Harmony Biosciences Holdings, Inc. (HRMY) Earnings Call Transcript & Summary

Good morning, everyone, again. Welcome to the second session with Harmony Bioscience. It's my pleasure to have Jeff Dayno, who's the Chief Medical Officer of Harmony. He has significant experience in leadership roles in multiple companies as well as clinical and academic medicine as a principal investigator. We are lucky to have him today in the session. Jeff, thank you so much. Perhaps if we could maybe just kick off with some questions, if that's okay with you.

Yes, that sounds great, Ami. So Good morning. It's a pleasure to be here with you this morning. And on behalf of the whole team at Harmony, thank you very much for the invitation and opportunity to participate in Needham's first annual Neuroscience forum. It's great to be here.

Okay. Great. And by the way, just for the listeners, if you have any questions, be free to send it over in the chat to me, and I will ask Jeff those questions. So let's start with WAKIX in idiopathic hypersomnia. The company recently announced that they're moving forward with developing pitolisant and IH. Can you talk about what are the IH patients treated with today? And help us understand the differences between IH and narcolepsy that we should keep in mind as we think about WAKIX for IH?

Yes. Sure, Ami. So patients with idiopathic hypersomnia or IH have been treated with medications commonly used to treat narcolepsy, those being traditional stimulants and wake-promoting agents. And now we have the first FDA-approved treatment for IH after the approval of Xywav last August for the treatment of IH in adults. So one of the main differences between IH and narcolepsy is the long sleep time in some patients with IH. And that also the sleep and the naps in patients with IH are not as refreshing or restorative as in patients with narcolepsy. So that also leads to another key difference between IH and narcolepsy, and that's the symptom of what's called sleep inertia or difficulty awakening from sleep. And it's also sometimes referred to as sleep drunkenness. So IH is more a primary problem of wakefulness rather than disturbed nighttime sleep, which is one of the other key symptoms in patients with narcolepsy. So this is why we think that the mechanism of action of WAKIX or pitolisant makes sense to treat patients with IH because as a histamine 3 receptor, antagonist inverse agonist, that works to increase histaminergic transmission in the brain, and histamine is a major wake-promoting neurotransmitter. It drives wakefulness during the day after once-daily administration in the morning.

Understood. Okay. We've received feedback from physicians that the symptomology of idiopathic hypersomnia seems to be a lot more heterogeneous than what they see with narcolepsy. They almost call it sort of like a kitchen sink indication. Can you help us understand what that means from the perspective of drug development?

Yes, I think that -- so I think that IH being a diagnosis of exclusion, I think that's what you're hearing from clinicians with regards to the symptomatology is more varied and variable. And it's really as opposed to narcolepsy, a diagnosis of exclusion, you're ruling things out. I think the other thing you're hearing is, historically in IH, there were those that were sort of long sleepers, at long sleep time and those without. So there is variability across the phenotype of IH, if you will, but when it comes to sort of drug development, in our Phase III trial, we will be taking all comers in terms of patients with both long sleep time and short sleep time and anyone who qualifies through an ICSD III diagnosis of idiopathic hypersomnia.

Got it. And help us understand what are these patients being treated with today? Is it wake-promoting agents stimulants sort of the typical things that other patients with sleep disorders get treated with?

Yes. I think it's the typical things, as you said, stimulants, wake-promoting agents. And as I mentioned, Xywav was approved for the treatment of IH adults last August, but IH as another central disorder of hypersomnolence, I think that's what clinicians have looked to those other traditional medications for narcolepsy. And the other thing Ami, is there's also the thinking and that's important also when it comes to drug development, the IH exists along a spectrum. There's sort of the spectrum theory of the central disorders of hypersomnolence. You have NT1 at one end, the narcolepsy type 2 and then idiopathic hypersomnia. So those medications used to treat narcolepsy have what HCP have been using historically for IH patients.

Understood. Have you decided on the dose that you would use for the IH study?

So for the IH study, we're actually -- actually put some slides together, and I can show the trial design as we advance this. Yes. So in terms of the dose -- so this being what you see here is the trial design for our Phase III registrational trial, idiopathic hypersomnia. And what you see, it is an open label -- it's a controlled randomized withdrawal design. So therefore, patients will be enrolled into an open-label dose optimization phase. And in that phase, they'll be dosed in a manner similar to the current label for narcolepsy in terms of a 3-week titration period, starting at 8.9 milligrams the first week, then up to 17.8 milligrams a second week to a maximum dose of 35.6 milligrams very similar to the dosing paradigm for narcolepsy. Then investigators will manage patients through their optimal dose based on the balance between efficacy and tolerability during the open-label dose optimization period. And then treatment responders will proceed into the randomized withdrawal phase in a one-to-one ratio that you see there between pitolisant and placebo at the optimal dose they were on during the open-label phase. So it is not a fixed dose study in a typical prospective RCT design.

Understood. Okay. Can you talk about sort of the variability in these patients? And what percent of patients do you think are likely to respond? Or as you have designed this trial, what have -- what assumption is to be made with regards to responders?

Yes. So we really sort of the assumptions on the thinking was built off of the narcolepsy data, again, as an adjacency and what we saw in the narcolepsy development program in the range of 66% to 75% of patients responding. So we're targeting about 200 patients being enrolled in the trial, going into the open-label dose optimization phase and then those who respond will go into the double-blind randomized withdraw phase.

Understood. Okay. Can you comment on just the market opportunity from the perspective of -- these patients are currently being treated with stimulants or wake-promoting agents. Is there any off-label views of WAKIX in this patient population or perhaps have there been a lot of off-label use of other drugs like Xyrem in this patient population just based on your knowledge?

Yes. So Ami, I think it's well known that there has been off-label use over the years of sodium oxybate for IH because there was really nothing approved. So historically, I think -- and also in the Xywav Phase III program, I think patients who'd have been on sodium oxybate were brought into that trial. With regards to WAKIX and the clinical experience now, I think it is possible. We know that there's a lot of interest in WAKIX from both patients and health care professionals in WAKIX for idiopathic hypersomnia. It's driving the interest in our Phase III trial that we're very excited for. But the difference is for this trial here, we will not be enrolling -- we'll only be enrolling patients who are naive to WAKIX pitolisant who have not been on it. But yes, we do think there is some off-label use because these patients have a significant condition and HCPs are looking for effective treatments. But our -- the Phase III trial will only enroll patients that are naive to pitolisant.

Understood. Perhaps the last question on IH before we move on to the next indication. Could you provide a rough time line around how long it would take to enroll this patient population? And roughly by when do we think that we could get top line results?

Yes, Ami. We're estimating about 2 years to trial completion and top line data for this Phase III trial in IH. And we're putting a lot of resources towards this program, as I mentioned, that we're very excited about. And I'm happy to report we're on track to initiate this trial in the first half of this year.

Okay. So sometimes by first half of '24, the trial will be completed and you'll probably get data soon after.

That's what we're anticipating. Obviously, once we get into the clinics and we see the enrollment rate, we'll provide an update as we move along.

Okay. Let's move on to PWS and DM. You have Phase II studies ongoing for both of these indications. Can you discuss the prevalence of excessive daytime sleepiness in both these patient populations?

Sure. So starting with Prader-Willi syndrome. So Prader-Willi syndrome is a disorder of hypothalamic dysfunction, and there's estimated to be about 15,000 to 20,000 patients with PWS in the U.S. And the literature suggests at least 50% to 60% of them experience excessive daytime sleepiness, or EDS, in that patient population. But what's also important, Ami, is that symptom, if we look at the range of symptomatology in that condition, EDS can also drive some of the other symptoms that we're looking at in the trial as secondary outcomes, including the behavioral manifestations in PWS, including feeding behavior and the impact on the potential hyperphagia as well as cognitive dysfunction and cognitive impairment and learning issues in these kids. So as we'll discuss, the Phase III trial, primary outcome around EDS, but key secondary outcomes around behavior symptoms and cognition. With [ LARS 2 ] up -- go ahead, I'm sorry.

Yes, why don't we cover DM as well? And then I'll go to the next question.

Yes. So myotonic dystrophy is a really interesting, also multisystem disorder and recent epidemiologic data from actually a study that Nicholas Johnson was the lead. And it's from the New York State Newborn Blood Screening Program demonstrated that there are about 150,000 to 160,000 people in the U.S. living with the genetic defect for type 1 myotonic dystrophy and the sort of the predisposition as symptoms develop. But we also know, based on claims data that currently only about 25% of them or 40,000 patients are diagnosed in the U.S. However, in DM1, EDS and fatigue are the most common nonmuscular symptoms in patients with DM1. And importantly, what's really interesting is studies have shown that the impact factor, meaning how it affects their daily functioning for EDS and fatigue is as high, if not higher, than the myotonia and the weakness in the muscle symptoms in that patient population.

Okay. That's very interesting. I want to talk about some of the preclinical data supporting the value proposition of pitolisant in these 2 patient populations. Can you talk about the correlation between [ risk ] and deficit and ESS score in both PWS and myotonic dystrophy type 1?

Sure. Let me -- so I mean, I also -- let's see. So let me go to this slide here. Yes. So really interesting is kind of the orexin biology and the science evolves. So what got our attention in terms of, I think, as everyone is familiar with, type 1 narcolepsy is the prototypical disease for orexin depletion or orexin deficiency. What got our interest in these other conditions, PWS and DM1 is there's evidence in the literature you see on this slide, of lower levels of orexin compared to normal controls, not as low as in narcolepsy. So the panel on the left is a small series of patients with PWS. And you see the narcolepsy in the middle there panel on the left, levels are below 100. The diagnostic -- these are CSF levels of orexin hypocretin. The diagnostic criteria is less than 110 picograms per ml, so very low levels in narcolepsy type 1. But what you see for both PWS patients on the left and DM1 patients on the right, their levels are around 200 or so, that's lower sort of intermediate levels than normal controls, which is around 300 to 350 picograms per ml. So there's evidence of lower levels of orexin hypocretin. And in that setting, histamine becomes a major player, and we can talk about that as well as to why in terms of stabilizing states of sleep and wakefulness. Now the correlation -- to your broader question, the correlation between these levels in EDS is better demonstrated, I think, for Prader-Willi in the literature than for DM1. And there's also the animal models of sleep disruption are better for Prader-Willi than for DM1. But for DM1, it was sort of a clinical picture from a -- as a neurologist and looking at the symptomatology, the EDS, the fatigue and cognitive dysfunction, so the main nonmuscular symptoms all of which are mediated or partially mediated through CNS histamine circuits are major symptoms in DM1. So that's what attracted our interest to integrate pitolisant in that rare neurologic disease patient population.

Okay. Yes, this is -- that's why I was sort of leading with in DM1, we found that the CSF orexin deficit in these patients was not as severe as narcolepsy. And the low CSF hypocretin level didn't show a correlation with severe EDS. And then so maybe the question for you is that, is that necessary to have to see at this stage in a preclinical data set? Or does the sort of symptomology certainly support further study?

Yes. Ami, it's a great question. And as I alluded to earlier, I think the whole science of orexin and orexin biology continues to evolve, and it's a really it's a really interesting area. And I think that it would be nice if we did see that kind of as a biomarker, if you will. But it's really interesting. I mean I just read a paper the other night that in patients with type 1 narcolepsy, they were showing -- it was a series where the cataplexy events were not as typical, if you will, as kind of the -- what you typically would see in patients with type 1 narcolepsy and very low levels of orexin. And they were patients diagnosed with type 1 narcolepsy with orexin levels above 110 picograms per ml in the 200 range. So I think that the levels of orexin hypocretin are likely a dynamic state, which is also consistent with that spectrum theory of central disorders of hypersomnolence, NT1, NT2, idiopathic hypersomnia, where it's not necessarily a static condition, but could be more dynamic state. And I agree with you, if we had it as a biomarker, it could inform our clinical trials and patient populations more effectively, but I don't think we're there yet. And the other comment is in the U.S. it's not common practice to measure -- to do a lumbar puncture, spinal tap and measure CSF orexin levels diagnostically. But our colleagues in Europe actually do that much more commonly even in the clinical setting as a diagnostic maneuver.

Interesting. Okay. Let's move on to the Phase II trials that are underway. Perhaps if you could -- we could grow maybe Prader-Willi syndrome first and then DM1. Can you give us an overview of the Phase II trial design and the endpoints that you're evaluating? Yes, let's start there.

Sure. Let me see. So we'll start with PWS.

Okay.

Okay. So this is a -- it's a Phase II proof-of-concept trial. And as you see here, it's different than IH, this is the more typical prospective randomized controlled trial, 3 arms, with a high dose, low dose of pitolisant and placebo after a screening phase and then randomization and then eligible patients that continue will be offered to go into an open-label extension. There's a titration period. And then based on the age group. So this is in patients ages 6 to 65 in our discussions with FDA in the pre-IND meeting, since PWS is a neurodevelopmental disorder in kids, they wanted us to study children from the start. So it's a broader age range. So we have dosing that's age appropriate in this small proof-of-concept trial. And we're currently at 15 clinical trial sites in the U.S. been actively, I think, enrolling as everyone is aware. The primary outcome is around EDS, as I mentioned before. And it's being assessed by both an objective measure, sleep latency on the multiple sleep latency test as well as what clinicians use more commonly; subjective measures. And in this case, it's the Epworth Sleepiness Scale for children and adolescents or what's called the ESS-CHAD, caregiver version through an observer, a close observer, usually the parent, that is a key observer of level of sleepiness in these patients that are not good reporters themselves, which is why you've got an observer version. Key secondary outcomes that I think I've mentioned before, are different behavioral -- validated behavioral scales. We'll be looking at cognitive function through a computerized, a validated computerized battery called the Cogstate battery. Also looking at the clinical global impression of severity CGI-S of PWS symptoms rated by the clinician, caregiver burden, which is important in this population as well as assessing hyperphagia through typical instruments that are used, the hyperphagia questionnaire for clinical trials in this patient population.

Okay. That's helpful. So this is roughly a 60-patient study. With the endpoint, what is the bar for success for the study that would support further development?

Yes. So Ami, so this is the first time pitolisant is being investigated in patients with Prader-Willi syndrome. And although our end point is around EDS as opposed to hyperphagia, this has been a difficult patient population to investigate and assess. So this is a proof-of-concept study. It is not designed, it is not powered to demonstrate statistical significance. We really wanted to field this initial trial to look for signals, to look for signals of effect around EDS, either the objective and/or subjective endpoint as well as probing around some of those other key secondary outcomes: behavior, cognition and its impact on hyperphagia since that's a key symptom in this condition.

So it sounds like -- and based on our discussions in the past, you're not necessarily looking for statistical significance here given the size of the study and given that sort of a proof-of-concept study. Could you give us a rough quantification of how much of a change in ESS score you would see as clinically meaningful? Perhaps could we anchor that based on the data that we've seen in narcolepsy?

Yes, the short answer, Ami, is yes. I think it's really the Epworth Sleepiness Scale and what's in the literature would suggest a 2- to 3-point improvement is clinically meaningful and clinically relevant. And that is sort of how you would model it based off of the data that we have in narcolepsy, but noting that these populations are very different. And Ami, what I would say is that's the reason in our pre-IND discussions with FDA for both this program and the myotonic dystrophy, the DM1 study. The reason when you move away from central disorders of Hypersomnolence where the Epworth is more recognized, that's where the agency really wanted us to do the effective testing in addition because the term is fit for purpose. The FDA questions whether the subjective scales that clinicians use much more commonly in their practice, whether they are what's called fit for purpose. So that's why we have both the objective measurement of sleep latency and sleepiness and the subjective ones, which are what clinicians use on a day-to-day basis.

Understood. And can you remind me the 2 doses that you're testing? Have you disclosed that?

So the 2 doses actually in this study, they are -- so they're modeled off of our low dose of pitolisant in the approved label for narcolepsy is 17.8 milligrams. Again, the titration schedule, first week is 8.9 milligrams, second week is 17.8, and then you go up to the maximum dose of 35.6 milligrams. In this study, they're modeled based on the different age ranges to the age appropriate modeled off of that. In the DM1 program, since they're all adult patients, the low dose is 17.8 milligrams, and the high dose is 35.6 milligrams.

Okay. So in Prader-Willi, you're probably closer to the 8.9 for the pediatric patients?

So we're at different increments, which is based on -- yes different PK [ ghost ] modeling. So children 6 to 11, it's modeled for that age range and then adolescents 12 to less than 18, there are the different dosage increments and then adults as I said, [Audio Gap] around 17.8 and 35.6.

Understood. Okay. That's helpful. What would be the regulatory path if this study supports an encouraging proof of concept for the development?

Yes. So the regulatory path, if we demonstrate proof of concept, then we would meet with the FDA sort of the data would inform our approach. And we would propose a single robust Phase III pivotal trial model based on how the data read out here to address an unmet medical need in this rare disease patient population and then have discussions with the agency about the robustness of a single Phase III trial. And that would sort of be the plan from there. And then given that opportunity, we would then go into the clinic with the Phase III trial design that hopefully would be able to satisfy an NDA submission.

Understood. Perhaps we can move to the Phase [ II ] design for the myotonic dystrophy trial? And maybe you could point out the differences in that trial versus the PWS trial. Of course, you already talked about the difference in the dosing, what other differences should we be thinking about?

Sure, Ami. Yes, so this slide shows similar, you can see the schematic DM1, the Phase II clinical trial. So a couple of differences here. So the overall design is similar in prospective 3-arm parallel group design. And as I mentioned, all adult patients with type 1 myotonic dystrophy, ages 18 to 65. The low dose is 17.8 milligrams of pitolisant, the high dose is 35.6% and a placebo arm. 3-week titration, stable dosing weeks 4 to 11 and then into an open-label extension to assess long-term safety and effectiveness of pitolisant. This is a larger patient population. So we have designed a larger trial targeting about 135 patients. We plan to be at -- about 20 to 24 clinical trial sites. We are working with a couple of key sites in Canada, in the Quebec region where there are large populations, what are called founder populations of patients with DM1. And the primary outcome is around excessive daytime sleepiness. But since this is in an adult population where the maintenance of wakefulness test is the common objective measure used, which is not really validated in kids, Ami, so that's why we couldn't use the MWT in the Prader-Willi study and the MSLT was more appropriate for that age range. So the primary outcome around MWT. And then we also have the subjective instruments, the Epworth Sleepiness Scale and another called the Daytime Sleepiness Scale. The key secondary outcomes here are around fatigue. As I alluded to earlier in our discussion, fatigue is a really important symptom along with EDS in this patient population. So we'll be looking at fatigue and cognitive function as well as an overall sort of disease burden. There's a well-validated instrument called the Myotonic Dystrophy Health Index, or MDHI, which is a multi-symptom disease-specific burden of illness inventory that was basically developed by a group up at Rochester University, Rochester, Chad Heatwole, and this was published in Neurology in 2012. So it's a really well-validated disease-specific burden of illness instrument. So that's another key secondary outcome.

Okay. So I'm going to combine some questions that I've received on the portal here, and that's around kind of the time line for the enrollment of these 2 studies. There was some discussion about that in the recent earnings call. Can you help us understand where you are seeing some delays and some of the steps you've taken to address that? And help us understand the need to access sleep labs that was unavoidable, that may have impacted the enrollment time line.

Yes, sure. So I think I'll start with -- I think that everyone is aware that the COVID pandemic impacted the clinical trial sort of enterprise across our industry. And everyone had to sort of pivot and design in flexibility through the trials. Specifically for our programs, And as I mentioned on our last quarterly earnings call, many of our sites, because these are specialized patient populations, are in academic medical centers, and ones that really had to shift resources and personnel to care for COVID patients, and specifically around the sleep labs that you alluded to. So since the primary outcomes included the objective sleep testing, getting into the sleep lab is an important component. So I think that was some of the headwinds that we were facing, but we quickly pivoted. And as I think many sponsors and the FDA started to put out some guidance, around what could you do to address that and keep the trials up and running, especially in these orphan rare disease patient populations. So on this slide, just a couple of key things that we did for the programs towards that end. So basically, a mechanism was put in place for e-consent which then allowed for remote screening. And Ami, the overall theme was trying to reduce the overall burden on patients and families and the clinical trial sites and a lot of that was related to flexibility in the protocols to reduce the amount of travel back and forth to the sites. And remember, for Prader-Willi, we're only at 15 sites around the country. So patients and families need to travel back and forth to sites, and we're trying to lessen that burden. We also wrote protocol amendments for remote visits during the trial. This included telemedicine visits to do efficacy assessments and then sending out visiting nurses to the patients' homes to collect labs and do ECGs for safety assessments along the course of the trials, further reducing the amount of travel. We had to set up alternative sleep labs at some of the academic medical centers where the trial sites were, that were closed down because personnel was shifted to caring for COVID patients, which is obviously a primary importance. And then we added some additional sites, especially in areas of the country that were more open as we were going through the different waves of the pandemic just to support the ongoing enrollment of the trials. But I'm happy to report that these maneuvers and these strategies that we've taken has shown some increased improvement in the enrollment rates. And based on the effects of the sort of mitigation efforts that the team has put in place.

Jeff, just to clarify, is the access to sleep labs required both for the PWS and the DM1 study?

Yes. Both of the -- so the MSLT test and the PS -- what's called a polysomnogram, PSG MSLT for Prader-Willi requires patients to go into the sleep lab, which is not an easy thing for kids with Prader-Willi to do, and we've been trying to balance that. And as well as on the MWT or the maintenance of wakefulness test also requires -- it's not as involved, this study, but it requires going into the sleep lab. And there's a series of naps, if you will, over the course of the day, but that's how the MWT is administered.

Understood. Okay. I'm getting an indication that we have 2 minutes left. So let me jump to a question that I think people want to hear about HBS-102. You've indicated that you guys are doing some sort of work around that product and sometime later this year, you'll announce how you plan to move forward with development. Perhaps you could give us some preview into where -- what got you excited with this product and where you might seek to develop it further?

Sure. Yes. So I mean -- so HBS-102 represents an early-stage asset with some really interesting science and it could take us in several different directions. Also, the deal was derisked with minimal upfront payment because it is a very early-stage asset followed by developed milestones should it succeed. So I sort of threw up this slide. So HBS-102 is a melanin concentrating hormone or MCH receptor 1 antagonist, and it's targeted at the MCH neurons. And as you see here, these neurons are located in the hypothalamus right near the orexin and histamine neurons. So a lot of what Harmony has been working on, we seem to be centered in the hypothalamus. We have a saying amongst our teams or we refer to as the 3Hs, histamine, hypocretin orexin in the hypothalamus. So now MCH is another group of neurons in a similar sort of area sort of centered in the hypothalamus of the brain. So you see kind of that structure circuit diagram on the left, multiple projections, not dissimilar to orexin hypocretin, to the cerebral cortex to other areas of the hypothalamus to metabolic centers, as well as to what's called the limbic circuit, behavioral processing and connections actually to the reward center of the brain. So that translates into multiple functions around controlled metabolic function, feeding behavior. It's the control center for REM sleep, sleep/wake state -- stability as well and mood and behavioral processing. So there are a lot of potential clinical targets for this early stage asset. We're working with experts in the area of MCH biology. So we can better understand some of the opportunities as well as the challenges to inform the development program going forward.

Okay. All right. We will stay tuned for an update, hopefully later this year. With that, we have run out of our time. Jeff, thank you so much for your time today, and thanks to the audience for listening.

Yes. Ami, thank you. The time went by very quickly. But again, thank you for the invitation and opportunity, and look forward to speaking soon.

Same here. Thank you.

Bye now.