformational here in a mechanism that is very unique in neuroscience, very, very unique, because we're going after a key symptom initially that has already been essentially de-risked in healthy volunteers and, of course, the rare hypersomnias.
But orexin agonists also have the potential to treat cognition, attention, and we've also seen orexin agonists, at least in the preclinical level, helping with mood symptoms because of the serotonin pathway. So it's something that we're incredibly vested in and is part of our plans here at Centessa.
Outside of the medicine by itself, what sort of differentiated Centessa over the last year was the speed at which you went through your healthy volunteer study and moved into phase II. G iven that, if everything goes well with the phase II -A program right now, how quickly can you get into the registration segment with 750?
We're going to have data, aiming to have data in NT1, NT2, and IH across all three indications this year. T his will pave the way for the end of phase II meeting, and we're going to have everything we need, and the aim is to have everything we need to move into registration studies. The great thing, Debjit, about our study design is that we can progress each indication decoupled from the other. I f for whatever reason one indication is where to progress a lot quicker than the other because of enrollment rate, we could move forward with an indication before the others. And right now, the plan very much is to be first in class in NT2 and IH and best in class across all three.
Ideally, given that Takeda is already well underway in the phase III study in NT1, would you prefer to be first in the clinic in NT2 and IH?
That's definitely the plan, Debjit. W e're coming, we're going to come to market hopefully with a single molecule to go after NT1, NT2, and IH, which has some significant advantages not only from the payer perspective, but also from the prescriber perspective. These are the same sleep docs that treat all three indications. Sometimes there is a lot of overlap even in the diagnosis, especially between IH and narcolepsy type 2. O ur goal is definitely to be first in class.
Awesome. I think we have run the clock, and for those who don't know, it's Mario's birthday. Happy birthday.
Thank you, Debjit. Thank you.