Alkermes plc (NASDAQ: ALKS) Q2 2025 Earnings Call dated Jul. 29, 2025
Corporate Participants:
Sandy Coombs — Senior Vice President, Corporate Affairs and Investor Relations
Richard Pops — Chairman and Chief Executive Officer
Blair Jackson — Executive Vice President, Chief Operating Officer
Todd Nichols — Senior Vice President, Chief Commercial Officer
Craig Hopkinson — Executive Vice President, Research & Development and Chief Medical Officer
Marcus Yountz — Vice President of Clinical Development
Analysts:
Paul Matteis — Analyst
Anastasia — Analyst
Andrea Newkirk — Analyst
Umer Raffat — Analyst
Unidentified Participant
Joseph Thome — Analyst
Uy Ear — Analyst
David Amsellem — Analyst
David Hoang — Analyst
Ami Fadia — Analyst
Ashwani Verma — Analyst
Leonid Timashev — Analyst
Jason Gerberry — Analyst
Luke Herrmann — Analyst
Marc Goodman — Analyst
Presentation:
Operator
Greetings and welcome to the Alkermes Second Quarter 2025 Financial Results Conference Call. My name is Rob and I’ll be your operator for today’s call. All participant lines will be placed on mute to prevent background noise. [Operator Instructions] Please note that this conference is being recorded.
I’ll now turn the call over to Sandra Coombs, Senior Vice President of Investor Relations and Corporate Affairs. Sandy, you may now begin.
Sandy Coombs — Senior Vice President, Corporate Affairs and Investor Relations
Good morning. Welcome to the Alkermes PLC conference call to discuss our financial results and business update for the quarter ended June 30th, 2025. With me today are Richard Pops, our CEO; Blair Jackson, our Chief Operating Officer; Todd Nichols, our Chief Commercial Officer; Dr. Craig Hopkinson, our Chief Medical Officer; and Dr. Marcus Yountz, Vice President of Clinical Development.
A slide presentation, along with our press release, related financial tables, and reconciliations of the GAAP to non-GAAP financial measures that we’ll discuss today are available on the Investors section of alkermes.com. We believe the non-GAAP financial results, in conjunction with the GAAP results, are useful in understanding the ongoing economics of our business.
Our discussions during this conference call will include forward-looking statements. Actual results could differ materially from these forward-looking statements. Please see Slide 2 of the accompanying presentation, our press release issued this morning, and our most recent annual and quarterly reports filed with the SEC for important risk factors that could cause our actual results to differ materially from those expressed or implied in the forward-looking statements. We undertake no obligation to update or revise the information provided on this call or in the accompanying presentation as a result of new information or future results or developments.
After our prepared remarks, we’ll open the call for Q&A. And now I’ll turn the call over to Richard for some opening remarks.
Richard Pops — Chairman and Chief Executive Officer
Thank you, Sandy. And good morning, everyone. We had a very successful second quarter. Our commercial and financial performance were strong. And last week, we took another major step forward in our pursuit of a medicine that has the potential to transform the treatment of narcolepsy. Now, midway through 2025, we’re on track to deliver on our key objectives across the business. In commercial, we had planned for strong sequential growth, and we exceeded our expectations in the second quarter. This result was driven by excellent operational execution by a seasoned commercial team.
With sustained profitability now, no debt, and more than a billion dollars of cash, we’re in a strong financial position with significant optionality. It’s clear to us now that the future growth of the business can be accelerated by the development candidates now in our pipeline. Last week, we reported positive topline results from Vibrance-1. This was our first phase 2 study of alixorexton, which you’ve formerly known as ALKS 2680 in narcolepsy type 1. Vibrance-1 was successful and a critical study in the development of our orexin portfolio — obviously for the efficacy and safety data it yields, but also for the operational foundation it provides for the phase 3 program.
Now, if you think back a year ago, we had data from our phase 1b study of alixorexton that suggested robust efficacy and a generally well-tolerated profile based on single-day exposures across a range of doses in small cohorts of patients with narcolepsy type 1, narcolepsy type 2, and idiopathic hypersomnia. These data were critical in defining the initial clinical profile and dosing range for alixorexton. This was step one.
Step two is to confirm and extend these observations in multi-week, multi-dose phase 2 outpatient studies. Vibrance-1 in patients with narcolepsy type 1 is the first of these studies. Now we have randomized placebo-controlled six weeks multi-dose data in hand from more than 90 patients with NT1. We’ve now answered key questions in phase 2 with rigorous assays across larger cohorts of NT1 patients over a longer period of time. These data and insights will be fundamental in preparation for phase 3.
So here are the key findings from the study at the topline. First, the results demonstrate a significant effect on wakefulness and a generally well-tolerated profile. This was our pre-test hypothesis, and the data were in line with our expectations. Second, the study provided entirely new and exciting findings relating to fatigue and cognition. These are among the most disruptive symptoms patients with narcolepsy experience, and they’re distinct from excessive daytime sleepiness. In this study, alixorexton showed robust and clear improvements on validated patient-reported measures.
Our view is that demonstrating effects in these domains establishes a new standard in the development of orexin 2 receptor agonists in narcolepsy. These emerging data also further support our hypothesis that the orexin system can be harnessed to address other neuropsychiatric and neurological conditions. We have two additional orexin candidates that we plan to develop for conditions beyond central order disorders of hypersomnolence. In Q2, we initiated first-in-human studies for one of these candidates, ALKS 4510. We plan to advance the second candidate, ALK 7290, into the clinic later this year.
So, for today, Craig and Marcus will take you through the topline results of Vibrance-1 with significantly more detailed data to be presented at the upcoming World Sleep meeting in September. But first, Blair and Todd will review the financial and commercial performance of the business for the second quarter. And with that, I’ll hand the call over to Blair.
Blair Jackson — Executive Vice President, Chief Operating Officer
Thank you, Rich. Our second quarter financial results were strong and reflected solid commercial and operational execution. We had planned for accelerated growth from the first quarter, and we exceeded our expectations. Financially, the year is progressing nicely, and we remain well-positioned to achieve our financial guidance for the full year, which we reiterated this morning.
For the second quarter, we generated total revenues of $390.7 million. For our portfolio of proprietary products, we generated net sales of $307.2 million, reflecting 14% year-over-year growth. These results were driven by strong underlying demand, which Todd will address in his remarks, and gross-to-net favorability, primarily related to Medicaid utilization rates and certain other credits during the quarter. These factors drove a one-time gross-to-net benefit of approximately $9 million for VIVITROL and approximately $11 million for ARISTADA. Taken together, these gross-to-net dynamics resulted in a proprietary product revenue tailwind of approximately $20 million in Q2.
As we move into the third quarter, we expect Q3 net sales from this portfolio in the range of $280 million to $300 million. Manufacturing and royalty revenues were $83.4 million for the second quarter, including revenues of $39.4 million from VUMERITY and $30.3 million from the long-acting INVEGA products.
Turning to expenses. Cost of goods sold were $49.5 million, which compared favorably to $61.5 million for Q2 last year, primarily reflecting efficiencies following the sale of our Athlone-based manufacturing business last year.
R&D expenses were $77.4 million compared to $59.6 million for Q2 last year, reflecting investments in the Vibrance phase 2 studies of alixorexton across narcolepsy and idiopathic hypersomnia. We expect R&D expense to step up slightly in the second half of the year as we complete our phase 2 studies in narcolepsy and continue to build momentum in our phase 2 study in idiopathic hypersomnia.
SG&A expenses were $170.8 million compared to $168.1 million for Q2 last year. For trending purposes, we expect SG&A expense to be fairly consistent in Q3 and a modest decrease in the fourth quarter of the year. This performance generated strong profitability of GAAP net income of $87.1 million, EBITDA of $101.6 million, and adjusted EBITDA of $126.5 million in the second quarter.
Turning to our balance sheet. We ended the quarter in a strong financial position with $1.05 billion in cash and total investments. We continue to have $200 million of remaining share repurchase authorization, and going forward, we may opportunistically repurchase shares dependent on market conditions and the capital needs of the business.
As we look ahead, based on our performance during the first half of the year and the expected contribution from our expanded sales efforts, we’re on track to deliver record revenues from our portfolio of proprietary products in 2025. As a result of this strong performance, we now anticipate finishing the year towards the higher end of our previously issued financial expectations in terms of both revenue and profitability.
With that, I’ll turn the call to Todd for a review of the proprietary portfolio.
Todd Nichols — Senior Vice President, Chief Commercial Officer
Thank you, Blair. And good morning, everyone. In the second quarter, we recorded net sales from our proprietary product portfolio of $307.2 million, reflecting 14% year-over-year growth. We drove strong end market demand across VIVITROL, ARISTADA, and LYBALVI by executing targeted growth initiatives and delivered strong sequential growth from Q1 to Q2. Due to this demand growth and the gross-to-net favorability during the quarter that Blair outlined, our second quarter proprietary net sales of $307.2 million exceeded the expectations that we provided in May of net sales in the range of $260 million to $280 million.
Starting with VIVITROL, net sales in the second quarter were $121.7 million. VIVITROL performance continues to be driven by growth in the alcohol dependence indication market and our ability to capitalize on highly localized market dynamics in certain states and payer systems. Looking ahead, we continue to expect VIVITROL net sales for the full year 2025 in the range of $440 million to $460 million.
Turning to our psychiatry franchise. The expansion of our psychiatry sales force, completed earlier this year, was an important element of our strategy to maintain a competitive share of voice for LYBALVI and reaccelerate growth for ARISTADA. The early returns from that expansion are encouraging, and we are pleased with our progress to date.
For the ARISTADA product family in the second quarter, net sales were $101.3 million. Leading indicators related to underlying demand were encouraging, with increased prescriber breadth and strong new-to-brand prescriptions during the quarter. For the full year 2025, we continue to expect ARISTADA net sales in the range of $335 million to $355 million.
Turning to LYBALVI. Net sales grew 18% year-over-year to $84.3 million. Underlying TRx growth was 22% year-over-year, driven by new patient starts and prescriber breadth. Gross-to-net adjustments were approximately 29% in the second quarter. We now expect gross-to-net adjustments for the full year will be approximately 30%. For the full year, we continue to expect LYBALVI net sales in the range of $320 million to $340 million. Across the portfolio, we are pleased with our second quarter performance and are focused on maintaining this momentum and driving demand in the second half of the year. With that, I will pass the call to Craig.
Craig Hopkinson — Executive Vice President, Research & Development and Chief Medical Officer
Thank you, Todd. Last week, we announced positive topline results from the Vibrance-1 phase 2 study of alixorexton in patients with narcolepsy type 1. The data further characterized the clinical profile of alixorexton and demonstrated that once daily alixorexton normalized wakefulness and excessive daytime sleepiness scores in highly symptomatic patients with narcolepsy type 1, with a generally well-tolerated profile across all doses tested. These topline results represent the first of a series of data sets that will emerge from the alixorexton phase 2 program.
The data are extensive and break new ground. We are looking forward to presenting the primary and key secondary endpoints related to wakefulness and cataplexy, and the safety and tolerability profile observed in the six week double blind period of the study, in an oral presentation at the upcoming World Sleep meeting at the beginning of September.
In addition to the topline results, we’ll also share data relating to the exploratory patient-reported outcomes collected in Vibrance-1, including the fatigue and cognition data outlined in our topline press release last week. These data are truly exciting not only in terms of the clinical profile for alixorexton, but also as we plan for additional clinical studies across our portfolio of investigational orexin 2 receptor agonists in disorders where impaired cognitive functioning and fatigue are key clinical features.
Following world sleep, we expect topline results from Vibrance-2, our phase 2 study in narcolepsy type 2 in the fall. Enrollment in Vibrance-2 is going well, and we expect to complete that soon. Topline data from Vibrance-3, our phase 2 study in idiopathic hypersomnia, are expected to follow in mid-2026. Each of these studies provides a significant amount of data to analyze and will deepen our understanding of alixorexton’s potential utility across central disorders of hypersomnolence and its differentiating features in the competitive landscape.
In parallel, we are preparing for key regulatory interactions and for the global phase 3 program in narcolepsy that we plan to initiate as rapidly as possible following the topline data from the narcolepsy type 2 study. Alkermes is at the forefront of development in this exciting potential therapeutic category, and the positive Vibrance-1 data represent an important stride forward for alixorexton development program and our broader portfolio of orexin 2 receptor agonists.
With that, I’d like to introduce Dr. Marcus Yountz to review the topline data from the Vibrance-1 study. Marcus is Vice President of Clinical Development and the clinical program lead for alixorexton here at Alkermes. Marcus?
Marcus Yountz — Vice President of Clinical Development
Thank you, Craig. Vibrance-1 is a six-week double-blind placebo-controlled parallel design study evaluating three different doses of alixorexton in patients with narcolepsy type 1 or NT1. The study enrolled a total of 92 patients, with most having moderate to severe disease at baseline. Patients were randomized to one of three once-daily dose levels of alixorexton — 4, 6, or 8 milligrams — or placebo.
The primary endpoint of Vibrance-1 was the change from baseline compared to placebo in the Maintenance of Wakefulness Test, or MWT. MWT is a standardized quantitative measure of how long patients can stay awake during a 40-minute test period when they’re in an environment that is conducive to sleep. These tests are conducted at 2, 4, 6, and 8 hours post-dose. The mean score is calculated by averaging the results of these four tests. While the MWT is less frequently used in real-world clinical settings, it is an important objective endpoint commonly used for regulatory purposes.
In Vibrance-1, alixorexton showed dose-dependent, statistically significant, and clinically meaningful increases in mean sleep latency at all doses tested at week six. Importantly, all dose groups achieved normative wakefulness, applying the standard convention of a mean sleep latency on the MWT of 20 minutes or more. The study also evaluated key secondary endpoints, including change from baseline on the Epworth Sleepiness Scale and weekly cataplexy rates compared to placebo.
First, the Epworth Sleepiness Scale or ESS. Unlike the MWT, this scale is widely used in the clinic as a diagnostic tool to assess for excessive daytime sleepiness. The ESS is a patient-reported symptom questionnaire asking about the patient’s likelihood of falling asleep across eight different scenarios, such as watching TV, riding in a car, or reading a book over the last week. Higher scores indicate a greater likelihood of falling asleep, with a score of 10 and below considered normal. The Epworth Scale is useful in that the seven-day look back period provides a holistic view of patient sleepiness beyond the 8-hour MWT test period.
Here, across all doses tested, alixorexton demonstrated statistically significant and clinically meaningful improvement at week six, with each dose group achieving normative levels. In addition to excessive daytime sleepiness, NT1 patients can experience a sudden, involuntary loss of muscle tone called cataplexy. Vibrance-1 evaluated mean weekly cataplexy rates.
To measure this endpoint, patients are asked to keep diaries of cataplexy events that they experience. The average number of weekly events across weeks five and six in the alixorexton-treated subjects were then compared to those experienced by the placebo group. Across all doses tested, alixorexton showed numerical and clinically meaningful improvements in cataplexy compared with placebo, and on the pre-specified analysis, met the threshold for statistical significance at the 6 milligram dose.
We are confident in the effects of alixorexton on cataplexy and believe there are learnings here related to our implementation of this assay that we will apply in phase 3 to reduce variability and the impact of outliers. We look forward to sharing additional analyses of these data at World Sleep. So while excessive daytime sleepiness is the hallmark symptom of narcolepsy, many patients also experience other symptoms such as fatigue and cognitive dysfunction. These can result in significant morbidity as well as impaired quality of life. Our hypothesis has been given the nature of the neurocircuitry affected, that alixorexton could have an impact on many of the aspects of this disease that affect patients’ day-to-day functioning.
The British Columbia Cognitive Complaints Inventory, or BC-CCI, and the PROMIS Fatigue scales capture two of these common and often debilitating effects of narcolepsy. The BC-CCI evaluates concentration, memory, expressing thoughts, word finding, slow thinking, and difficulty solving problems. The PROMIS Fatigue measures patients’ frequency as well as intensity of fatigue, along with its impact on physical, mental, and social activities. It’s important to note here that fatigue is a symptom that patients experience, which is distinct from sleepiness. While sleepiness is a general feeling of being tired and wanting to sleep, fatigue is a broader feeling of exhaustion that can be long-lasting and may not be resolved by sleep.
We also looked at the Narcolepsy Severity Scale. The NSS captures a holistic assessment of disease by evaluating five key narcolepsy symptoms: excessive daytime sleepiness, cataplexy, hallucinations, sleep paralysis, and disturbed nighttime sleep. And on each of these exploratory patient-reported outcome scales, the BC-CCI, the PROMIS Fatigue, and the NSS alixorexton demonstrated clinically meaningful improvements from baseline compared to placebo that were statistically significant. Of course, the p-values here are nominal due to the exploratory nature of these endpoints.
So from a clinical perspective, these results are compelling due to the robustness and particularly the consistency across all doses of alixorexton as well as across various complementary assays. This is the first time that we’ve seen data from the orexin class on these fatigue and cognition scales, and we believe this differentiates alixorexton from other development programs and builds upon the evidence base that orexin 2 receptor agonists with appropriate pharmaceutical properties could have broad potential utility across a range of neurological or neuropsychiatric disorders.
And now we’ll turn to safety and tolerability. Overall, alixorexton was generally well-tolerated in this study. The majority of the treatment-emergent adverse events were mild to moderate in severity, and no treatment-emergent serious adverse events were seen. The TEAEs that did occur were generally consistent with the events that we observed across the phase 1 studies in healthy volunteers and in subjects with NT1, NT2, and IH. And among the many clinical safety assessments we conducted in the study, two of particular interest are hepatic labs and ophthalmic exams, and importantly, there were no treatment-emergent safety signals seen in these assessments.
So overall, we are very pleased with the safety and efficacy profile thus far, and we look forward to presenting these data sets at World Sleep. I’ll hand it back to you, Rich.
Richard Pops — Chairman and Chief Executive Officer
Well done. Thank you, Marcus. So that’s a summary of the topline findings. There’s a lot more to come, and you’ll begin to see it in a few weeks at World Sleep. These data in NT1 represent a meaningful step forward for the alixorexton development program, and they provide a substantial new data set that significantly expands our understanding of orexin biology, not just relevant to narcolepsy, but its potential across a broad range of neuropsychiatric and neurological disorders.
We’re now moving forward with confidence and a sense of urgency as we prepare for the initiation of our registrational program in narcolepsy. And with clear findings now relating to cognition and fatigue adding to what we’ve seen for excessive daytime sleepiness, we now have further data supporting development of additional orexin candidates in other disease states beyond sleep disorders.
As you’ve heard throughout this call, the business is in a strong position. Our commercial team is on track to deliver proprietary product next sales in excess of a $1 billion and robust profitability in 2025. Our balance sheet is strong and provides strategic optionality with more than a $1 billion in cash. Our pipeline products are advancing. Alixorexton is the first major potential commercial opportunity to emerge from our orexin portfolio. But we also believe that sleep disorders are just the beginning for this exciting new therapeutic category.
So thank you for your patience. With that, I’ll turn the call back to Sandy to run the Q&A.
Sandy Coombs — Senior Vice President, Corporate Affairs and Investor Relations
Great. Thanks, Rich. Rob, we’ll now open the call for Q&A, please.
Questions and Answers:
Operator
Thank you, Sandy. We’ll now be conducting a question-and-answer session. [Operator Instructions] Thank you. And our first question today comes from the line of Paul Matteis with Stifel. Please proceed with your question.
Paul Matteis
Hey, good morning. Thanks so much for taking my question, and congratulations on all the progress. I don’t want to front-run the World Sleep presentation and ask a specific data question, but taking a step back, there’s been a lot of focus, right or wrong, from Wall Street on visual adverse events with the orexin program. And I guess I wanted to ask the Alkermes team, one, do you think the focus on visual AEs is warranted, like, are we on the right track, and kind of thinking about whether this is clinically significant?
And then two, where would you draw the line on a visual AE signal that is benign and might not have much regulatory commercial consequence versus something that might be more significant and might require certain things like driving studies or could result in certain restrictions on a drug label? Thanks so much.
Richard Pops
Good morning, Paul. Hey, let me start, then I’ll hand it over to the experts. But I can just tell you, in my experience dealing both with clinicians and investigators and patient groups and dealing with Wall Street, the focus is almost entirely on the Wall Street side on the visual AEs. And it’s an important contribution that we made in this particular study because, as I’ve said before, there was a reasonable scientific medical question at the beginning of this program about whether orexin 2 receptor agonists can have a direct effect on the eye.
So with this rigorous baseline ophthalmic exam at baseline and then six weeks later, having 90 plus patients’ worth of data to establish now that we saw no changes was an important step forward in that. But I think from a clinical perspective, I’ll let these guys comment on that and what their experience has been, but we won’t provide any more specific AE data on this call other than what’s in the press release. But I think they can give you some qualitative sense of it.
Marcus Yountz
Yeah, exactly. As Rich pointed out, we’re not providing additional AE data, but I think in general we do feel that, based on some of our discussions, that any AE that’s thought to be mild and not interfering with patients’ daily activity is not going to be something that’s going to be overly concerning, both for physicians, or for their patients.
Craig Hopkinson
Yeah. And maybe just to add to that, we had a data safety monitoring board in place overseeing all safety from the Vibrance program. And they’ve met a number of times and given us the green light to proceed. In addition to that, as Richard pointed out, the ophthalmologic exams were normal. We established a baseline. So that gives us some confidence there as well. And then sort of directionally, as we said, as we’ve sort of said in our disclosure, the adverse event profile is in line with what we saw across the phase 1 healthy volunteer and patient cohorts from that phase 1 program.
Richard Pops
Hey, Paul. Let me just add, because it might anticipate some other questions that we’re going to get. Just as a matter of fact, the safety database is actually not even closed until probably mid-August, because recall we have a seven-week extension that follows the six-week double-blind. So, actually, we couldn’t populate data tables with specific numbers until that database is locked. So part of the reason for the level of disclosure at the topline was simply to characterize accurately what we found in the six-week double-blind period. We’ll actually have very specific data for you by the time we come to World Sleep in September.
Paul Matteis
Great. Thanks for all the perspectives.
Richard Pops
You’re welcome.
Operator
Our next question is from the line of Akash Tewari with Jefferies. Please proceed with your question.
Anastasia
Thanks so much for the question. This is Anastasia [Phonetic] on for Akash. So in phase 2, it looked like TAK-861 may have left some efficacy on the table in MWT versus 994. How confident are you that 2680 may be able to fully explore that exposure response range between 4 and 8 mgs in NT1 and differentiate on efficacy versus Takeda?
Richard Pops
Yeah, we’ve always thought that a range of doses would be a competitive advantage. And I’m not going to comment on the competitive programs. I think that one of the features of this program has always been the ability to dose across a wide range in NT1 leading into NT2 and IH as well. So we’ll wait to see the data from the NT2 and IH cohorts, but we’re very pleased with the dose range that we selected for this NT1 study.
Operator
Thank you. The next question is from the line of Andrea Newkirk with Goldman Sachs. Please proceed with your question.
Andrea Newkirk
Good morning. Thanks for taking the question. Rich, I was just wondering if you might be willing to speak a little bit about how you’re thinking about the regulatory path from here, recognizing you do still need to meet with the FDA. But is there the possibility that these Vibrance studies could serve as registrational trials? And if not, would you expect one registration-enabling trial per indication would be sufficient to support approval, or would you need to for each?
Richard Pops
Good morning, Andrea. Thank you for the question. Yeah, I think — okay, the first stipulation will be that FDA is a fairly fluid place right now. But I’m going to answer the question based on what we know coming into things. And that is that — our expectation was to complete the NT2 study. And because we’ll be seeking a label that encompasses narcolepsy writ large, which would encompass NT1 and NT2, we would wait for those data from NT2 before scheduling our formal end of phase 2 meeting with FDA, where we’ll agree on the phase 3 design.
The reason we’re doing that is because we’re currently the only player in NT2 at this late stage. And so that’s a very differentiating part of the product and potentially the label. And as we understand the NT2 doses relative to NT1 doses, then we’ll have the ability to sit down with the FDA and map out the phase 3 program. Our assumption right now is that our phase 3 program will look very similar to the competitors, i.e., a three-month study in NT1 and probably a similar study in NT2, one each. But we would confirm that at our end of phase 2 meeting.
Andrea Newkirk
Okay, thank you.
Operator
The next question is from the line of Umer Raffat with Evercore ISI. Please proceed with your question.
Umer Raffat
Hi, guys. Thanks for taking my question. Just two quick ones. One, could you confirm the dose response is, in fact, linear on MWT? And I asked because there’s been some questions around the cataplexy observation. And my question is, I realize there’s a numerical trend, but it’s not stat sig [Phonetic] at 8 milligrams. Is it reasonable to assume, based on how the data and the variability looked, that you guys tripped the threshold on [Indecipherable] and ended up using negative binomial distribution to drive the p-value? And did that explain partially why p-value was broader for 8 milligrams? Thank you.
Richard Pops
Good morning, Umer. It’s Rich. I’ll — we haven’t talked anything specifically about the linearity or lack thereof of the dose response. You’ll actually see all the by dose information in just a few weeks’ time at World Sleep. Your statistical question on cataplexy is impossible for me to answer, so I’ll pass it to the pros.
Craig Hopkinson
Sure. Yes, I can answer that. So we did use the negative binomial analysis, and that was actually pre-specified for us in conjunction with discussions with the FDA. So we did use that analysis and pre-specify it. And based on our data, that was the appropriate analysis to use.
Umer Raffat
So can I just clarify then, [Indecipherable] as not your base case. It went straight into negative binomial.
Richard Pops
That’s correct.
Craig Hopkinson
That’s correct. Yeah.
Umer Raffat
Wouldn’t that create a discrepancy when we look at p-values for Takeda data set versus Alkermes data set? It wouldn’t be apples to apples on p-value basis?
Marcus Yountz
It’s — yeah, now — it’s a good question. I mean, we’re not comparing apples to apples directly with their data set regardless, obviously, they were different patient populations, so we wouldn’t compare across trials directly. But this is the analysis that we used in a pre-specified manner.
Richard Pops
And I would say, Umer, that’s one statistic that you use as a lens to look at the data. When you go to World Sleep, you’ll see other perspectives on that data that I think very clearly show alixorexton’s effect on cataplexy at these doses. So, for us, we think of it as more of a methodological learning for phase 3, which statistic and which method are we going to apply in phase 3, recognizing we see a very clear cataplexy signal.
Umer Raffat
Got it. And Richard, I’m sorry, since this is so important, I just want to be clear. Numerical trend-wise, you think the data is as competitive as Takeda on cataplexy?
Richard Pops
Well, again, it’s apples and oranges. So what’s — you can make the decision yourself when you see the data itself. But I think we feel quite comfortable that we have a clear cataplexy signal that you’ll see. There’s some aberrant, there’s some outlier data that really confuses this statistic. So you can try to correct for that using various statistical methods, or you can just look at the data, and you’ll see the data in more complete revelation at World Sleep. And I think we can talk about that afterwards. But I think you’d be satisfied that we have a very clear signal on cataplexy.
Just as an aside, without math associated with it, remember, the Narcolepsy Severity Scale picks up cataplexy as one of its key domains. And we’ve normalized patients on the NSS. So all the data tended to work complementary.
Umer Raffat
Thank you.
Richard Pops
You’re welcome.
Operator
Our next question comes from the line of Jessica Fye with JP Morgan. Please proceed with your question.
Unidentified Participant
Hello. This is Adam on for Jess. Thank you for taking our question. I just wanted to ask the alixorexton’s [Phonetic], could you please remind us of the potency selectivity towards the OX2R over the OX1R? Thank you.
Richard Pops
Adam, I thought you were going to ask how to pronounce alixorexton because it’s not the easiest word to pronounce.
Craig Hopkinson
5,000 fold.
Richard Pops
So it’s 5,000-fold more selective.
Unidentified Participant
Great. Thank you.
Operator
Our next question is from the line of Joseph Thome with TD Cowen. Please proceed with your question.
Joseph Thome
Hi, there. Good morning. Congrats on the progress, and thank you for taking my question. Maybe when we do look at the full safety profile of alixorexton, maybe how much of that can be extrapolated to some of your follow-on compounds 4510 and 7290? It looks like there’s obviously some class orexin side effects that we’re seeing. But is there anything different about the targeting or the dosing of 4510, 7290 that you think could result in a different AE profile? Or I guess, how much will the alixorexton and initial data de-risk follow-on? Obviously, we need to see the data, but how are you thinking about translatability there?
Richard Pops
Go ahead, Marcus.
Marcus Yountz
We think — yes, they’re working on the same receptor, so we do think there will be some similarities. That being said, the molecules were purposely developed to have different pharmacokinetic profiles and different structures that could lead them to not all look exactly the same on an AE profile. So, of course, the human data will answer that question for us completely. And we hope to have that for you in the near future on both 4510 and 7290.
Richard Pops
And Joe, if you don’t mind, let me build on your question because to anticipate questions we’ve been getting from investors, and that is how do the NT1 data anticipate what we would see in NT2? And I think I just want to make clear our original hypothesis, which has been confirmed by our own data, which is we believe that there’s actually a frame shift in terms of the tolerability and sensitivity to orexin agonists as you move from NT1 to NT2, meaning you’ll need higher doses to drive efficacy, and you’ll need higher doses to drive the on target side effects as well. So we imagine just the dose response curve shifting to the right in that. And that’s what is consistent with the data we saw in our phase 1b study. We’ll know more definitively, obviously, when we get those data. But that’s the pre-test hypothesis.
Joseph Thome
Perfect. Thank you.
Operator
The next question is from the line of Uy Ear with Mizuho Securities. Please proceed with your questions.
Uy Ear
Hey, guys. Yeah. Thanks for taking our question. Given that the NT2 study, I guess the primary endpoint is complete — expects to complete in August. Just wondering, is there a good chance that you also present the NT2 data at World Sleep? Thanks.
Richard Pops
Uy, just to be clear, so we should complete enrollment in the next couple weeks, which then, with a two-month primary analysis, puts us into the fall. So there won’t be any NT2 data at World Sleep? There will be plenty of NT1 to review, so you won’t be hungry.
Uy Ear
Thanks.
Operator
The next questions are from the line of David Amsellem with Piper Sandler. Please proceed with your question.
David Amsellem
Hey, thanks. So number one, are you planning to build in dosing flexibility or any sort of titration in the phase 3 as a means of minimizing treatment-emergent adverse events? And then secondly, this is a commercial question. It’s not, of course, unheard of to have a wakefulness-promoting agent that does not have cataplexy in the label. I guess my question here is, just given the wakefulness-promoting properties of alixorexton, how important is it to have cataplexy in the label in terms of a commercial adoption perspective? Thank you.
Richard Pops
So, good morning, David. I’ll give you my view, and then I’ll hand it over to the guys. We haven’t made the call yet on the phase 3 dosing because we haven’t finished all the analyses, and we also want to see the NT2 dosing data before we decide on the range of doses. So stay tuned there. We have a — what’s so exciting about running phase 2s of this quality is that now we have 92 patients’ worth of data to model. And our view is that that level of exposure time is essential for actually those dosimetry decisions for the registrational program. So stay tuned on that.
Wakefulness study, my own view, and Todd can answer the question, is that we believe alixorexton is going to have a cataplexy claim in the label because the signal is quite clear. So it’s more a matter of tuning the assay to make sure that we can see that in the data. But Todd, you can comment on whether you think cataplexy is an important part of the presentation.
Todd Nichols
No, I agree. I think you captured it right.
Richard Pops
Okay.
David Amsellem
All right. Thanks.
Operator
The next question is from the line of David Hoang with Deutsche Bank. Please proceed with your question.
David Hoang
Hi. Congrats on the quarter and taking my question. So I just wanted to ask about the upper dose in the NT2 and IH studies. I believe it’s 18 milligrams. Could you just remind us how you landed on 18 specifically, given 25 was used in the earlier phase 1B. And what would be, I guess, the ideal number to take forward into phase 3? Thank you.
Richard Pops
Craig, you want to talk?
Craig Hopkinson
So we employed some sophisticated modeling basically taking into account all the data that we had collected from our healthy volunteer studies as well as our phase 1b program and ultimately, modeled out the doses for phase 2 and that’s how we came to the dose selection.
Operator
Thank you. The next question is from the line of Ami Fadia with Needham & Company. Please proceed with your question.
Ami Fadia
Good morning. Thanks for taking my question. Going back to the focus on visual AEs, can you give us sort of your high-level view on how important is it to avoid a dose that might have such an AE, even if it were to be transient. And would you choose to take a dose into phase 3 that might have seen a transient visual adverse event? And just separately speaking, do you think that that is an on target effect of this class? And as you think about sort of your next gen assets, do you think that it’s just a broadly an on target effect or is it more specific to the structure of a given drug? Thank you.
Richard Pops
Let me start off. Ami, I’m just going to try to pull you back up from the visual AE obsession to the question about AEs, in general. I mean, what we’re looking at in this data set is doses where we know there are on target AEs, namely insomnia and pollakiuria that we and others have observed as on target effects that could be — that would limit your dose. So I think the virtue of this program is we’ve run three different doses for six-week periods of time in the outpatient setting, and we’ll get a complete time course as well as severity map of those AEs.
And we have so much more data to look at together. Because let’s say, for example, if you have a numerical AE, did it happen every day over the 42 doses, or did it happen once in the first week and then went away? There’s a lot of nuance to this, so I don’t think it’s easy to answer the question about what doses we would take forward because I think that there’s a range of different AES that we’re going to be focusing on and we think our competitive advantage is going to be this range of doses because not all patients are the same and people react to different doses. And I think that that would be true for efficacy as well as for AEs. Marcus, do you have your view on it?
Marcus Yountz
Yes. Yeah, I would agree. We’re, as we speak, doing complicated modeling on exposure response, exposure safety analyses, and that’s going to encompass everything we’ve seen. We’re looking at this data in every way we can to help us determine what the best range of doses is going to be to move forward into phase 3. And so that’s going to encompass efficacy, it’s going to encompass safety, and even ease of dosing. And so we’re thinking about all possible criteria when we think about dosing moving forward.
Ami Fadia
Thank you.
Operator
The next question is in the line of Ash Verma with UBS. Please proceed with your question.
Ashwani Verma
Hi. Thanks for taking my question. So just on these, like phase 2 studies, like what time of the day are patients dosed with 2680? I understand it’s down in the outpatient setting except for the MWT days, but just what’s your direction to the patients? And then secondly, on the visual disturbance that you mentioned, so the PR says that you’re not seeing any signal, but is that statement based on AEs that you may have seen in and outside of the scheduled exams? Thanks.
Craig Hopkinson
Sure, I can answer the first part. So we tell them generally to take it in the morning. So roughly around 8 AM, there’s of course a window around that just to accommodate for various lifestyles. But roughly 8 AM in the morning is when we ask them to dose.
Richard Pops
And we didn’t made any specific AE table elements until the phase 1 — I’m sorry, until the phase 2 safety databases closed in mid-August. So you’ll see those data at World Sleep in September.
Ashwani Verma
Thank you.
Operator
The next question is in the line of Leonid Timashev with RBC Capital Markets. Please proceed with your question.
Leonid Timashev
Thanks, guys. I just want to ask on the endpoints, just given what you’ve seen out of Vibrance-1, are you still thinking that MWT is the best way to go forward compared to ESS, and both for, as you think about, I guess the dual design for Vibrance 2 and then the ESS focus for Vibrance 3. And then related to that, given all the secondaries that you hit on, including some of the exploratories, how are you thinking about elevating some of those exploratory endpoints to secondaries or key secondaries for a phase 3 program like cognition or fatigue? And would you try to aim to get those on the label? Thanks.
Craig Hopkinson
So, from our perspective, we believe both MWT and Epworth are really important measures. MWT, obviously, the more objective measure, Epworth really sort of expressing the patient’s sort of subjective assessment. The reason that we elevated Epworth to a dual primary in the NT2 study is because of our sort of thought process around planning for regulatory interactions, really taking a look at the late stage clinical trial landscape and planning for our phase 3 program. So yes, we believe both are important. And I think the data from the NT2 study will help us plan for our phase 3 program with the dual primary endpoints.
Richard Pops
And on the key secondaries. Marcus, you might want to comment on, but from my perspective, these were really exploratory. Some of these scales I don’t think had been used in narcolepsy studies at all. And so we were curious to see whether we would see movement. And boy did we. I mean, we really saw clear signals. And you’ll see more of those data at World Sleep. Given the magnitude of the response and how clear it was, I think we have to do a lot of thinking now about how we might elevate those in the hierarchy. Please comment on that.
Marcus Yountz
Exactly right, Rich. The teams here are working diligently to look at exactly that. All of these various exploratory outcomes, the ones where we succeeded that again were hypotheses that we’ve now shown significant data on. And so the team is really working on how do we pull these in through the phase 3 and elevate them into the key secondary realm in a way that allows us to really test the aspects of this drug more thoroughly.
Operator
Thank you. The next question comes from the line of Jason Gerberry with Bank of America. Please proceed with your question.
Jason Gerberry
Hey, guys. Good morning. Thanks for taking my question. So maybe just wanted to probe a little bit, just your understanding of the relationship of PK and the weekly cataplexy endpoint, and whether or not you feel comfortable enough to rule out QD versus BID approaches conferring any advantage in terms of the WCR measurement. And then also, just I know you’re not commenting on phase 3 dosing plans yet, wanting to see the NT2 results first. But I’m wondering if we can take from your comments that at least in the context of NT1 that there’s at least enough confidence in the tox profile of the high 8 mg dose that you could potentially push dose in future NT1 pivotal. Thanks.
Richard Pops
Hey. You guys want to talk about the PK WCR relationship?
Marcus Yountz
Sure. I mean, at this point, those analyses are ongoing. So we’ve reported, as you know, sort of the topline results, and we’re really digging deeper into the PK again, as I mentioned, sort of exposure response, exposure safety. So that’s going to be coming soon, and potentially something you’ll see in the future.
Richard Pops
But is it fair to say, Mark, I don’t think there’s a QD versus BID difference that we think is meaningful.
Marcus Yountz
We don’t. No, that’s right.
Richard Pops
And with respect to the phase 3 dosing plans.
Marcus Yountz
Yeah, I mean, at this point for phase 3 dosing, again, we’re still making those determinations, so it’s hard to speak to them. We do have confidence fully in our tox profile at 8 milligrams, so there’s no concern there. And we’re obviously taking that into account, as well as everything else we’ve talked about to determine our phase 3 doses.
Jason Gerberry
Got it. Thank you.
Operator
The next question is from the line of Luke Herrmann with Baird. Please proceed with your question.
Sandy Coombs
Luke, do we have you?
Luke Herrmann
Oh, sorry, I was on mute. Sorry. Just a quick one for me. It looks like some nice step-ups in revenues across the Board. Can you just talk about the relative contribution from inventory or seasonal dynamics as compared to underlying demand for the commercial portfolio? Thank you.
Blair Jackson
Yeah. This is Blair. I’ll start with on the number basis, and I’ll turn it over to Todd to get in some detail. I think number one, as you look at our proprietary programs, we saw demand increase across all three programs over the time period. We actually didn’t have any inventory dynamics that contributed to this quarter. And so this is in line with normal seasonal dynamics and a contribution of our strong commercial performance. Todd, anything you want to add?
Todd Nichols
Yeah, I would just agree with what Blair said. We saw strong demand for all three products, ARISTADA, VIVITROL, and LYBALVI. And we were really pleased with Q2 was really the step up in new patient starts. As we said earlier, exceed our expectations. To Blair’s point, nothing to look at with inventory. Inventory is just growing with demand right now. So we’re pleased with that.
Luke Herrmann
Great. Thank you.
Operator
Thank you. Our last question comes from the line of Marc Goodman with Leerink Partners. Please proceed with your question.
Marc Goodman
Yeah, just back on this cataplexy endpoint discussion. Like when you talked about the learnings for the phase 3, are these learnings basically just the statistical methods that you were talking about, or are you talking about using different assays? I’m a little confused. Maybe you can shed a little more light there. I mean, we’re all still a little confused. What happened with the cataplexy data, and we’re wondering, are we even going to learn? Like, once we see the data in Singapore, we’re going to feel much better about it. I mean, because you were talking about statistical significance of what needs to get on the label. Like doesn’t it need to be stat sig to get on the label? I mean, it’s a pretty important part of the story, right? So maybe you could just give us a flavor for that?
And then, just since I’m the last question, I’ll ask as well. The working assumption is that insomnia is only seen at the very beginning. So, should we still have that assumption that any type of insomnia is still only seen in week one? Thanks.
Richard Pops
Go ahead, Marcus on cataplexy. Let’s clear that up because it’s important and I think we have a really clear picture of this, Mark.
Marcus Yountz
Sure we do. We do think the cataplexy results had a lot of what we saw likely had to do with outliers that we saw in the Vibrance-1 study. And we think some of that could have to do with the operational implementation of the assay. So, particularly things like standardization of how patients are recording cataplexy at sites and even across sites in what is a global study. So these are the types of things that, for phase 3, we’re going to apply in attempts to reduce this variability and minimize any outliers that we might see.
Richard Pops
And to your point, of course, you’ll need to see statistically significant results for an inclusion in the label, and that’s — that would be our objective. And when you see the data at World Sleep, I think you can draw your own conclusions about the strength of the signal on cataplexy, which shouldn’t be surprising given all these other endpoints are moving so strongly toward normal. You wouldn’t expect to see a whole lot of discordant data in a cataplexy. But the statistic itself, as you saw in the topline release significant at one dose and not at the other two doses, I think you’ll get a little bit more understanding about why that might have happened when you see the variability. And we’re not going to comment on the time course of the various AEs, you’ll see more of that at World Sleep as well.
Operator
Thank you. At this time, we’ve reached the end of the question-and-answer session. And I’ll hand the floor back to management for closing remarks.
Sandy Coombs
Great. Thanks, everyone, for joining us on the call today. Please don’t hesitate to reach out to the company if you have any follow-up questions. Thank you.
Operator
[Operator Closing Remarks]