Greetings. Welcome to the Alkermes conference call to discuss ALKS 3831 ENLIGHTEN-2 data. My name is Rob. I'll be your operator for today's call. If anyone should require operator assistance during the conference, please press star zero on your telephone keypad. Please note that this conference is being recorded. I'll now turn the call over to Sandy Coombs, Vice President of Investor Relations. Sandy, you may begin.
Thank you. Welcome to the Alkermes plc conference call to discuss ALKS 3831, our investigational agent for the treatment of schizophrenia. Last fall, we announced top line data from ENLIGHTEN-2, and today we are presenting detailed results of the study from the Congress of the Schizophrenia International Research Society in Orlando. With me today are Richard Pops, our CEO, Craig Hopkinson, our Chief Medical Officer and Senior Vice President of Medicines Development and Medical Affairs, and special guest, Dr. René Kahn. Dr. Kahn is the Chair of the Department of Psychiatry at the Icahn School of Medicine at Mount Sinai. Please note that during today's call, we will reference slides that are available on the webcast. A PDF of the slides will be made available on our website following the conclusion of the call.
If you have not done so already, please go to the investor section of our website, alkermes.com, to access the webcast player. We will be making forward-looking statements based on our current expectations relating to, among other things, the future clinical development of ALKS 3831, its therapeutic value and commercial potential, and our regulatory filing strategy and timelines. These forward-looking statements are neither promises nor guarantees and are subject to a high degree of uncertainty and risk. Please see our press release issued earlier this week and our SEC filings for important risk factors that could cause our actual performance and results to differ materially from those expressed or implied in the forward-looking statements. We undertake no obligation to update or revise the statements provided on this call as a result of new information or future results or developments. After our remarks, we'll open the call for Q&A.
I'd like them to turn the call over to Craig.
Thank you, Sandy, and good morning, everyone. We're here in Orlando at the Congress of the Schizophrenia International Research Society, or SIRS, to present data on both of our schizophrenia programs, ALKS 3831 and ARISTADA. Alkermes has been deeply involved in the area of serious mental illness for many years now, and we have a growing leadership position in this space. Before we get to ALKS 3831, I'd like to spend a brief moment on ARISTADA. Earlier this week, we announced positive top-line results from ALPINE, first of its kind, 6-month study evaluating the efficacy, safety, and tolerability of ARISTADA and INVEGA SUSTENNA when used to initiate patients experiencing an acute exacerbation of schizophrenia in the hospital and maintain treatment in an outpatient setting.
These preliminary data were also presented yesterday here at SIRS and represent another data set in the growing clinical body of evidence supporting ARISTADA as an increasingly important treatment option for schizophrenia. We're excited about these results and look forward to sharing more details from the ALPINE study in the coming months. In the interest of time today, I will focus on ALKS 3831 and the results from ENLIGHTEN-2 pivotal weight study, as well as our interim data from the ongoing long-term safety extension study. I'll turn it over to Dr. René Kahn, who will share his perspective on the schizophrenia treatment landscape. Let's get started. Schizophrenia is a disabling disease that affects an estimated 1% of the population or roughly 3.5 million people in the U.S. and generally emerges late in adolescence and early adulthood.
This devastating disease is characterized by symptoms that include hallucinations, delusions, disorganized thinking and speech, blunted emotions, as well as cognitive impairments. Given the progressive neurodegenerative nature of schizophrenia, patients need effective, well-tolerated medicines to help control their symptoms and prevent relapse. While we know long-acting injectable antipsychotics are important treatment options that have proven long-term benefits, the vast majority of patients still use oral medications. olanzapine has played an important role in the treatment paradigm of schizophrenia since its approval in 1996. Known for its efficacy, olanzapine offers other clinical attributes, including no required titration and a rapid onset of antipsychotic efficacy that make it an important treatment option for healthcare providers and patients. These characteristics make it a leading choice for quickly stabilizing acutely ill patients.
Interestingly, in meta-analyses, olanzapine did not cause significantly more extrapyramidal effects compared to placebo and was associated with a smaller increase in prolactin compared to risperidone and paliperidone, reinforcing olanzapine's reputation for being generally well-tolerated outside of its significant weight and metabolic liabilities. This profile was underscored by data collected in the NIH-sponsored CATIE study, one of the largest studies ever conducted in schizophrenia, where five antipsychotics, including olanzapine, were studied in a head-to-head manner in nearly 1,500 patients. What you see on the left side of this slide is the rate of discontinuations of the antipsychotics based on lack of efficacy. olanzapine, shown in the red line at the top of the graph, stands out from the pack with the lowest rate of discontinuation due to lack of efficacy. At its peak, olanzapine was the most widely used antipsychotic in the country.
As its profound effect on weight and long-term metabolic liabilities became more broadly understood, regulators and the treatment community were forced to reconsider its use in the treatment paradigm and more restrictive guidelines were put in place. Today, olanzapine remains an important treatment option for patients and clinicians because of its efficacy, but it has dramatically decreased over the years in favor of more weight-sparing antipsychotics. This leads me to ALKS 3831. ALKS 3831 is an investigational novel oral antipsychotic drug candidate designed to provide the robust efficacy of olanzapine while mitigating its weight gain and many of the related long-term metabolic consequences. It's a once daily bilayer tablet composed of olanzapine and samidorphan, our proprietary opioid receptor antagonist.
The efficacy, safety, and weight profile of ALKS 3831 has now been demonstrated across multiple studies and patient populations, including two large phase III studies, and more than 1,600 subjects have had exposure to 3831. A broad spectrum agent in the brain, olanzapine's efficacy is thought to be mediated primarily through a combination of dopamine and serotonin receptor antagonism. Samidorphan is our new molecular entity that functions as an opioid receptor antagonist and works through three prominent opioid receptors, mu, kappa, and delta. Samidorphan has a unique binding pharmacokinetic and activity profile compared with other opioid antagonists. Our initial hypothesis was to leverage combination pharmacology of these two agents to retain the antipsychotic efficacy provided by olanzapine and see whether the addition of an opioid antagonist would affect the reward system, which is known to play a role in mediating food intake.
By doing so, we aim to harness the antipsychotic properties of olanzapine while mitigating the deleterious effects of weight gain and associated long-term metabolic sequelae. The registration program for ALKS 3831 that was agreed upon by the FDA is comprised of two foundational phase III studies, ENLIGHTEN-1 and ENLIGHTEN-2. ENLIGHTEN-1 was designed to test the antipsychotic efficacy of ALKS 3831 versus placebo in acutely ill patients. An olanzapine arm was included as an active comparator. The stage 3 study was successfully completed in June of 2017 and subsequently presented at the 2017 ACNP meeting. ENLIGHTEN-1 was critical to the registration program for ALKS 3831 as it provided a clear demonstration that 3831 retains the robust antipsychotic efficacy of olanzapine. The second key study in our registration program, ENLIGHTEN-2, was a six-month head-to-head comparison of weight gain for 3831 versus olanzapine.
We completed this study last fall, achieving statistical significance on the two co-primary endpoints. Before diving into the results of ENLIGHTEN-2, I'd like to spend a moment on our rationale for the study design. This figure is a model of how we predicted the distribution of weight gain for patients receiving olanzapine might look in a six-month study based on CATIE as well as our own 300 patient phase II study. The histogram captures the magnitude of weight gain on the x-axis and the percentage of patients on the y-axis. Two specific measurements matter. The first is the center dotted line, which represents the average or mean weight gain across the whole population. A continuous variable, the mean captures information across all patients in the study.
The second measurement is the shaded area on the right, which measures just the proportion of patients who gain more than 10% of their body weight. By focusing on both these measures, we aim to capture a sense of the entire distribution. Our hypothesis for ENLIGHTEN-2 was simple. ALKS 3831 would shift the overall distribution to the left. You can see how we arrived at the two co-primary endpoints for the study. One focused on the population mean change in weight and the other on the proportion of patients gaining an excessive amount of weight at a level pre-specified at 10%. By shifting the entire distribution, you would favorably affect both of these measurements. Importantly, this has key potential clinical implications for patients.
You can see here that a shift in the distribution to the left means that on the right side of the figure, fewer patients gain clinically significant amounts of weight, and on the left side, there are more patients who lose weight or gain insignificant amounts of weight. This theoretical model informed our design for ENLIGHTEN-2, which I'd like to walk you through now. ENLIGHTEN-2 was a multi-center, double-blind, randomized, phase III study that evaluated the weight gain profile of ALKS 3831 compared to olanzapine over six months in patients with stable schizophrenia. The study's two co-primary endpoints and statistical analysis plan were developed in collaboration with the FDA. In order to reach a positive study result, both of the co-primary endpoints had to achieve a P value of less than 0.05.
Subjects who completed ENLIGHTEN-2 were eligible to roll over into a 52-week extension safety study, and I'll speak to that later on this call. The baseline characteristics and subject disposition for ENLIGHTEN-2 were similar for both ALKS 3831 and the olanzapine treatment groups, as shown here on slide 16. A total of 561 patients with stable schizophrenia were randomized in the study, and 550 patients who received at least one dose of study drug were included in the safety population. We know that gender, race, and BMI are characteristics known to be associated with weight gain, and as you can see here that these covariates were well-balanced between the two groups. At the bottom of this slide, we list the completion rates for patients.
For ENLIGHTEN-2, we knew that the retention throughout the six months duration, and particularly keeping those patients experiencing substantial weight gain, would be of critical importance in order to capture the separation between olanzapine and 3831. We modeled for a discontinuation rate of 40% in ENLIGHTEN-2, which is typical for schizophrenia trials of this duration. Our overall discontinuation rates were slightly lower than those expectations and similar between arms. Treatment of missing data is an important consideration in clinical trials. Historical data tells us that olanzapine-associated weight gain is underrepresented in controlled settings, and early discontinuations have an important effect. While there are a number of statistical methodologies to address missing data, we aligned with the FDA that multiple imputation was the most appropriate method for the co-primary weight-related endpoints in ENLIGHTEN-2. Moving to the results of ENLIGHTEN-2.
Here are the data which demonstrate that ALKS 3831 successfully mitigated olanzapine-associated weight gain. The two co-primary endpoints met statistical significance at a p-value of 0.003. We had pre-specified a key secondary endpoint, the proportion of patients reporting weight gain of 7% or more from baseline, and that hit as well with a p-value of 0.001. Importantly, this mitigation of weight gain was maintained when we conducted our subgroup analyses based on race, gender, age, and early weight gainers. Expressed in terms of odds ratio, the olanzapine group had two times the risk of gaining clinically meaningful weight in the categorical 10% and 7% cuts compared to the 3831 group. Going now a level deeper into these data, let's look at how 3831 impacted the overall distribution of the weight curve.
Here on slide 19 is the right side of the distribution or the proportion of patients who gained meaningful weight gain. In addition to our primary and key secondary endpoints of 10% and 7% weight gain, we also observed a clear differentiation from olanzapine at the 5%, 15% and 20% weight gain thresholds. At each of these categorical cuts, ALKS 3831 consistently demonstrated fewer patients with these levels of weight gain than olanzapine. Let's examine the left side of the distribution curve. Here we see that more patients on 3831 lost weight as compared to olanzapine. Together, these data reinforce the success of our design intent for ENLIGHTEN-2, a clinically meaningful shift in weight distribution, creating a distinctive profile for ALKS 3831. Shifting the weight distribution for ALKS 3831 was a necessary prerequisite to achieve success on the primary endpoints of ENLIGHTEN-2.
Clinicians and patients are also concerned with olanzapine's weight gain trajectory, which is well known to persist over many months and potentially years before subsiding. It was important for us that we replicate the flat profile demonstrated in our large phase II study for 3831. As you can see here, we did just that. This is the longitudinal plot of weight gain in the two groups over the six-month period. There are a number of important conclusions from this plot. First, the curves begin to separate early and reach statistical separation by week six. Second, as the study progressed, you see continued divergence between the two arms. Perhaps most importantly, the 3831 weight curve stabilized at week six and remained flat for the rest of the six-month treatment period, in contrast to the ascending olanzapine weight curve.
Another way to capture the continued divergence between ALKS 3831 and olanzapine is to plot the mean difference in body weight over the six-month study. You can see here that the line continues to creep down, reflecting the growing difference between the two treatment groups. As important to us as the results from patients who completed ENLIGHTEN-2 are the profiles of the patients who did not. Here are side-by-side depictions of the subjects who discontinued early by treatment group. The solid blue and green lines denote the weight gain curve of patients who completed the entire six-month study for olanzapine and 3831. While the dashed lines denote mean weight gain curves for those patients who discontinued prematurely. The numbers by each line reflect the number of subjects who discontinued at each week.
For example, in the olanzapine chart on the left, eight subjects dropped out of the study at week 4. The dotted line reflects the mean weight gain for those eight subjects. You can see despite similar overall retention rates between the treatment groups, the weight gain trajectories are different for the patients who discontinued olanzapine versus ALKS 3831. The majority of those patients in the olanzapine group who discontinued prematurely gained a greater percentage of baseline weight than olanzapine completers consistent with historical data that suggests that olanzapine-associated weight gain is typically underrepresented in controlled settings. This pattern of discontinuations is reflective of what you'd expect to see in the real world. Patients who gain a lot of weight stop taking their medications. In contrast, you see a different pattern on the right-hand side for 3831.
This tells us that the observed treatment differences between olanzapine and 3831 may have been underestimated in the study. The bias that these discontinuations introduce further underscores the significance of the separation that was observed in ENLIGHTEN-2. In summary, the weight profile of ALKS 3831 demonstrated in ENLIGHTEN-2 was consistent with our expectations. ALKS 3831 mitigated olanzapine-associated weight gain with mean weight stabilizing at week 6 and remaining flat thereafter. For the first time, we showed you that ALKS 3831 clearly shifted the weight distribution curve as compared to olanzapine with fewer patients on 3831 who gained significant amounts of weight. Lastly, we showed you the weight trajectories of those patients who discontinued the study early.
While we are extremely pleased with the positive results from ENLIGHTEN-2, those discontinuation plots are an important clinical reminder that the observed treatment differences between olanzapine and 3831 may have been underestimated. Let's turn now to the metabolic data from ENLIGHTEN-2. Presented here on slide 26 are the combined metabolic parameters observed across the 6-month treatment period. Parameters include HDL, LDL, total cholesterol, triglycerides, HbA1c, and glucose. The top panel shows these parameters for ALKS 3831 and the bottom panel for olanzapine. The observed changes in these parameters from baseline were generally small, with similar changes for both treatment groups. In this study, we saw that LDL and total cholesterol, shown in light blue, remained essentially flat for both treatment arms and showed no clinically meaningful change over the 6 months.
HDL, shown in dark blue, went down approximately 6 milligrams per deciliter for both groups and then remained stable. This effect on HDL may be olanzapine induced and a parameter that samidorphan may not impact. For triglycerides, shown in gray, we saw a small increase for both groups of just under 30 milligrams per deciliter. When we look at the 52-week extension study in a few moments, we will show you that this acute phase phenomenon stabilized and then returned towards baseline. For glucose, we saw small changes from baseline over the 6 months treatment period for both olanzapine and ALKS 3831, while HbA1c, which is considered a clinically superior measurement of 6-week glycemic control, remained completely flat in both groups. The small changes observed in ENLIGHTEN-2 were not surprising to us given the study of this length.
The literature demonstrates that the impact of olanzapine on lipids and HbA1c are variable from study to study, sometimes showing noteworthy changes in these metabolic parameters and other times demonstrating little change from baseline. In real-world practice, olanzapine's long-term metabolic disturbances are likely driven by weight gain associated with its long-term exposure. As we talk to the differentiated weight and metabolic profile for ALKS 3831 compared to olanzapine, it's important to remember that it's not only absolute weight gain that matters, but where the weight accumulates. Olanzapine is known to lead not just to excessive weight gain in some patients, but weight gain highly concentrated around the abdomen. This is referred to as central or visceral fat and is generally stored in the abdominal cavity around a number of important internal organs such as the liver, pancreas, and intestines.
Carrying a high amount of visceral fat is associated with a number of long-term health problems and is a leading indicator for developing metabolic complications such as cardiovascular disease and diabetes. See how the results for the change in patients' waist circumference from baseline measured in centimeters. Change in waist circumference is a well-established metabolic parameter and is an indicator of metabolic risk as denoted by the guidelines from the American Diabetes Association. In ENLIGHTEN-2, ALKS 3831 demonstrated a clinically meaningful and statistically significant difference from olanzapine very early in the study, with continued separation for the entire duration of the six-month treatment period. Notably, the separation in waist circumference for ALKS 3831 versus olanzapine occurs earlier than the separation in weight curves. Shown another way, the mean difference from baseline and waist circumference between olanzapine and ALKS 3831 continued to increase over time.
Overall, the changes in key metabolic parameters observed in the six-month ENLIGHTEN-2 study were small and remained stable. ALKS 3831 demonstrated an early and significant impact on waist circumference, which occurred earlier than the observed shift in weight at week six. The last data I'll review for ENLIGHTEN-2 is antipsychotic efficacy as measured by the Positive and Negative Syndrome Scale, or PANSS total score. While the study's co-primary endpoints are focused on weight, we also measured efficacy as an exploratory endpoint because at the end of the day, efficacy is what will differentiate ALKS 3831 in today's treatment landscape. Here on slide 32, you see that both ALKS 3831 and olanzapine demonstrated and maintained improvements in PANSS total scores over the six-month study, underscoring the powerful efficacy of these agents.
To put these data into context, patients in both groups entered the study with stable disease, however, still moderately ill as based on the entry PANSS scores of approximately 70. By the end of the study, both groups had demonstrated meaningful improvements in their symptoms with PANSS scores considered to be mildly ill. The continued and sustained improvements in symptoms in the stable population is important as it reinforces the potent efficacy of olanzapine and ALKS 3831 over a six-month period. The most common adverse events reported in the ALKS 3831 treatment group were weight gain, somnolence, and dry mouth, and the most common adverse events reported in the olanzapine group were weight gain, somnolence, and increased appetite. Now let's take a look at the ongoing open label extension study.
We're excited to share the interim results with you as you begin to see how ALKS 3831 performs over the long term. As a reminder, all patients who completed ENLIGHTEN-2 were eligible to enroll in the open label 52-week extension study. 76% of subjects who completed ENLIGHTEN-2 chose to do so. The subjects who were randomized to 3831 in ENLIGHTEN-2 remained on 3831 in the extension period, and those who'd received olanzapine were switched to 3831. Let's take a look at what happened to the weight and long-term exposure to 3831. Here on slide 37, you see the weight curves of patients in the extension study normalized by each group's baseline. The green line represents the weight curve for those patients who continued on 3831. Their weight remained stable for the duration of the 52-week extension.
Shown in blue are the patients who switched from olanzapine to 3831. Their previously ascending weight curve stabilized and remained flat during the extension period. This is in contrast to what we know about the long-term olanzapine use, where weight gain is known to continue for many months and years. The stabilization of mean weight for the newly switched patients and the durability of effect for the 3831 continuation patients replicates the findings from our phase II study. Here are the weight results for all subjects in the extension period, regardless of previous treatment. ALKS 3831 demonstrates durability of effect over the 52-week period. Since the extension was a single arm, open label study and all patients received 3831, it is more appropriate to show you the combined data. I will focus on all subjects for the remaining slides.
Next, I will turn to the interim metabolic results, where all parameters remained stable with long-term treatment with ALKS 3831. On slide 40 is a graph showing that metabolic parameters in the extension period remained remarkably stable with no clinically meaningful changes observed. The previously observed bump in triglycerides reverted, suggesting the previous variability may have been an acute effect that is not sustained in the long term. The decrease in HDL observed in ENLIGHTEN-2 was maintained but remained stable throughout the 52-week extension, reinforcing a hypothesis that HDL may be one parameter where samidorphan does not mitigate olanzapine-induced effects. At week 52, the mean change in fasting glucose for patients in extensions was about 2.5 milligrams less per deciliter, and the change in HbA1c was 0.07%. Slide 41 shows the change from baseline in waist circumference over 52 weeks.
The line remains flat, demonstrating ALKS 3831's durable stabilization of central fat accumulation. Finally, let's look at the antipsychotic efficacy, which was sustained by ALKS 3831 in the extension. After completing ENLIGHTEN-2, patients rolled over into the extension with a mean PANSS score of approximately 59. Durability of antipsychotic efficacy was maintained, with patients reaching a mean PANSS score of 57 at week 52, representing a mildly ill, well-controlled patient. You have now seen detailed results from ENLIGHTEN-2 phase III study and interim results from the ongoing 52-week extension. ALKS 3831 mitigated olanzapine-associated weight gain, demonstrating clear separation at week six, which continued to diverge through week 24 and remained flat for up to 76 weeks of treatment. The data also showed how ALKS 3831 shifted the weight distribution curve to the left compared to olanzapine, resulting in fewer patients who gained clinically meaningful amounts of weight.
ALKS 3831 demonstrated small and non-clinically meaningful changes in metabolic parameters, which remained stable with long-term treatment. The exception here was for waist circumference. ALKS 3831 demonstrated an early and significant impact on mitigating olanzapine-associated increases to waist circumference, which may represent a long-term metabolic benefit. Finally, the antipsychotic efficacy of ALKS 3831 should not be overlooked. Both ALKS 3831 and olanzapine demonstrated significant reductions in psychotic symptoms, and this efficacy was sustained for patients throughout the 52-week extension period. ALKS 3831 performed exactly as we expected it to, demonstrating a favorable impact on weight and clinically meaningful antipsychotic efficacy. With the ALKS 3831 registration program complete, the team is busy preparing the New Drug Application. We expect to have a pre-NDA meeting with the FDA in the coming months to ensure alignment and are on track to submit the NDA later this year.
We're very excited about the profile of ALKS 3831 and its potential to help address the unmet needs of people living with schizophrenia. Now I'd like to introduce Dr. René S. Kahn, who will provide his insights on the schizophrenia treatment landscape and the clinical relevance of these data. Dr. Kahn is Professor of Psychiatry at the Icahn School of Medicine at Mount Sinai with more than 30 years experience in clinical research. We're pleased to have him here with us this morning. Dr. Kahn.
Thank you very much, Craig. I'm happy to say a few things about schizophrenia in general and about this particular drug and the treatment of schizophrenia. As Craig mentioned, I've been working in the field of schizophrenia for the last 30 years, and actually have conducted quite a few large clinical trials, which one of the drugs that we used was olanzapine, so I have both clinical and actually scientific experience in the field. Just to give you the context, schizophrenia is, as you probably know, quite a prevalent disease, as Craig said, with a lifetime risk of 1%. But to, again, put it in perspective, it's more prevalent than Parkinson's disease, type 1 diabetes, and multiple sclerosis. And different from many other illnesses, it starts very early in life. It starts late adolescence, in the very productive period of people's lives.
People very often don't get married, don't get a job, have to discontinue education. It's a very costly disease for society and obviously also very costly from a personal perspective. That's important to realize. That may just sound 1%, but it's 1% throughout the lifetime, which is very different, for instance, compared to Alzheimer's disease, which is more prevalent but only starts when you're in your 80s, usually. Olanzapine is an extremely effective drug. In fact, we have used it. Craig presented the CATIE data, and we had an equivalent study, more or less, in first-episode schizophrenia called the EUFEST study. We found the exact same thing, that the patients stayed longest on olanzapine because of its efficacy. And that's particularly important in first-episode schizophrenia because that's when you want to have an effective treatment from the get-go.
Also from all the meta-analyses that have been done, the most effective drug is clozapine, but immediately followed by olanzapine. There's no question why this drug was so widely used when it was introduced to the market because it is indeed one of the most effective antipsychotics that we have around. Unfortunately, as you know, it became clear that patients gained considerable weight, especially as Craig has also said, weight distribution was extremely unfavorable with the increase around the waist, with all the subsequent consequences for cardiovascular disease and cancer. Unfortunately, more or less, we've lost one of our most effective antipsychotics because of the side effects. I think this particular combination with samidorphan has now provided or provides the opportunity to kind of reintroduce olanzapine.
What's important, as you've just heard, that the efficacy is, which is not surprisingly, is the same for the combination drug, the Alkermes drug, and olanzapine alone, because it's obviously the same active compound that's in there, but it's important that samidorphan doesn't detract from its efficacy. We have a very effective antipsychotic without the potential of excessive weight gain. I think there is a real need for that. I know from clinical practice that olanzapine is hardly being prescribed anymore, that doctors go for the newer antipsychotics, which as meta-analyses show, have not been shown as effective as olanzapine and clozapine, for instance, or use some of the other antipsychotics like risperidone, which has also probably less efficacy according to the meta-analyses and the other studies, and also has a lot of side effects like sexual side effects.
There is, I think, a real need for a drug like this. Just to emphasize a few things that I think are particularly important. One was slide 23 that Craig has shown is the early dropouts. Why is that so important? It's never caught in clinical trials because these are the patients that don't complete the study. In fact, this is one of the main clinical impediments of olanzapine that many patients drop out early on because of their weight gain, especially women, but men also. Therefore, the patients don't get the opportunity to be treated with the most effective drug.
The other thing I want to emphasize is that there is a separation between the Alkermes drug and olanzapine in weight gain at week six, but then there is a stabilization on the Alkermes drug and olanzapine, as we know, and that's also shown in the first six months, is actually still increasing in weight. That's exactly what we know about olanzapine, that the weight gain continues to accumulate. That's very important in the 52-week extension of the ENLIGHTEN study that in fact there's no additional weight gain. It's important to know that although there is some weight gain after six weeks, it stabilizes, and then for the rest of the year and a half follow-up almost, there's no additional weight gain.
My conclusion is that this really is a great opportunity to kind of reintroduce one of the most effective antipsychotics that we have, that we know are available, and so that patients are able to be treated with it again so doctors will again prescribe it, which there really is a need for a drug like this for the greater potential for the treatment of patients. Thank you very much.
Great. I'll now turn it back to the operator so we can open the call for Q&A. Thank you, Dr. Kahn.
Thank you. We'll now be conducting the question and answer session. If you'd like to ask a question today, please press star one on your telephone keypad and a confirmation tone will indicate your line is in the question queue. You may press star two if you'd like to remove your question from the queue. For participants using speaker equipment, it may be necessary to pick up your handset before pressing the star keys. Thank you. Our first question comes from the line of Cory Kasimov with JPMorgan.
Hey, guys. This is Matthew on for Cory, and thanks for taking my question. The first one is on the weight gain chart on slide 23. Apologies if I missed this in the prepared remarks, but what did you observe with the 13 olanzapine-treated patients that lost weight but discontinued treatment nonetheless prior to four weeks? Any thoughts on why this wasn't seen in the 3831 arm?
I'm just going to get to slide 23 here.
I'm sorry. Could you please repeat the question?
On slide 23, that looks like there's 13 patients in the olanzapine arm that lost weight but discontinued treatment prior to four weeks.
Yes.
Just curious to get your thought on this patient subgroup and why we didn't observe this in the 3831 arm.
I think it's really important to emphasize that these are invariably patients who came off a study not for reasons of weight gain, but for other reasons. As you can see across the entire profile of the olanzapine discontinuers, these patients had substantial amounts of weight gain in addition to their reasons for coming off a study. For those 13 patients, the patients that discontinued early, it could have been for a number of different reasons outside of weight gain. It may have been very variable amongst the individual patients in that subgroup of 13.
Got it. A question for Dr. Kahn. Out of all of the efficacy measures provided on the call today, what do you see as the most relevant for use in a real-world setting?
I think the reduction in the PANSS scores is the most important. It's a very widely used gauge for not only psychosis, but also negative symptoms and general symptoms. In both studies, this is only in the weight gain study, these were not very ill patients. They start out with a PANSS score of 70. Despite that not being a very high score, they still go down to 60. Don't forget that zero is not zero at the PANSS. I think it's around in the 30s or 40s. Even though they were moderately ill, they still improved. In the ENLIGHTEN-1 study, the efficacy study, there was actually a very large drop. I think the PANSS is the most important gauge for treatment response.
Great. If I can sneak one more in. Just curious if you've began discussions with payers, and if so, if there's any color you can provide around these.
Hey, Matthew, this is Richard. No, we're just starting that now that we have these data. Now getting them out is an important part of that because now we can begin the medical affairs and education process that will be necessary to make sure the payers are ready for this medicine.
Maybe just to add to what Richard's just said. I think the first step of engaging payers really begins with sharing the data with the clinical community, and we did that yesterday. There's a lot of excitement around the product. We had some physicians coming up to us after the presentation yesterday asking when this would be available. I think there's that education which needs to occur over the coming months. Then obviously, as Richard said, we'll start working with the payers and educating them as well.
Awesome. Great. Thanks for taking my question.
Thank you. The next question comes from the line of Chi Fong with Bank of America. Please proceed with your questions.
Hi. Good morning, guys. This is Chi Fong for Jason Gerberry. Thanks for taking my question. Just going back to slide number 21, the 24-week data. It looks like olanzapine weight gain is probably still on the upward trajectory where Alkermes drug has stabilized. Can you talk about maybe how much that 24-week cutoff underappreciate the potential weight gain difference between the two drugs on longer term exposure? It looks like when I look at the 52-week extension, ALKS 3831 stay more or less flat, maybe within 1% change of difference. I have a follow-up after that. Thank you.
Sure. I'll give my perspective and then I'll ask René to give his as well. Essentially what you're seeing there is exactly as you described it, is that there's continued divergence through 24 weeks of treatment for the olanzapine arm. The 3831 arm from week six onwards is completely flat, and we see that all the way through 76 weeks of follow-up. Once again, I'll go back to the discontinuation curve because I think there's a lot that lies within that curve. That is really that, as we know in controlled clinical settings, that olanzapine weight gain is often underappreciated. I think when you look at those discontinuations, you're seeing virtually all of the olanzapine discontinuations are above the line of completers within the olanzapine group for weight gain.
That tells a story because these patients invariably came off a study for reasons other than weight gain, and yet they were experiencing significant amounts of weight gain. I think our belief is also that patients on olanzapine who experience large amounts of weight gain just stop taking their medications. There's an adherence component there. I think the other component is also that then leads to relapses and hospitalizations. From our perspective, we would have expected to see those curves continue to diverge if we had managed to do a blinded study through the 52-week extension. Obviously, it wasn't ethical for us to do that. I think there's a lot of literature as well which demonstrates now that patients who were on olanzapine long-term, especially coming in with lower BMIs, experience weight gain which can span over years.
Obviously the flatness of the 3831 curve is incredibly important. I'll ask René maybe to give his perspective as well.
I agree with Craig . I think the impressive finding is here that there is separation at six weeks, which is quite early, and then there is no additional weight gain in the Alkermes arm. Which is impressive because like you said, the olanzapine goes up, and we know from clinical experiences will continue to go up. I also want to emphasize, although it was not one of the main outcome measures, that the waist circumference actually separates even earlier. It separates at one week and then again at four weeks. So it precedes the overall weight gain, which is very important because, like I said, the distribution of the weight gain is actually one of the biggest risk factors for cardiovascular disease and cancer as well. I think that is very important that there is almost an immediate benefit in reducing the largest risk factor for long-term complications.
Got it. If I may, I just want to sneak in a follow-up on that. If I look at slide number 37 on the expansion interim analysis through 52 weeks, it looks like subjects who switched from olanzapine to ALKS 3831 had attenuated or have stopped the weight gain from going. Just going to want to see the clinical meaningfulness of stopping the weight gain of olanzapine and whether that is clinically meaningful enough for physicians to switch current olanzapine patients from using olanzapine to ALKS 3831. Thank you.
Once again, I will give my perspective, I will hand over to René to give his clinical perspective. I think what we are seeing is that patients who switch from olanzapine to 3831, you do see that stabilization. The damage that is already done is there. I think from our perspective, it is most efficient to start patients on 3831. However, having said that, olanzapine patients gain weight over years. So what you are seeing here is, at very least, you can stabilize those patients by switching them over to 3831, which we believe is clinically meaningful. Also bear in mind that you see similar patterns on the waist circumference as well. As René said, that is a really important cardiometabolic risk factor as well. Our belief is yes, but ideally, we would like to have patients initiate on 3831. I will hand over to René to give his perspective.
Yes, I echo that. I think the clear advantage is to start patients on a drug like this because you do not want to have the long-term weight gain that the patients who have been treated with olanzapine have. It is, however, remarkable that there is no additional weight gain after those first six weeks. That is essentially what this says. Like a year-and-a-half follow-up, there is only a weight gain in six weeks, and then there is a stabilization, which I think is a very potent finding because a one-and-a-half-year follow-up is a really long time for clinical trials. Yes, you could switch patients from olanzapine to the Alkermes drug, but that is not the main strength of this drug. The main strength is that it should be a first-line drug in schizophrenia.
Thank you.
The next question comes from the line of Biren Amin with Jefferies. Please proceed with your question.
Yeah. Hi, guys. Thanks for taking my questions. Just a question for the company. Were the patients in the trial on exercise or diet regimen during the double-blind phase or the open label extension phase of the study?
No, the patients did come in with stable weight profiles over the three months coming onto the study. We did not implement any additional measures in the trial. It would have been too difficult to implement.
Got it. I've got a couple of questions for Dr. Kahn. Given these data, how would you compare the 3831 to other weight-mitigating treatments? I think there's some literature on metformin and topiramate potentially as add-on therapies.
Yes. Well, there are two things. One is this is extremely well demonstrated. I think all the other trials are small trials, not very large, not very well controlled, not very well demonstrated. That's one. Two, it's very difficult to combine it, what dose to use, at what point to use it. I think this has an added advantage. One, because of simplicity, of course, of administrating a single compound drug. But the most important thing is the long-term follow-up. I've been emphasizing that several times now. This is a one-and-a-half-year follow-up. You don't have those data on any of the drugs you mentioned.
Got it. Maybe just one additional follow-up, Dr. Kahn. If the drug's approved in second half 2020, what percentage of your patients would you switch off of olanzapine to 3831, and what percentage of patients on other atypical antipsychotics would you switch off of those therapies and prescribe 3831?
Yeah. The first one is easy. That's everyone. There's absolutely no reason for them to continue on olanzapine. You have this drug which is as efficacious and has no additional weight gain. To switch, you would switch a patient only if there is a need for better efficacy. Now, there are many patients who are partial responders who are still psychotic. I showed you in the ENLIGHTEN-2, they still had a PANSS of 70, which is not that high, but most of our patients will still, in fact, have a PANSS of 70. If I would then give the Alkermes drug, you could still oral olanzapine, which you won't do. You could still have an additional 10-point reduction in the total PANSS score. That's worth it. Everyone who would be moderately ill and is not responding totally, I would definitely try to switch.
I mean, try to switch, like I said, the current data suggests that that will be beneficial. Again, the most important thing is that you would start new patients. In short, everyone who's currently on olanzapine, I would switch. The patients who are partially responsive, which are most of the patients, I would probably switch unless they are doing really well, the new patients I would start on this drug.
Great. Thank you.
Our next question is from the line of Chris Shibutani with Cowen. Please proceed with your question.
Hi. Thanks for taking my question. This is Pam on for Chris. First, I have a question about the metabolic results. Were any of the results significantly different between arms?
Let me start off by just putting the metabolic results into context. The shifts in the metabolic parameters were small, and what we believe to be not clinically meaningful for the most part. The study wasn't actually designed to look at statistical separation between the arms. Numerically, in the 24-week study, we saw similar separations from baseline for olanzapine and 3831, but these were not clinically meaningful shifts. Very small shifts from baseline. However, in terms of long-term follow-up, whatever shifts in metabolism that we saw on the lipid parameters, except for HDL, these trended back towards baseline for 3831. If you look at HbA1c, which is probably the better marker for 6-week glycemic control, there was really no separation from baseline all the way through 0.07%.
Obviously, one of the most important metabolic parameters is waist circumference, and we really discussed that where there was very early separation of waist circumference. Thereafter, the profile looks very similar to the weight curve, where you see a complete flattening on 3831 and divergence with olanzapine throughout the 24-week study period. Obviously, this has long-term cardiometabolic consequences. We believe that represents a metabolic benefit over the long term with 3831.
Got it. Thank you. For Dr. Kahn, are any of the differences in laboratory parameters observed between arms clinically meaningful regarding the metabolic parameters? Thank you.
I think the most meaningful, although there's not a difference because there was no control, is that you can actually see the triglycerides normalize in the long-term follow-up study. I think that is quite significant. For the rest, in the first 24 weeks, there wasn't much difference, not much change in general. I think it's very important in the 52-week follow-up, you see a normalization, a total return to baseline of the triglycerides. That's very encouraging, I think.
Got it. Thank you very much.
Thanks, Pam. The next question is from the line of Umer Raffat with Evercore ISI. Please proceed with your question.
Hi. Thanks so much for taking my question. I just wanted to zoom in on the data in a little more detail, just to understand each of the moving pieces of it, because I know we had very good visibility on it previously as well, and there's some incremental more information. I'm just trying to focus more on some of the newer things. The first thing was, as I look at the image on weight gain by percentage of patients that gain more than 5%, 7%, 10% weight, et cetera, what catches my eye is if patients gained more than 7%, then obviously 3831 clearly outperforms it. I think it's 27% versus 43% in that ballpark.
If I dip below that, if I say patients that only gain 5% to 7%, et cetera, the number of patients that gain weight is actually lesser on just olanzapine. I'm just trying to figure out why that is, and is that 7% sort of like a threshold? If a patient's gonna gain more than 7%, you should be on 3831. If it's gonna be less than 7%, maybe you just stay on the generic olanzapine. I just wanted to get your thoughts, both from the company side and also from Dr. Kahn on that, and I had a follow-up.
Let me start off by giving our perspective. We'd pre-specified 7% and 10% weight gain as clinically meaningful thresholds. In addition to that, we had pre-specified in our statistical analysis plan that we'd be looking at a number of different thresholds. We looked at weight gain of 2%, 5%, obviously the 7% and 10%. We also looked at 15% and 20%, and on all of those cuts, there was significant separation in favor of 3831. In actual fact, in terms of the actual weight distribution curve, which Dr. Kahn presented yesterday, virtually that entire curve across the entire distribution is shifted to the left in favor of 3831.
If I may just sort of quickly add to that. I guess what I'm doing is I'm separating out the vintages five to seven, separate from seven to 10, instead of a more cumulative above five and above seven. You see what I'm getting at? If I just do five to seven, the percentage of patients that gain weight is actually slightly more on 3831, which is sort of the genesis of the question. What is it about that 7% magic number?
I think more patients are then getting more weight on olanzapine, the curve is shifted further to the right. I think that's kind of an artificial way of looking at it. When you look at the actual data and you look at the weight distribution curve, that entire curve for the entire distribution is shifted to the left.
I see. Fair point. Thank you very much for that. The second one was, I was just trying to understand. I know there's a slide showing 3831 in one graph and just olanzapine in a graph below it on the metabolic parameters.
Yes.
I know it wasn't done on the same graph, what I'm curious about is there more hyperglycemia on 3831? Also, if you could also just compare head-to-head the LDL on both the drugs.
Yeah, sure. In terms of the glycemic control, there was a lot of variability in the measure for glycemic control because it's really difficult in this population of patients to ensure fasting status. There were really small shifts in serum glucose. Numerically, it was slightly higher in the 3831 arm. The better measure to be looking at is obviously HbA1c, which is an indicator of six-week glycemic control and is less dependent on fasting status. There you see the exact opposite, that you actually see numerically lower separation from baseline for ALKS 3831. I would say in both of these parameters, these are not clinically meaningful shifts. That dynamic range in terms of these shifts from baseline is really not clinically meaningful. I'll ask René for his clinical perspective on that as well.
Yeah, I agree. With both points, it's very, very difficult to do in these studies to have everyone really fasting. I think the changes are clinically not meaningful.
Got it. If you could just quantify for us exactly how much did blood sugar go up on the active arm, and exactly how much went up on the comparator arm?
Yeah, the other aspect I just wanted to probably jump in and add is, the shifts we've seen from baseline we don't think are clinically meaningful, so it's really difficult to tease apart differences when there really weren't differences across the 24-week study. Over the long term, these trended towards baseline except for HDL. I think the most important aspect to capture is also what we would have expected to see in terms of the long-term weight gain of olanzapine. Because I think that's more tied to as their waist circumference increases, as their weight increases, you start seeing longer-term metabolic consequences with olanzapine, which start to kick in, and that's where you start seeing more significant aberrations in terms of lipid profile, glycemic control, et cetera.
The data that we're seeing from this study is actually not that different from what you would have expected to see in the 12-week olanzapine studies. It's pretty much in that same dynamic range of about a slightly higher 5% shift in total cholesterol, 3% for LDL, and about 20% for triglycerides. I think at the end of the day, it's really the long-term consequences that I think we need to be focused on to really get the clinical message out of this. Once again, I'll ask René just for his perspective.
No, I agree with you. The changes are very minimal. In fact, an increase in HDL is what you want to see. That's actually a good thing. An increase.
Sorry, your follow-up question?
Oh, no, that was it. On the glucose and A1c. Thank you very much.
Thanks, Umer.
The next question is from the line of Vamil Divan with Credit Suisse. Please proceed with your question.
Hi, this is Oy Favamo. Just have a few basic questions. First, just want to know, how long do patients typically stay on olanzapine? Secondly, the metabolic lab data, at least from the way I was looking at it, doesn't seem like there was, as you indicated, one meaningful or significant difference. You also said that it wasn't designed to measure these differences. I'm just wondering, is this what is typically seen? Does it suggest that perhaps the weight does not have an impact on these labs, or is the study not long enough? I just want to have a better understanding of that. The third question is, when the data was presented yesterday, I just want to know what the receptivity of the data was. Were people excited? Thanks.
Let me start off with your metabolic lab question. I think once again, let me reinforce what I said earlier on, that is, we don't think these are clinically meaningful changes. Yes, we do think that the delta that we saw on weight and the waist circumference data are critically important in terms of really sort of analyzing what we expect to see long term with olanzapine. With continued weight gain on olanzapine, with continued increases in waist circumference, we would expect to start seeing more significant metabolic aberrations in terms of both lipid profile as well as glycemic control. We actually think that's how this would sort of have mapped out if we had managed to perform a longer-term blinded metabolic study, which obviously would not have been ethical to do so.
In terms of the receptivity of the data yesterday, I think there was a lot of excitement. I'm going to ask Dr. Kahn because he actually presented the data yesterday. I will also ask him to speak to the duration of olanzapine use. René?
Yeah. I did get applause and no one walked out during my presentation. I think that was about as much as I could expect. No, I think it was well received, but it's always very hard to gauge how that exactly is. I think there is enthusiasm about it. A lot of people approached me afterwards. There was certainly a lot of interest from the media, I can tell you that. I'm doing at least three interviews today with various media outlets. I think that tells you something about the interest in this drug. How long patients usually stay on olanzapine, it depends very much on the kind of study. In our own studies, it's actually in a one-year follow-up, it's about 60%. That's not bad, despite side effects. Again, those are in controlled circumstances, and that's because it's a highly effective drug.
You've seen probably also here that it's, what is it, 67% or something, two-thirds, which is more or less reflecting the clinical practice and the trials that we have done. The compliance is actually pretty good despite the side effects. Although you've seen that the patients who do drop out, mostly drop out, while that may not be the cause, but they gain more weight than the average. Actually the patient staying on the drug may actually even be higher than two-thirds if you mitigate the weight increase.
Can I just ask a follow-up question? If a patient is on olanzapine already, given the data that's presented, do you see some sort of emerging profile to put in clinical practice, such as when would you switch patients over, like at what % weight gain?
Well, again, one of your colleagues asked more or less the same question. I think switching, of course, some patients you'll switch, but I think the strength of this study, of this compound really, is to use it as a first-line drug. You want to use the most effective drug at the earliest possible stage. That's exactly what happened with olanzapine. Olanzapine, it was the first-line drug used for all patients, and then it was more or less fall into disuse because of the side effects. It should actually regain this position of the first-line drug. If you have a patient with psychosis, this is the drug you should give because it's the most effective one, again, apart from clozapine, which is not a first-line drug. Just keep them on it, and the patient will probably stay on it.
Now, your question about which patients would you switch? Yes, patients on olanzapine, and I will be irrespective of weight gain, except for the patients who really tolerate olanzapine very well with no weight gain, I would leave them on it. There's no reason to change. If the patients are early in their treatment, I would switch them because of the ultimate risk that we know about olanzapine as a treatment. Like I said, there are quite a few newer antipsychotics on the market with their own problems. Some have EPS, some have sexual side effects, and some are just not as effective as, for instance, olanzapine. There, if the patients don't do really well, are not, as we call it, in remission, meaning they still have a PANSS score of around 70. They still are not totally functioning. They still may have hallucinations.
All those patients, I think, will be eligible to be switched to this compound.
I think maybe just to add to that, we've also got a study which is ongoing at the moment, the ENLIGHTEN-Early study, which is really looking at younger patients, early in illness.
Yeah
population of patients where we expect there to be significant amounts of weight gain on olanzapine because they are especially vulnerable to weight gain. That study is ongoing at the moment, and Dr. Carlis is actively involved in that study with us as well.
Okay. All right. Thank you.
All right. We've got time for one more question.
That question is coming from the line of Danielle Brill with Piper Jaffray.
Hi, this is Niraj Shah on for Danielle Brill. I just had a quick question. It was relating to something that you've covered before, there's a lower difference in the 5% threshold between the two groups. I was just wondering if there would be some sort of, I guess, separation for clinicians for prescribing generic olanzapine for patients who are at lower risk of gaining weight, particularly for the ones who are going to be gaining weight for less than 5%?
Yes, it's interesting. One of the aspects that we looked at when we analyzed the data was really early weight gainers, because we expected that there may be a differential profile there. As it turns out, there's absolutely no difference in the patients who gain early weight versus the entire population of patients. The same benefit is construed across the population. On top of that, as I said earlier on, when you look at the distribution curve of the weight, that entire curve is shifted to the left. I think from a clinical perspective, it's really difficult to try and tease out which patients are going to gain early weight, and there's really no clinical benefit. The entire population benefited in ENLIGHTEN-2.
Yeah, just to add, it's impossible. I wish we could. It's impossible to predict who will gain five and who will gain 15%. We don't know. There's no marker that can predict that.
Got it. In the thresholds that were defined, there was no demographic specificities that were seen between the 7% or 5%, right?
No. Well, if I may jump in. This is obviously a great study, but it's only 500 patients. There's been studies in thousands and thousands of patients on olanzapine, and we haven't found a single predictor of gaining weight. There's a wall of effort been put in that because obviously everyone would like to know that, but there are no predictors of who's going to gain weight or not.
Got it. Okay. Thank you so much.
Thank you. I'll now turn the call back to Sandy Coombs for closing remarks.
All right. Thank you everyone for joining us on the call this morning. If you have any follow-up questions, please don't hesitate to reach out to us at the company. Thank you.
This concludes today's conference. You may disconnect your lines at this time. Thank you for your participation.