Alkermes plc (ALKS) Earnings Call Transcript & Summary

All right. Good morning, everyone. I'm Sandy Coombs, I lead Investor Relations here at Alkermes. I am happy to welcome you all today here in the room and on the webcast. I know it's a busy time a year, so thank you for making the time to learn a little bit more about Alkermes Orexin portfolio. Some housekeeping items for today, there will be two opportunities for Q&A during our roundtable panel with Dr. Maski from Boston Children's and Dr. Plante from the University of Wisconsin-Madison; Monica Gow, Co-Founder of Wake-Up Narcolepsy is here to join us, and we're fortunate to have them here with us. So please take advantage of the opportunity. During today's presentation, we will be making forward-looking statements. So please take a look at our -- so please take a look at our SEC filings for important risk factors that could cause our actual results to differ materially from those are expressed or implied in our forward-looking statements. We undertake no obligation to update or revise the forward-looking statements provided on this webcast as a result of new information or future results or developments. With that squared away, it's my pleasure to introduce Dr. Craig Hopkinson, to provide some introductory remarks and the agenda for the day. Thank you.

Well, good morning. I'm Craig Hopkinson, Chief Medical Officer and Head of R&D at Alkermes. And on behalf of our scientific and clinical development teams here, I'd like to thank you all for joining us here this morning. We've been looking forward to a meeting of this type for some time now as it gives us an opportunity to share some of our developing insights into the potential utility of therapeutics that harness the Orexin pathway. Orexin is considered to be the master regulator of wakefulness. As you know, the first embodiment of the Orexin 2 receptor agonist pharmacology has been focused on the restoration of Orexin signaling in disorders where excessive daytime sleepiness is the key symptomatic domain. Narcolepsy type 1, narcolepsy type 2 and idiopathic hyposomnia or NT1, NT2 and IH as you'll hear them referred to throughout today's session. Alkermes is the only company with patient data supporting phase II development of an Orexin 2 receptor agonist in NT1, NT2 and IH with a range of doses expected to accommodate the spectrum of potential indications with once-daily administration. As we review the data generated across the space and the competitive profiles that have emerged, we are confident in the differentiated features and profile of ALKS 2680. ALKS 2680 results from our understanding of the critical elements necessary to design a small molecule Orexin 2 receptor agonist. The program is built off of a strong scientific foundation beginning preclinically and extending into a Phase Ia study in healthy volunteers and a Phase Ib study in patient's populations that we intend to treat NT1, NT2 and IH. The early clinical program was designed to provide data relating to the safety, tolerability, the pharmacokinetics and dose response as well as the pharmacodynamics, which enabled data driven selection of our doses for phase II. ALKS 2680 is advancing along a well-defined clinical and regulatory pathway. We're currently enrolling in Phase II studies for NT1 and NT2 in multicenter global studies called Vibrance-1 and Vibrance-2. Today, we are sharing our decision to expand the Phase II program and advance ALKS 2680 into a study in patients with idiopathic hypersomnia. This will be a Phase II study. We made this decision based in part on clinician and patient feedback following the results of our 1b IH cohort and the limited treatment options available to IH patients. Advancing into a Phase II study will allow us to more fully characterize the profile of ALKS 2680 in this patient population. As we advance the development program, we are focused on three core elements. First, the quality of our clinical study conduct as well as the data that we're generating. Second, the speed at which we're advancing our program; and third, integrating the principles of patient-focused drug development throughout our program. Our Phase II program is designed to provide a strong operational foundation across all three of these priorities as we prepare to enter a potential Phase III program. We believe that Orexin therapeutics have the potential to fundamentally change the way that narcolepsy and IH are treated. All of this is exciting in its own right, but the Orexin system is also associated with the activation of multiple downstream neurotransmitters and neurocircuitry. Orexin-based pharmacology has the potential to extend to multiple CNS disease settings where sleepiness, fatigue, cognition and mood are prominent clinical features. We are levering this understanding and pursuing a multifaceted preclinical research program to identify the most promising lanes of development and evaluating the effects of Orexin 2 receptor agonist either as monotherapy or superimposed for other mechanisms to extend the spectrum of pharmacologic activity. This hypothesis and work are the foundation of our Orexin expansion strategy, Project Saturn. We're looking forward to sharing insights into this exciting work with you today. So with that is by way of introduction, this is how we're going to proceed today. Dr. Brian Raymer, one of our lead chemists and neurobiologists will present a brief overview of Orexin signaling in the brain and the role what it plays in narcolepsy as well as its direct relevance to other CNS disease states. Understanding the biology is critical to identify the clinical opportunities. However, the complement to the biologic understanding is the ability to develop suitable chemical matter. Medicinal chemistry has historically been an area of strength at Alkermes, engineering new molecules with appropriate pharmaceutical properties along with features important to the real-world needs of patients. Following Brian, Dr. Julie Himes, our Head of Clinical Development, will summarize the data that we've generated for ALKS 2680 in our 1b clinical program, providing a snapshot of the safety, efficacy and dose response across all three indications, NT1, NT2 and IH. Julie will also discuss our Phase II program, including the study design as well as the objectives for our two ongoing Phase II studies in NT1 and NT2. Rapid and efficient execution of our Phase II program is a critical priority for us as we look forward to the data readouts in our two studies next year and initiating our Phase II study in IH. Charlie Pak, who heads up our new product planning group will then share important insights into the epidemiology of narcolepsy and IH along with patient research that we've conducted. This research reveals significant unmet needs of patients despite the availability of current drugs and captures the critical impact that these diseases have on patients beyond the core representation of daytime sleepiness. These data and insights inform important elements into our clinical strategy and our clinical trial design. After Charlie, we have a thought leader panel comprised of two sleep specialists, Dr. Kiran Maski from Boston Children's and Dr. David Plante from the University of Wisconsin-Madison. We're also fortunate to have Monica Gow, a Founder of the Patient Advocacy Group Wake-up Narcolepsy here with us today to share patient and caregiver insights. Sandy will moderate the discussion, which will be fueled by questions that many of you have submitted. Following our panel, Dr. Bhaskar Rege, our Head of Early Development will outline how we use data-driven decision-making in development and give you some insights into the sophisticated experimentation and biostatistical modeling that we use to rigorously explore dose response relationships to inform clinical dose selection. Bhaskar will then shift his focus to Project Saturn to outline our approach to identifying our next potential clinical candidates and indications, selecting areas, which we believe have high translational fidelity. He'll introduce a framework of preclinical assessments that serve as the foundation of our decision-making process including microdialysis, qEEG and behavioral assays. Taken together, we use these data to identify and select potential indications at our next candidates for evaluation, either as single agents or as polypharmacy with other established CNS active compounds. Next, upto Julie Brooks, a seasoned neurobiologist, will share a selection of data from this program that support Orexin-2 receptor agonists as a potential treatment option in a number of potential indications such as ADHD and certain mood disorders where we see opportunities to develop new medicines with differentiated profiles to address unmet patient needs. Now we won't share the full extent of our work with you today for competitive reasons, but Julie's presentation will provide a window into our approach to fully harness the potential of the Orexin 2 pathway. Rich will then finish up with a summary of key themes for the session and the view ahead. We're excited to get -- to share this data with you, so let's get started. Brian, over to you.

Thank you very much for the kind introduction, Craig. I'm Brian Raymer, I'm a PhD chemist by training and a research project leader here at Alkermes. I led the research team that discovered Alkermes portfolio of Orexin 2 receptor agonist. I'm here today to tell you how the team utilized advanced molecular design principles to harness the broad potential of the Orexin mechanism. Orexins also known as hypocretin or neuropeptides produced in the hypothalamus. Initial drug development focus for medications utilizing the Orexin mechanism has been in sleep disorders given that Orexin is a key master regulator of wakefulness. But there's more to it than just wakefulness. Orexin neurons are multitasking neurons that regulate a set of vital functions including sleep and wake states, energy homeostasis, reward systems, cognition and mood. In today's presentation, I'll share Alkermes approach to Orexin 2 receptor agonist and how understanding complex interactions between a number of critical variables is key to designing differentiated small molecule Orexin 2 receptor agonist. The broad potential of the Orexin mechanism starts in the hypothalamus, colored blue in the lower center of the sagittal diaphragm of the brain. Orexin neurons project from the hypothalamus into multiple brain regions and modulate an array of downstream neurotransmitters. For instance, Orexin neurons can exert a central control of wakefulness connecting with different regions of the brain associated with wakefulness such as the dorsal raphe, basal forebrain, prefrontal cortex and other areas in the cortex that are involved in many essential functions, including sensory processing. Pathways modulated by Orexin may also be involved in the control of mood. As you can see by the dark red arrows, many of the brain regions involved in mood have very similar and may even have interacting connections with the Orexin pathways. These areas include connections from the dorsal raphe to the hypothalamus and from the substantia nigra through the to the prefrontal cortex and other areas of the brain. Pathways modulated by Orexin may also be involved in the control of attention. As you can see, represented by the black arrows, some of the neural pathways related to attention emanate from the dorsal raphe to the tuberomammillary nucleus and basal forebrain. Many of these pathways, for instance, those proceeding towards the prefrontal cortex, cingulate and cortex area, parallel the light blue pathways of the Orexin originating from the hypothalamus. How do orexin neurons activate downstream pathways via the orexin 2 receptor? Initially, the orexin 2 receptor is found in its inactive state. The orexin 2 receptor is a GPCR or G protein-coupled receptor, also known as a 7 transmembrane receptor represented here in light gray. With the lipid bilayer of the cell or neuro wall on either side, also in gray. Represented on the right is a fragment of one of the orexin peptides in orange. Orexin peptides are released from orexin neurons and are endogenous activators of orexin receptors in the brain. While starting in the inactive state, orexin signaling is off. Once the orexin peptide is released from the orexin neurons, it interacts with the orexin 2 receptor. The receptor then moves to an active state, as shown by the shift from the light gray representation to the blue representation, the orange arrows highlight some of the transmembrane areas that undergo confirmational and location changes that facilitate downstream cellular neural signaling via the release of G proteins. Understanding the orexin peptide interactions with the orexin 2 receptor is key to replicating this activation with a small molecule. Key activator interactions include hydrophobic interactions with phenylalanine 227, a key hydrogen bond with histidine 350, as shown by blue circle and by not blocking certain areas of the binding pocket allowing movement of glutamine 134, a key step in the activation of the receptor. And orexin 2 receptor small molecule agonist at most mimics these activating interactions and allows for movement of key residues could replace the endogenous orexin peptide activation. So that being said, mimicking the interactions of an approximately three kilodalton or larger orexin peptide with a small molecule less than 500 daltons is easier said than done. Drug developers face numerous challenges in replicating the orexin peptide interactions with a small molecule including balancing complex and competing critical variables, essential to designing small molecule orexin 2 receptor agonist. Among these variables, for instance, Alkermes use key orexin 2 receptor structural information to engineer in potency and selectivity into our portfolio of molecules. Together with metabolic properties, these attributes are key to enabling a wide range of doses with the potential to be well tolerated. We also focus on oral bioavailability and brain penetration to allow for low overall dose levels for -- and oral administration and particularly important for medicines seeking to mimic a natural sleep wake cycle. We focused on molecules with suitable pharmacokinetic properties. We have a deep expertise at Alkermes in designing complex potent and selective molecules, integrating a diverse array of pharmaceutical properties with a particular emphasis on the relationship between pharmacokinetics and pharmacodynamics. In the case of ALKS 2680, we designed a highly potent and selective GPCR agonist. We synthesized and screened hundreds of compounds to arrive at ALKS 2680 and the other compounds in our portfolio that balance these variables. The pharmacokinetic profile is one of the key features of orexin 2 receptor molecules and I'd like to delve into that area in a little bit more detail. The pharmacokinetic profile of a small molecule is a key player in the balance between safety and efficacy features. Let's walk through an illustration of a few key PK parameters that informed how we approach the molecular design of our orexin 2 receptor agonist. In the diagram, the Y-axis is the plasma or brain concentration and the x-axis represents time moving from left to right. The blue line represents an illustrative efficacious concentration in plasma brain levels for given indication here using NT1 as an example. From the start of the program, we wanted to design a small molecule that would mimic the natural sleep wake profile in duration provided by endogenous orexin with once-daily dosing. We also targeted high potency to drive meaningful efficacy with low overall doses and exposures. The black waveform line represents a hypothetical plasma concentration over time. Once the plasma concentration of the orexin 2 receptor agonist rises over the efficacious concentration level, in this example corresponding to an indication such as NT1. Wakefulness would be promoted by sufficient interactions of the small molecule with the orexin 2 receptor. While above this concentration level, the orexin 2 receptor agonist would stimulate downstream neurons and neurotransmitters turning on the wakefulness system. Over time, as the plasma concentration decreases, the downstream signaling and wakefulness promotion would subside. This hypothetical PK profile can provide flexibility in the magnitude and duration of pharmacodynamic effect and provide differentiated molecular attributes. A PK profile can be targeted to yield dose-dependent proportional increase of AUC or area under the curve and a lower than proportional or nonproportional increase of the maximal concentration after the dose. The lower than proportional maximal concentration is designed to reduce the risk of tolerability issues, and I'll circle back to it shortly. As you can see in the schematic doses A, B and C provide different profiles and durations of effect, relative to the efficacious concentration, in this case, representing NT1. Different indications may require different efficacious concentrations and dose levels. This schematic shows a second set of higher doses. In this example, that would allow for efficacious concentrations to be achieved in indications that may require a higher level of orexin 2 receptor agonism such as NT2 or IH. A wide therapeutic margin also allows for dosing flexibility across indications and patient variability within an acceptable tolerability range. You'll see that at the top of the graph, there is a red line representing a hypothetical tolerability limit. This concentration would be associated with unacceptable safety and tolerability profile or toxicity. So exposure levels must stay below this threshold. All these PK design principles were considered in our development of ALKS 2680. We targeted high potency to allow for low doses and low overall exposures that drive wakefulness, while keeping the maximum concentrations low to increase tolerability. ALKS 2680 was designed to enable dosing flexibility with the potential to accommodate variability and indications as well as individual patient profiles and treatment objectives. In this introduction, I hope you have seen how Alkermes utilized advanced molecular design principles to engineer ALKS 2680 and harness the broad potential of the orexin mechanism. As you know, orexin neuropeptides are master regulators of wakefulness that work in a diurnal manner and we designed our small molecules to mimic this profile via once-daily dosing. In addition to wakefulness, orexin 2 receptor agonist may benefit additional symptomatic domains such as fatigue, mood, cognition and attention as we saw on the overlay of the different but overlapping neural systems. Understanding how the orexin peptides activate the receptor and cascade signaling across the brain is key to designing small molecules that harness the potential of the mechanism. Alkermes chemistry design approach is focused on key parameters such as potency and a targeted pharmacokinetic profile that may address the needs across a range of potential orexin relevant indications. Thanks very much for your attention and looking forward to connecting with you after the presentations. And I'd like to introduce Julie Himes, our Head of Clinical Development.

Thank you, Brian, and good morning. I am delighted to take you through our Phase Ib program and tell you about our strategies in Phase II. My name is Julie Himes, I lead Clinical Development at Alkermes. I am an infectious diseases physician by training. I spent 10 years at NIH doing both clinical as well as bench research and then the last 22 years in industry, I've worked mainly in orphan diseases and rare diseases, and I have been a company such as Millennium, Vertex at Takeda. So let's begin with a look at the clinical features of our target indications, central disorders of hypersomnolence, including narcolepsy and idiopathic hypersomnia. As you can see with few exceptions, the individuals impacted by these conditions share many of the same symptoms, including excessive daytime sleepiness, brain fog and disrupted nighttime sleep. And while there are elements that can help differentiate these conditions such as sleep onset REM periods and cataplexy, the variability and subtleties of these disorders can be challenging even for the most seasoned clinician, as there are considerable diagnostic challenges. ALKS 2680 is being developed as a potential new therapeutic option for all three of these conditions and is currently the only orexin 2 receptor agonist advancing into Phase II in these areas. We believe that this is an important differentiating feature of ALKS 2680. And today, I will share the highlights of our program, our Phase Ib study and our plans for Phase II. However, there are comprehensive data sets available on our website from presentations and posters given at major medical meetings and I encourage you to look at those. Before we dive in the data, I want to provide an overview and hit on a few salient features that further differentiate ALKS 2680. ALKS 2680 is a highly potent and highly selective molecule. And we -- in addition, as Brian has described, we've specifically engineered suitable pharmaceutical properties into its creation. It is orally bioavailable and we specifically designed a PK profile that mimics the sleep-wake cycle and is designed to improve support improved daytime wakefulness with single daily dosing. And because of our positive data in our Ib study in patients, including safety, efficacy and PK, we have moved quickly and efficiently into Phase II with a high degree of confidence in our chosen doses. And I know Bhaskar will walk you through this data in later this morning. This slide shows the trial design used in our Phase Ib proof-of-concept study, which was key and I am forming our Phase II design and our dose levels. This is a highly efficient design that used a 4-way crossover design which permitted the testing of three different doses of ALKS 2680 and placebo in every patient enrolled. This crossover design was an important attribute that led to our Phase II dose selection as it allowed for each patient to service their own control, thus reducing and mitigating the types of interpatient variability that is commonly seen in clinical trials. Our patient population had to meet the ICSD criteria for their respective disorders. However, eligibility criteria did not mandate baseline maintenance of wakefulness testing results. So the variability was consistent with what one would expect in the real world setting. And the objectives included safety, efficacy as measured by the maintenance of wakefulness test. So let's move to the data, starting with safety. Principally, the key objective in a Phase I program is to identify any early safety signals. ALKS 2680 showed itself to be generally safe and well tolerated at all doses tested in NT1, NT2 and IH. Nearly all of the adverse events observed were transient and mild in severity. And in clinical trials, the designation of mild actually is firmly defined in the protocols. It means that while the adverse event was noted, it was tolerable and did not interfere with acts of daily living. There were no serious and no severe AEs and no AEs leading to early treatment discontinuation. The most common AEs in each population that were considered treatment-related or listed here, but is principal included insomnia, polycoria and dizziness. Importantly, there were no clinically meaningful trends or results or findings in laboratory values, in vital signs and ECGs, there was no evidence of a cardiovascular safety signal. This slide provides the results of our primary efficacy outcome, the mean sleep latency, which is essentially how long patients can resist falling asleep during the maintenance of wakefulness -- with maintenance of wakefulness test, which was a 40-minute test. ALKS 2680 demonstrated clinically meaningful and statistically significant improvement from baseline in mean sleep latency compared with placebo at all doses across all indications. The graphs that we show here include the observed or the absolute mean sleep latency and below the graphs, we provided the primary efficacy analysis. The least square mean difference in sleep latency versus placebo. And you can see from that analysis that at each dose tested in all indications were highly statistically significant. I think it's important to note in presenting the data in this aggregated form that there are some key takeaways I'd like to leave you with. First of all, ALKS 2680 improved mean sleep latency in all patient populations. And this is important because it's irrespective of the underlying orexin levels in these different conditions. Secondly, as I mentioned the results were statistically significant and clinically meaningful. But what's really notable is the consistency of this result, irrespective of the variability of baseline assessments. And finally, the dose-dependent and clinically meaningful improvements in wakefulness that resulted in patients achieving mean wakefulness levels of normal individuals, normal healthy individuals that typically have sleep latencies of 30 minutes, plus or minus 10. So as these data that give us great confidence to move into Phase II is the consistency, is the magnitude of the effect and is the efficacy that we're observing in patients who endogenously produce orexin and those who do not. So let's move forward. So in summary, we are ready to move into Phase II based on the data we've showed you so far. The safety profile shows that ALKS 2680 is generally safe and well tolerated. And as I described to you, the efficacy shows consistency shows a high magnitude of effect and it shows it works across three different indications, as I described. Now I'm going to take you through the designs of our Phase II studies. With two Phase II studies ongoing, Vibrance-1 in patients with NT1 and Vibrance-2 in patients with NT2. These studies share some of the same design features. Both studies have a double-blind treatment period where we will be comparing three different dose levels against placebo for a period of time. And then all patients will roll into an open-label extension where there will be an adjustment of dose. All patients will begin at the mid-dose level of 6 mg -- or I'm sorry, will depending on the trial, they'll begin at a mid-dose level, and then we'll be able to adjust that dose. And this is a unique feature of the study that we believe is going to yield some very important information in terms of patient preferences that we will then use and build into our Phase III program. This table gives some additional details and points I'd like to emphasize across these 2 studies, we are exploring a range of doses and the ability to dose adjust in the open-label extension is a unique and differentiating feature of this program. We are also in the process of opening a long-term open-label extension study in the very near future, the patients will be able to roll into and continue to receive ALKS 2680. And in addition to these two Phase II studies I've described here, we are in the design phases of Vibrance-3, which will be a Phase II study in patients with IH and we plan to initiate this next year. So I just want to pause here and emphasize that these are serious diseases and they impact so many aspects of patients' lives and their families of hives. And as you'll hear from Charlie next, the impact of these disorders far exceeds what can be measured in the sleep lab. Hence, we designed our Phase II program to collect data to support the profile of ALKS 2680 both in terms of classic symptoms, registrational outcomes as well as those symptoms and outcomes that are important to patients. And we're doing this using validated scales, including sleep quality, cognition, fatigue and the ability to engage in daily activities and the impact on relationships. So in my experience, across multiple therapeutic areas, including rare diseases, the success in organizing and executing clinical trials is enabled by the quality of the data supporting the potential medicine. The more information we can share with our clinical researchers regarding the safety, the tolerability, the efficacy of an experimental agent, the better they are to determine whether or not they'd like to offer this to their patients that they take care of as an option. With the data that we presented at major medical meetings by key thought leaders associated with this program, a generally well-tolerated profile has been demonstrated across a range patient types, and we have shown some exciting efficacy signals. So we believe we have all the elements in place as we accelerate our program into later-stage development. Our investigators are very enthusiastic based on the data we've generated in our Ib program and we are building momentum in Phase II. That's with an intense focus on execution, we expect to have the data from Vibrance-1 and Vibrance-2 sometime in the second half of next year. And in parallel with conducting our Phase II, we are preparing for success and doing the foundational work to plan for our potential Phase III program. So we're in a strong position and we're looking forward to updating you in the future. So now I'd like to welcome Charlie Pak to the podium. Thank you, everyone.

Thank you, Julie. My name is Charlie Pak. I head up new product planning here at Alkermes, and I look forward to sharing with you the insights that we've gathered around narcolepsy and IH as it pertains to both prevalence and the experiences that people living with these conditions have. Now as we deepened our understanding of these patients, there were a couple of nonobvious insights that we gathered. One around prevalence as it relates to the prevalence of NT1 and NT2 and also perhaps the discrepancy between the view of unmet needs coming from treating physicians as well as patients. And the reason why we are looking at this is to better inform ourselves as Alkermes in terms of how and the why of developing drugs for these conditions. So let's turn to prevalence. Unfortunately, narcolepsy and IH affect people around the world. For today's session, I'd like to focus in on the U.S., both for the patient experiences and the prevalence. Having said that, we have conducted primary research with patients outside of the U.S. and unfortunately, the experiences that they have are also uniform across the different regions. And as far as the prevalence goes, again, I will focus in on the U.S., but these can be scaled to the rest of the world. There are about approximately 200,000 overall prevalence narcolepsy patients, of which 100,000 are diagnosed. Now that simple statistic that 50% diagnosis rate really doesn't do justice to the patient journey. They can take 10, 15 years between exhibiting the first symptoms to actually receiving the diagnosis of narcolepsy, which really talks to some of the challenges that Julie had hinted at about getting there. It's our thought that greater awareness of this condition, greater availability of novel therapies may increase the diagnosis rate, which, of course, in turn, can improve and increase the treatable patient pool. Now in addition to the narcolepsy patients, there are 40,000 diagnosed IH patients. And again, this is another area of diagnostic confusion between IH and narcolepsy. Again, given some of the overlap that Julie had talked about in terms of the symptomology. Another area of diagnostic challenge relates to the prevalence between narcolepsy type 1 and narcolepsy type 2. Now earlier studies, suggested there were more NT1 patients than NT2. And these are in the silver long stress both based on the medical records coming from single counties in single states. So Minnesota and Washington and these suggested that perhaps it's a 2:1 ratio of NT1 to NT2. However, the understanding is evolving over time. And if you look at the later publications, looking at a variety of methodological approaches, they suggest that there are actually more NT2 patients than NT1. So [indiscernible] Also adopted an audited medical record approach. However, with updated ICSD-3 diagnostic criteria. There was also a claims-based approach presented in sheer and also internal work that we had done. And then yet again, another approach in Ohio where they looked at prospective surveys that were sent out to residents of 15 states in the U.S. Despite the commonality -- despite the different approaches, the commonality is that they all suggest there are more NT2 than NT1 patients, approximately a 1:2 ratio. Now regardless of the prevalence of NT1 to NT2, what we know and what I'll share to you is the unmet need that's present in all of these patients. And it's our belief that any therapy that's indicated for the treatment of both type 1 and type 2 would be a welcome addition. There are, as you know, drugs approved for the treatment of narcolepsy. Some of these approved more than 20 years ago and the net sales across these different therapies generated over $2.5 billion. Now given the availability of these therapies, we were interested to understand what is the current satisfaction almost by physicians, the treating physicians with these available therapies. So we conducted primary market research with treating physicians and asked them on a 10-point scale of satisfaction. So one represents completely unsatisfied with the current therapies, 10 representing I am fully satisfied with the therapies that are available today. On average, the treating physician gives it about a 6. So middle of the road. And this is not what we hear from the patients a share. There is no ambiguity from what we've heard from the patients in terms of their satisfaction levels. So let's turn to that. I think the -- at a high level, the overall unmet need is well illustrated with a survey conducted with over 1,000 patients for narcolepsy, I'm summarizing this review article. Almost all of them had been previously treated with at least one of these approved therapies. And despite that, the majority of them complained of daily narcolepsy symptoms 34%. And an even greater proportion talked about the negative impact that it has on their work, on their school performance, 84%. So this high level of you really caused us to want to understand this at a deeper level at an individual patient level. And so we've conducted primary market research with narcolepsy and IH patients. And what we've learned from that really is well captured under the construct of Maslow's hierarchy of needs. So as a reminder, this is a hierarchy of needs that was described that is relevant for all of us, as human beings that we all have needs that these can be classified into lower order needs that need to be met and addressed for us to think about these higher order needs. And these lower order needs are foundational needs, the need for well-regulated sleep, just as important as the need for water, air, shelter. And if these needs are met, we don't have the opportunity to think about the higher order needs of how do we actually fit in with society? How do we interact with friends and family? What is it that I want to be and the opportunity to actually self actualize and achieve those goals. And unfortunately, what we've heard from narcolepsy and IH patients is that they have to focus their energies on these foundational needs. And they avoid "Wasting energy on higher order needs." And really what brought it home was one example, one analogy that one of the narcolepsy patients had mentioned, and he talked about it analogous to your phone. So at the end of the day, if you're like me, you plug in your phone, then you wake up, unplug and you have 100% charge. You're fully in the green. And this patient talked about when he wakes up, he does not start with 100% charge, he starts the day in the red and having to adapt to that on a daily basis. And that adaptation on a daily basis means when am I going to be able to take a nap because I have to do this. Am I going to actually interact with friends, how am I going to interact with family? And this daily struggle unfortunately, extends into longer term where it actually talks about what sort of job choices can I make? Should I pursue higher education? Can I have a relationship? Should I get married? Should I have children? Who's going to take care of the children and this is an ongoing struggle. When we ask the patients, what is the impact of the symptoms? And how does it affect your lives? They talk about the impact on their safety, productivity, mental health and their relationships. And I think what really brings it home is to actually hear from the patient's words. Under safety, one example, this narcolepsy patient talked about. I was at work one day. I fell asleep holding a piece of industrial glass and I dropped it and almost cut my artery, my neck. On productivity, this was a patient recalling back to when she was in school. I used to fall asleep in class all the time and everyone would get upset. I would end up into detention for sleeping in class. I would try so hard to stay awake. Everyone just thought I didn't care enough to stay awake. And you can imagine the toll that this takes on their mental health. I felt inadequate. I had low self-esteem due to not having the energy to do basic things my peers did and lonely because I would spend so much time on my own sleeping. As far as relationships, this was coming from a mother who is on vacation at the beach. I don't feel comfortable taking my son or my foster son on vacation alone because sometimes I have to sleep. I cannot sleep. So we're at the beach and I have to bring somebody with me so that I can make sure that children are safe. So when we ask these patients, what are the areas of unmet needs? What is it that you would like to have addressed. They talk about getting better symptom control. They talk about having a disease-modifying therapy that is not a stimulant. For symptom control, I would be alert. I wouldn't constantly feel like I need to take a nap despite all the treatments I'm taking. Disease-modifying I wish there was an option from my sleep disorder that could treat the cause and not just the symptoms. You would allow my body to do what it should be able to do on its own. And on stimulants, I'm fearful that I'm hurting my body or that I'll lash out at people. It's scary being on something that so strictly controlled. I can't really trust stimulants to help me. So in closing, I want to take us back to a comment that I made at the start is why are we looking at this? Why are we looking at patient insights. And it's our belief here at Alkermes that understanding the patient experience and instilling it into development program is just as important as understanding the science that underlies narcolepsy that Brian talked about. It's just as important as clinical outcomes, regulatory considerations and we intend on weaving all these together into our clinical development program, into our clinical trial design with the aspiration that we will be able to generate a therapy that meaningfully addresses the needs of the patients. Thank you. And with that, I'll turn it over to Sandy.

I'll ask our thought leaders to join me at the front here. I should have mentioned it at the beginning, but there is an opportunity to submit questions via the webcast viewer for folks viewing from their computers at home but I also invite the folks in the room to join in the discussion. I'm delighted today to be joined by Dr. David Plante from the University of Wisconsin-Madison and Dr. Kiran Maski from Boston Children's and Monica Gow, who as a co-founder of Wake-Up Narcolepsy. They're here to share with us their clinician point of view and also their lived experience dealing with these serious diseases. So maybe just to warm up and start the conversation. I invite all of you to -- based on your experience, provide some context on what the current unmet needs are for patients with narcolepsy and IH.

Yes. I think the major thing that people are dealing with is lack of efficacy, as kind of shown here in some of the slides, it seems very consistent with the reality of what people are experiencing where they're operating at maybe a 6 out of 10. And so they're making decisions all the time about how to prioritize work, school, how to balance that and I work with a younger population of patients, so children up to age all the way up to age 25. And so it is interesting to see what kind of has to be left out of those calculations. And oftentimes, it's social consequences so they're prioritizing school and early work, but don't have the bandwidth to form relationships or hang out with friends in a meaningful way. And then I think from the survey that was highlighted here the patient voice survey. Other details seem really clinically relevant. So lifestyle factors of taking the medications. So medications such as oxybate really take quite a bit of planning. You have to have a regular amount of sleep, you can't drink alcohol. You can't take any sedating substances. And if you think about young adults, like sleep irregularities, socializing with friends with alcohol is a big part of culture and so they feel sort of shut out from that. And I think other things are having to redose medications through the day, sometimes gets onerous, especially for younger populations. They have to go to the nurse's station or something like that in the middle of the day to get a second dose. And then side effects, I think, is the other big thing. I think people struggle with a high degree of side effects with the current therapies. So things like insomnia are common, but more issues related to headache actually are big problems for us as well as cardiovascular issues, which I think is getting more attention these days from the literature.

That's excellent. Monica, since your son was diagnosed when he was a child and is now an adult, and maybe you can provide some perspective on Dr. Maski's comments and add what you think the patients are still where the unmet need is.

Sure. In addition to what Dr. Maski said, I think getting people diagnosed early is key to having a productive life while trying to manage narcolepsy. We were fortunate. My son was diagnosed within 3 months, and we learned that people, once we -- the reason we started wake up narcolepsy is what we learned after we found this out on -- of course, we turn to the website to the Internet to find out what was happening in the space and not much was happening at that time, that was 2008. So I think the early diagnosis is key finding the right doctor who knows what they're doing and how to prescribe the medicine is key. A lot of the sleep doctors are focused on sleep apnea and not all of them are knowledgeable in narcolepsy. And then also finding that community where they can be accepted. We hear so often that people go years and years without meeting someone in person who has narcolepsy. So we try to have community events to get people together to just talk and meet someone else who's going through what they're going through. A lot of the stigma to can be reduced with awareness and education people are embarrassed. They don't want to talk about their narcolepsy, especially to employers who would not understand the need to sleep. I just see it as -- this sleep is the thing that is so prominent and you're a person with narcolepsy from what I see is you're just always managing that. When am I going to nap again? How am I going to go to school and stay there all day like, where am I going to strategically nap? We used to -- when my son was in middle school, I would show up in the parking lot, so he could run out and nap and then go to like the Halloween party or whatever what was happening so, but not every child has that. So you just have to provide support as a family.

Maybe Dr. Plante, can you talk through some of the challenges that you faced when diagnosing NT1 versus NT2 or IH?

Sure. So it's -- and I would just echo what's been said so far. I mean, I think most of the treatments are kind of like band-aids like they kind of help people get through the day but they don't sort of oftentimes, especially in the case of narcolepsy type 1 treat the underlying problem. In terms of the differentiation diagnosis of NT1 versus NT2, that's challenging, all of these diagnosis really hinge on a very complicated in laboratory setup testing called the polysomnogram in multiple sleep latency test or MSLT, where people have to spend mainly 24 hours in the sleep lab with an overnight sleep and then multiple repeated naps across the course of the day. And we're really distinguishing these disorders using those tests to see how quickly people fall asleep on the naps and also if they go into early onset REMS, which is a marker of narcolepsy. In the case of NT1, though, we know that it's caused by fundamentally orexin deficiency. There's also another way to make the diagnosis, which is with a lumbar puncture. That's challenging because a lot of patients aren't super excited about doing a lumbar puncture. And also up until relatively recently, it was hard to actually measure that value in a laboratory, either had to be via a special academic institution, but now you can send it away to Mayo. So there's challenges in terms of like just access to the testing for a lot of people. There's a challenge of recognition. You can't make a diagnosis, if you don't think about it in the first place, and a lot of providers, I think, out in the community, weren't necessarily thinking about narcolepsy sort of being kind of up on the differential. And then oftentimes, people by the time they get to us in clinic and sleep clinics or specialty, they've been treated with a number of different medications for other processes, comorbid depression and things like that, that will impact our ability to use our standard sleep test. So we will not be able to oftentimes detect early onset of REM periods if people are on an antidepressive medications. So even just sort of where people come to us can make it challenging based on sort of a number of different factors. And then I guess the final thing that I'll say is not everyone manifests with cataplexy who has type 1 narcolepsy. So it's another challenge in our field. So cataplexy, which is the loss of muscle tone in response to emotional stimuli, classically laughter or mirth is a very specific symptom of narcolepsy. So when as clinicians we hear that in combination with excessive daytime sleepiness, we automatically gravitate thought this person may have type 1 narcolepsy. It turns out, though, about 20% of people who meet criteria for narcolepsy. So type 2 narcolepsy don't have cataplexy. They don't complain about this clinically, even when you ask them. But it turns out when you do a lumbar puncture, 20% of those folks will actually have NT1 so they'll have orexin deficiency. So even identifying the pathology based on clinical presentation can be very challenging as well.

Okay. In the room? Okay. Joe, please. Hold on one second, we'll get you a mic, and then we can be on the webcast, too.

Joe Thome from TD Cowen. You mentioned earlier the lack of efficacy of current available agents. I guess when you look at the MWT, is there a sweet spot for efficacy on that? Because obviously, the orexin class is getting towards the end of this. And the company mentioned the opportunity for flexible dosing. Is that when you would use flexible dosing? Or how do you incorporate that?

Yes. I think the MWT questions. Really good. I was on the practice parameters for the treatments of CNS disorders of hypersomnolence and so we looked at the MWT as sort of an objective gold standard measure. Interestingly, compare to sort if you look at the classes across all of the narcolepsy therapies, they really range anywhere from like 1 minute to maybe up to like 6, 7 minutes, maybe maximum 12 minutes in terms of the total mean sleep latency improvement. So the threshold of clinical meaningful improvement, we have to set it very low. Otherwise, we would have had no medications really to evaluate. So the MCID is only 2 minutes. So when you're looking at these orexin class drugs, this is really exciting for us to be seeing these numbers. And then to the flexibility question, I think I do see more heterogeneity about that particular need. So some people, I think most people, I would argue want to take a single medication that lasts through the day. But some people do benefit from taking naps or things like that and like the flexibility of taking a nap, then taking a second dose after. So there is some variability in the field about that. But I would say the majority of working and school-aged people once a day would be helpful.

Marc Goodman from Leerink. I was hoping that each of the physicians could describe how many patients they treat and how many are NT1, how many are NT2, how many are IH and whether the diagnosis changed over time? And we're trying to fast -- second question is kind of fast forward 4 or 5 whenever these orexins are on the market. If there's an orexin that's only for NT1, but then you have an orexin all of narcolepsy, how much that's going to matter. And if you only have one of them versus the other, we're trying to gauge like how this whole competitive dynamic is going to matter, too, by having a full narcolepsy indication versus just NT1.

So I guess I can start without having an EMR open in front of me and looking at the list of -- I think that -- and also, please keep in mind my practice is probably going to be a little bit biased just because of my sort of areas of expertise and also the people who might seek me out. So probably I think it's about 50-50 in terms of NT1 and NT2. And then for me, I have about double that in terms of IH patients. But again, I'm not sure that you can extrapolate that to the general population in terms of sort of your other question about like sort of the specificity of an orexin agonist related to diagnosis. I mean I think that if you have NT1, you know that it's an orexin deficiency. So you're basically treating directly sort of the fundamental problem. These other disorders, I think we don't know what causes NT2. We don't know what causes IH fundamentally. So the question is like what's the best treatment option. We know that probably people are going to need different doses, right, if they don't have an orexin efficiency. So I think that will affect the potential choice of agent but then also having the ability to try different agents is really important as well. If a drug is only approved for 1 type of narcolepsy, then we'll have to be very careful and can only use it a limited number of people at a limited range.

Before we go to the next question in the room, Monica, can you give us a sense of how aware patients are of new drugs in development and the research that's being done in this space?

I would say there are many different variances that would go into that, one would be the relationship they may have with their physician and then the physician's knowledge of what's going on in the space with clinical trials, two their financial ability and their geographic location, definitely would play a role. I would also say whether or not they're connected with any patient advocate organizations that share this type of information. And also, lastly, just their interaction with people on social media, which is where they learn and gather a lot of their information depends on those four things, I would say.

That's really helpful.

David Amsellem at Piper Sandler. So wanted to ask a question about safety and tolerability of the clinicians up onstage. So in the Phase Ib studies, the couple of the AEs that were identified. One was visual disturbances, the other was daytime urinary urgency. And I was hoping you could contextualize those AEs, the extent to which you're concerned about them. And help us understand how to think about those kind of AEs in the context of the other treatments that are available, such as the oxybate or stimulants or modafinil and help us understand how you're thinking about the body of safety, tolerability data for 2680 in terms of what we've seen thus far relative to what you're currently using for your patients?

I can start. I mean I think currently with side effects of the medications we're seeing, it is probably around 30% to 80% of AEs. And the more common side effects are increase in mood disturbances to the point of suicidality and hypertension, so increases in blood pressure and then headaches, appetites, depression, weight loss. I would say those are sort of the big 5 for us. So the side effects with the oxybate class, David can chime in, are relatively new for us, things like increased urinary urgency, hypersalivation, I think, are not something that we're encountering. Visual disturbance also, I mean, frankly, I talk to my patients about anything that comes out in the literature about orexin agonist and I've mentioned the urinary urgency. It does not seem to bother them when they really talk about the benefit of what the wakefulness component are. I mean, anecdotally, I have a couple of patients who we're like that. It's not a problem. I drink cold water all through the day, anyway, like I'm going to the bathroom already a lot. I think it is the magnitude of what the results are in a larger study, though that might impact that. The visual disturbance, I think, is concerning. I mean I think anyone who has visual disturbance, that's affecting their ability to perform or drive is going to be of concern and that's something that obviously needs to be evaluated in a larger study but I think looking at the published data doesn't really suggest that, that's a big concern at this point, but I agree, it has to be watched.

I mean I would just build on that. I think you have to take sort of the data that's available currently like in context, right? So that study, people were given a specific dose. We're very tightly monitored sort of in a sort of very careful setting for full 24 hours. The report of sort of what may be a side effect when you're out in the world, taking medication on your day-to-day life may be very different. That said, my understanding to about the visual disturbances that we weren't a study site, so I would defer to the company and all this is that really those occurred at higher doses, they were short lived, transient and did not require any specific treatment. And I do think that also for Vibrance 1 and 2 studies, there is like an ophthalmology examination as part of the protocol, so to kind of really keep a close eye on that because obviously, that would be a concern in terms of polycoria. I agree, I think that most patients would see it is that would be worthwhile of a trade-off, right? And the other thing is we don't know how it's going to play out over time because, again, in the early study, it was a transitory administration watch for 24 hours. We don't know that is something that people sort of equilibrate to over time.

Joel Beatty from Baird. In a slide earlier today, we saw that oxybate seem to be the best-selling class for narcolepsy and IH. If the ALKS 2680 data looks as it does so far, how would you decide between using oxybates and ALK 2680? And would there be some patients that you might use both in?

So I mean, in terms of like -- I'll get to the last question first because there some patients who may need both potentially, yes. So it really depends on the person's clinical situation. We use polypharmacy quite frequently in these disorders. I think it would depend a little bit on how sort of an orexin agonist since it would be dosed during the day affects sleep and I think that is kind of an open for trying to understand big picture what that's going to look like in terms of like would I choose one over the other? I mean I think that you would look first at efficacy data, sort of once these sort of things move through the pipeline and have also a sense of like what the person's diagnosis is. And I think for NT1, pretty straightforward that probably an orexin agonist would make by far and way the most sense because again you're targeting the fundamental pathology for NT2, for IH, it's a little bit less clear. The phenotypes can get a little bit more complicated in those disorders. And we don't really know exactly how that's going to play out in terms of like things like long sleep time, excessive sleep inertia and some of those other symptoms that come up, especially more in the IH phenotype side. I think it really depends on sort of how people do with the medication in these studies.

Yes, thank you for the time today. In the real world, I guess, could you explain how, if at all, you really differentiate between NT2s versus the IH patients? I think sometimes they come in to your practice already having been or may be diagnosed, maybe not diagnosed, maybe just saying they were off in sleepy as opposed to really knowing underlying REMS patterns or things like that. How squishy is that label in your opinion? And I guess -- how does it matter if at all, kind of that delineation?

So that's a great question. And I'm happy to talk a lot with you too about it. So the differentiate between NT2 and IH is really complicated. And there's a lot of that go into it. Fundamentally, these are disorders of excessive daytime sleepiness that we don't know the cause of it. Except for that sort of 20% of people with NT2, right, who actually have NT1 who get miscategorized because we usually aren't getting a lumbar puncture on these folks. So when you use the word real world, I have to keep in mind that my practice is probably not the real world, right, because I'm doing something probably a little bit different being at an academic center and thinking about these things pretty deeply. Some of the things that I will do or use things like HLA, DQB10602 testing and people who don't have cataplexy when it's important because we know from the data that if you don't have cataplexy and you don't have this HLA alleles but you're at risk for having narcolepsy in the first place. Then you don't have NT1, right? So you can use some other sort of pieces of data to move people into sort of probabilistic diagnostic categories to put it in a different way. So there's -- in terms of like what the fundamental differences between people with NT2 and IH, it's complicated. And even within the field, there is some discussion about whether or not these diagnostic constructs should be merged at some level, probably people who sleep long amount of time who have IH are probably a little bit different than people who don't sleep long amount of time. So it's a complicated issue to say the least.

I can also chime in. I mean, I think for research purposes, we're always trying to have practical implementations as well. So we use scales like the idiopathic hypersomnia scale, which really has domains of long sleep duration, sleep inertia as well as daytime sleepiness. And the majority, I want to say like 70% of idiopathic hypersomnia patients do have a longer sleep phenotype is really difficulty getting up in the morning. And those seem to be -- they're overlapping, but sort of distinguishing in the magnitude of severity of those symptoms. So that scale is helpful. And then while it might not be as real world, we're trying to implement protocols where essentially that long sleep duration is actually captured on the polysomnogram and then an MSLT is done so we can kind of capture both symptomology. And I think that's clinically feasible for other groups, but they want us to test it out first.

And one real world thing I want to also mention in terms of thinking about some of the data that were presented in terms of like NT2 is a lot of time, people will end up in our clinic who have been seen elsewhere diagnosed with type 2 narcolepsy. Sometimes that diagnosis is made by a well-meaning physician because historically the drugs that always been approved to treat narcolepsy. And people with IH for a lot -- for a long time, we're kind of stuck with no options that were approved to treat them. So sometimes people would say, "Well, you're sleepy enough on the sleep study, you seem like you -- you have narcolepsy, so I'm going to give you a diagnosed of narcolepsy." So I always take all the claims data with a big grain of salt based on some of those sort of where the rubber meets the road of clinical care, I guess, is what I put out the two.

And then just one follow-up. How much does cost of the medicine come into play? You obviously have -- for NT2 or IH patients, you have 2 options or 3 options cheap things like. Adderall and Adderall XR and Provigil that are very inexpensive, if not almost free and largely safe. And then oxybate, which is not free at all but approved. Some of you even said that it's actually easier to get oxybate approved in IH than NT2. Would you agree with those statements? I guess to what extent does it matter the cost of the medicine in either of those 2 buckets?

I mean, in the U.S. system, we're using payers -- insurance payers and the contracts that are going on between pharmacy benefit managers and the companies, and it's something that is more black box and opaque at my level. So yes, we've had that experience of an oxybate is actually easier to get than a stimulant even independent of cost because it's FDA approved and whatever contracts are behind the scenes are happening. So that is true. But for narcolepsy, at least in this state of Massachusetts, payers have different stuff therapy and so oftentimes, you do need to fail like 1 or 2 traditional stimulants, then get to the next class and you get to the next class. Whether that would change with like such robust efficacy data, I think would be helpful. And I don't know where practice parameters will fit into that versus expert consensus.

Monica, I'd love to get your perspective on that from the patient groups. How much does cost play into treatment decisions? And are there significant challenges there that patients have?

It has an impact, a very high impact on what people are able to use for their medication. And a lot of people with narcolepsy are on disability, so they don't have insurance. Some of the pharmaceutical companies do have programs where they assist with the payment of the medication but it's definitely a huge impact. And just for the real-world narcolepsy type 2 and IH, the patients are so confused because they're back and forth, doctors telling them they have IH and narcolepsy type 2. So there's so much confusion inside the patient's head and what's going on in the community.

That's really helpful.

This is Chi for Jason Gerberry at BofA. I have a couple of questions. I guess the first one is what we've heard is that physicians care about the quality of wakefulness as much as the quantity of the wakefulness. I'm curious in practice, is there any scale that you would use to monitor the quality of wakefulness? And is there anything that you will look for either standard type test or a new test that you will hope to see in clinical trial to help inform therapeutic potential of the orexin class in this aspect?

Yes. So I mean, clinically, the most common scale that's going to be used to assess sort of daytime sleepiness is going to the Epworth Sleepiness Scale. Now the challenge for Epworth Sleepiness Sleeping Scale is that, it's a very sort of cursory scale that measures your likelihood to dose under real or imagined situations and 8 possible or like 8 situations and scaled 0 to 3 and then you just add it up. So it's very simple and it has -- it's useful in sleep medicine and it's sort of applicable to a bunch of different disorders. So that's why everybody gets it. The challenge is with these type of disorders a lot of times even with the medication sometimes people can do if they let themselves pretty easily. So even if they're actually saying, I'm actually doing really well. I'm feeling pretty good. They may have an elevated upward scale. So it's -- you have to move way beyond just like basic symptom scales. And I think really get at quality of life measures, I think, is really what's most important based on some things that have already been discussed today and what people experience. So what I think we spent a lot of time really focusing on like how awake and alert are you, whereas, realistically that may not be the thing that really impacts people, it's usually function in those kind of downstream thing.

I mean, I think I'd just add to that, that no, there is no scale that's been talked about that we have four sleep stages, how come we don't have like differentiation of wakefulness in a similar type way. But I think that some of the features of like good wakefulness, if you will, relate to cognition. So it is really encouraging to see new clinical trials implement cognitive outcomes, things relating to alertness and focus, I think, kind of go hand-in-hand with a good quality of wakefulness.

And I've got a question from Charles Duncan at Cantor for the Vibrance-1 and Vibrance-2 trials, how do you see even mild AEs, such as insomnia, dizziness and polycoria impacting effect size. Are you concerned about these potential confounding variables due to the study design or sample size, reducing risk or data interpretation? I am sorry that should have been? Are you not concerned about these confounding variables?

Sorry, what are the concerns?

So the question is, are you concerned that even mild AEs may have functional unblinding on the study and would you expect that to impact effect size?

Well, so I think that there's always the risk of unblinding with any confound that is highly effective, right, especially if people are on placebo. And I think that would go for any potential study design. I mean I think that one of the reasons that things like the MWT are used to is that there's less of a placebo effect with the MWT and some of these other sort of purely self-administered scale. So that's kind of how I would look at it. So that's why the MWT and subjective report are both important.

Okay. Do we have any more questions from in the room? One more?

[ Mark Hetrick ] from Stifel. Just a few questions about the treatment paradigms in your practice. So out of the patients sort of that are diagnosed in your practice, what would you say the percent that gets treated? And then sort of just a piggyback on that. Like what's the refractory group, like what patients tried the drugs and that don't end up working after drop it? And then also just wanted to get an idea, those that are on the therapy. Like what's the sense of urgency to like try something new? They're currently on oxybate or the stimulants and then really like trying to get a sense of the new therapies that are out there? Are they brand loyal going to stay? And then is there any difference across the indications, NT1 and NT2 and IH?

I'll start. I mean I think that -- yes. So there's a lot of excitement for these orexin classes because it's a very specific therapy for the problem that they have because most of our patients, as David mentioned, are on polypharmacy and there is this feeling like the runway is ending. They're on all these medications and they're still at a level 6. So this -- the development of these compounds coming like a really, I think, important time. And I'm sorry, what was the second question?

Yes, sorry. So like out of the patients that are in your practice, like the ones that are diagnosed, what person actually getting treated. So like what's that group that's like. Maybe not on therapy, they tried it didn't work. They had to go off.

Got it. I think -- I want to say like 95% start and stay on therapy. And then it's adding additional therapy. Things like oxybate are probably our highest side effect profile medications. That's the one that comes to mind when people are coming off medications. And that might be for lifestyle reasons that might be from side effects like nocturnal enuresis is a common reason or nocturnal eating. But sometimes, if it's a lifestyle issue, they might want to restart it in a couple of years. But I would say most of them are on pretty stable medication.