Greetings, and welcome to the BioXcel Therapeutics Neuroscience R&D Day. At this time, I'd like to turn the call over to Dr. Frank Yocca of BioXcel. Thank you. You may begin.
Thank you, operator. Good afternoon, everyone. I'm Frank Yocca, Chief Scientific Officer at BioXcel Therapeutics. Thank you for joining our virtual Neuroscience R&D Day. As a reminder, during today's presentation, we will be making forward-looking statements regarding our R&D plans. These statements are subject to risks and uncertainties that may cause actual results to materially differ from those forecasted. The risk, the risk factors that may affect results are detailed in our most recent public filings with the SEC, including our most recent Form 10-Q on file with the SEC and available on our website. Next slide, please. It's been two years since we last hosted an R&D presentation for our neuroscience franchise emerging pipeline, and today we are very pleased to share with you several important and exciting updates.
We believe our approach, taken with our AI engine, has been highly productive and on the heels of our success with the approval of BXCL501 for the treatment of agitation associated with bipolar disorders and schizophrenia, we will discuss two of our most exciting programs. First, for the potential chronic treatment of neuropsychiatric symptoms with agitation related to Alzheimer's disease, we will unveil the compound behind BXCL502 and our progress in its development, as well as a new program with candidates that target apathy, also a key symptom of dementia as well. Finally, as part of our land and expand strategy, BXCL501 showed positive results in its phase Ib/II study in treating opioid use withdrawal and may be poised to help with what has become a national health crisis. Today's event will include prepared remarks with time at the end for Q&A.
Joining me for today's presentations are Dr. Vimal Mehta, our founder and CEO, Dr. Jeffrey Cummings, Dr. Sandra Comer, Dr. Michael De Vivo, and Dr. Friso Postma. Dr. Jeffrey Cummings is Research Professor in the Department of Brain Health and Director of the Chambers-Grundy Center for Transformative Neuroscience at the University of Nevada, Las Vegas. A world-renowned Alzheimer's researcher and leader of clinical trials, he has published nearly 900 articles and 44 books devoted to neuroscience, Alzheimer's disease, and clinical trials. Dr. Sandra Comer is Principal Investigator of the National Institute on Drug Abuse-funded Trial of Investigational BXCL501 for potential withdrawal treatment of patients diagnosed with opioid use disorder and professor of neurobiology in the Department of Psychiatry at Columbia University. She is also Director of the Opioid Laboratory in the Division on Substance Use Disorders.
Her research focuses on the clinical testing of medications for treating opioid use disorders, devices for treating opioid overdose, methods to maximize the use of naloxone by opioid users, and evaluations of the comparative abuse liability of prescribed pain medications. Dr. Friso Postma is Vice President, Artificial Intelligence and Emerging Portfolio for BioXcel. He has over a decade of experience in nonprofit and biotechnology, developing and translating innovative solutions, addressing unmet needs in neurodevelopmental and neurodegenerative disorders. Over the last five years, he has incorporated artificial intelligence methodologies to develop digital biomarkers for CNS-related disorders and is leveraging this experience to further develop BioXcel Therapeutics' AI capabilities. Dr. Michael De Vivo is Vice President of Neuroscience here at BioXcel. He has more than 20 years of experience developing drugs for neurological and psychiatric indications.
He was Director of Neuroscience at Pfizer for eight years and led several cross-disciplinary project teams during that time. Prior to that, he worked at Memory Pharmaceuticals and Trophix Pharmaceuticals, where he was Head of Preclinical Discovery. Next slide. With that, I'm pleased to turn the microphone over to our CEO, Dr. Vimal Mehta.
Thank you, Frank, and good afternoon, everyone. This is an exciting time for BioXcel Therapeutics as we are laser focused on advancing our late-stage TRANQUILITY and SERENITY programs for BXCL501, while continuing to generate new pipeline candidates through our unique AI-based reinnovation approach. We are making tremendous progress and have a clear development path for our two late-stage programs following our two recent meetings with the FDA this quarter. First, let's review TRANQUILITY, our highest priority for capital allocation. I'm pleased to announce that we have developed a clinical trial protocol designed to evaluate the potential at-home use of BXCL501 for the acute treatment of agitation associated with Alzheimer's dementia. The trial will evaluate the 60 microgram dose and is expected to enroll approximately 100 patients. The primary focus is safety and tolerability in this setting, and we will also be collecting additional efficacy data.
We will refer to this trial as TRANQUILITY At-Home going forward. We plan to submit the protocol to the FDA, and in parallel, we are planning to open multiple clinical sites for rapid recruitment of patients in this population. We look forward to initiating the trial, and we'll share the timing of our four data catalysts as we progress. Currently, we are initiating instructions for patient and caregiver use at home, which is required for our TRANQUILITY At-Home study. Turning to SERENITY, where we are evaluating the potential at-home use of BXCL501 for agitation associated with bipolar disorders or schizophrenia. We are pleased to have just received the FDA meeting minutes last Friday. These minutes are consistent with our prior understanding to evaluate the 120 mcg dose, with safety as the primary endpoint, with efficacy evaluations as exploratory endpoint.
We have developed the protocol to share with the FDA and working with clinical research organization partners to prepare for trial initiation. We will refer to this trial as SERENITY At-Home going forward. Both the TRANQUILITY and SERENITY At-Home trials are expected to be capital efficient and to have predictable recruitment rates and timelines based on our multiple pivotal trial experience. To advance these programs, I'm especially pleased that we just announced the strengthening of our existing clinical leadership team. We welcome Rajiv Patni in a newly created position as a strategic clinical advisor, and Vince O'Neill in his broader role as Chief of Product Development and Medical Officer. We are pleased to have already completed the amendment of our strategic financing agreements with Oaktree Capital Management and Qatar Investment Authority. This gives us additional operational and financial flexibility.
At the same time, we are acutely focused on strengthening the company's balance sheet. In that regard, we are actively exploring multiple financing options in parallel, having recently developed clear path forward for our two pivotal phase III programs. Let me now comment on our innovative clinical pipeline, the focus of today's presentation. The value of BioXcel Therapeutics extends far beyond our late-stage programs, and we are excited to host an R&D Day to highlight pipeline candidates in Alzheimer's-related symptoms, particularly BXCL502 for chronic treatment of agitation associated with dementia. In addition, with support from the federal government, we are committed to further advancing our development program to potentially address the enormous national opioid crisis with BXCL501. I am proud of the entire company for putting us in a solid position to move forward in 2024.
All of us at BioXcel Therapeutics are motivated to develop new treatment options for a greater number of patients. The discussion today reflects our passion and commitment in doing so. With that, I will turn it over to Frank to discuss our R&D strategy. Frank?
Thank you, Vimal. Next slide, please. Our R&D strategy is focused to build breadth and depth to our pipeline. We hope to achieve this by continuing to pursue different indications for BXCL501, while adding new programs from our drug discovery efforts. Besides our TRANQUILITY and SERENITY efforts, which are driven internally, the opioid withdrawal program, which you will hear about in a later presentation, as well as our PTSD program, are prosecuted by external clinical teams in academic centers with funding from government grants. Both the OUD and PTSD programs have completed successful Phase Ib/II trials and have moved on to Phase II. The emerging discovery pipeline is driven by our AI engine. Today, we will discuss our lead candidate, BXCL502, for potentially treating neuropsychiatric symptoms and agitation in Alzheimer's disease, as well as introduce a new program concept, BXCL503, focused on the potential treatment of apathy.
There are additional programs that are being evaluated as well, such as BXCL504, for treating aggression. These and other early concept programs are developed through discoveries made also with our AI engine and created with knowledge graphs. Can I have the next slide, please? This slide shows an excerpt of a map called a knowledge graph, that normally contains millions of nodes depicting brain regions, neural circuitry, and how it is linked to behavior and symptoms. The different neuropsychiatric disorders are, in essence, a collection of symptoms. BXCL501 is thought to act on the locus coeruleus, modulating the release of norepinephrine and is linked to the stress axis and stress-related symptoms such as agitation. While BXCL502 is believed to act on different circuitry, consisting of the habenula and raphe nuclei, which also mediate stress-related responses involving serotonergic signaling.
We will also show you behavioral data that suggests that BXCL502 may be having an indirect effect on the noradrenergic system. The BXCL503 concept is thought to act on yet a different but also evolutionarily conserved circuit that regulates reward behavior and linked to apathy, a hallmark symptom of dementia and Alzheimer's disease. Next slide, please. Taking the learnings from our knowledge graphs further, our emerging discovery effort is focused on mechanisms and targets which affect prominent symptoms experienced by Alzheimer's disease patients. If you examine this figure, you will see that the three approaches that we have taken with developing BXCL501 and now BXCL502, have the potential to treat a number of neuropsychiatric symptoms like agitation, irritability, sleep disturbances, and anxiety. With the BXCL503 program, our goal was to broaden our efforts towards treatments for apathy, which you will be hearing about shortly.
At this time, let's turn our attention to BXCL502. Dr. Michael De Vivo will discuss the discovery and development of BXCL502, which is our novel agent for the chronic treatment of neuropsychiatric symptoms, which include agitation. Mike?
Thank you, Frank. My name is Michael De Vivo, and I'm the VP of Neuroscience for BioXcel Therapeutics. I am speaking to you today about our progress with our pipeline, specifically BXCL502. Next slide. BioXcel uses knowledge graphs to identify stress-related pathways that underlie neuropsychiatric behaviors, such as agitation or apathy. The purpose of this effort is to uncover assets with hidden value that can be developed in an efficient and timely manner. Neuropsychiatric behaviors are a large unmet need in the field of neurocognitive disorders, such as dementia due to Alzheimer's disease. BioXcel identifies unexpected links among neural pathways, drug targets, and clinically evaluated compounds. This diagram demonstrates the strategy. First, identify known parts of the brain that respond to stress, a trigger for neuropsychiatric behaviors.
Next, find drug targets that regulate those pathways of interest, and finally, probe a compound library consisting of compounds with good drug properties to find an appropriate compound to develop. Using this strategy, we identified the combination of 5-HT signaling through the raphe and noradrenergic signaling through the locus coeruleus as promising drug pathway to target. The algorithm suggested that the 5-HT7 receptor, a GPCR in the serotonin receptor family, regulates these pathways. Probing our compound virtual library, we found that latrepirdine would be a good candidate to develop. Latrepirdine, also known as Dimebon, was previously under development by Pfizer and Medivation for cognitive decline in patients with Alzheimer's disease. Latrepirdine did not meet the cognitive endpoints in those studies, and development was halted. However, as a result of those studies, the safety of latrepirdine in the patient population, Alzheimer's patients, was established.
Moreover, pooling the data from three large studies, there was a change in the NPI. NPI was a secondary endpoint in these studies. NPI is the Neuropsychiatric Inventory, a clinically useful scale that measures changes in neuropsychiatric behaviors. Next slide, please. In the field of neurocognitive disorders and dementia, it is increasingly obvious that neuropsychiatric behaviors drive the increasing level of care required for these patients, possibly even more than cognitive decline. These behaviors include agitation, apathy, delusions, and more. These symptoms are shown in the graph on the slide. The graph shows results from a large cohort of Alzheimer's patients. The prevalence of symptoms was quantified. Of particular interest for latrepirdine, there was a high prevalence of agitation in these patients, about 40%. Dr. Cummings will talk more about these behaviors during his presentation.
Suffice it to say, there are a few drugs approved for the treatment of these behaviors, and the need is great. Given the data already available, we believe latrepirdine is a promising candidate to treat neuropsychiatric symptoms associated with Alzheimer's disease. Next slide, please. So why latrepirdine? This slide shows our excitement. Latrepirdine was well tolerated in Alzheimer's patients, dosed up to one year. Based on the Pfizer and Medivation studies, there were no significant safety concerns. An early sign of potential efficacy was established using the NPI. Although the NPI measures a host of behaviors, BioXcel did extensive preclinical work to identify those behaviors that likely contribute to the NPI change seen in these patients. Also, in five preclinical animal studies, latrepirdine was active in improving behaviors related to agitation, both anxiety and aggression.
These data suggest that the NPI score change seen in the patients may be driven by factors in the NPI related to agitation. For both the safety and early signs of potential efficacy reasons, BioXcel proposes that latrepirdine may be suitable for development for the indication of chronic agitation in patients with neurocognitive disorders, including Alzheimer's disease. Next slide, please. As mentioned, Pfizer conducted two studies: the CONCERT and CONNECTION studies, with the intent of demonstrating improvement in cognitive decline in Alzheimer's patients over 26 weeks, and safety up to 52 weeks. An earlier study was conducted for Medivation.
Although the trials failed to show benefit, measure the frequency and severity of multiple behavioral symptoms, including agitation, it is worth noting that patients may have been chosen for these studies that had a low baseline of behavioral symptoms in order that they would be able to complete the cognition batteries. This chart shows the pooled data from those studies. Despite the low baseline in all three studies, there was an improvement in the NPI score versus placebo, and the pooled result is significant. In the corresponding safety analysis, there are favorable safety and tolerability results. As can be seen here, adverse events were similar between treated and placebo groups, as were serious adverse events, and dropouts were identical. No adverse events of special interest, such as somnolence or cardiovascular changes, were noted. This explains our excitement about latrepirdine. Next slide, please.
In the clinical studies, as mentioned, there were signs of activity or an early sign of efficacy using the secondary endpoint, NPI. Because NPI measures many behaviors, including agitation, BioXcel planned a series of preclinical studies to increase our confidence and rationale that the NPI score is driven, at least in part, by agitating behaviors. As seen in two representative studies, latrepirdine reduced anxiety, agitation, behavior in the elevated plus maze and aggressive agitated behaviors in the resident intruder study. In the elevated plus maze study, conducted by administering the drug yohimbine to mice, yohimbine causes the mice to become aroused and anxious. Yohimbine activates the locus coeruleus. Anxious and agitated mice will not explore the open arms of the elevated maze. Either diazepam or latrepirdine restored their exploration, which we believe indicates that anxiety was reduced.
This is a gain of function response, meaning the animals must be more active, explore the open arms, and therefore, the activity cannot be attributed to any locomotor or sedating effects. In the resident intruder study, a mouse is housed alone for a period of time, then a new mouse, an intruder, is added to the environment. The introduction of the new intruder mouse initiates aggressive behavior, behaviors such as biting or scratching. Latrepirdine and the positive control significantly reduced these behaviors. No changes in locomotor activity were noted at these doses. That data is not shown in the graph. In summary, in summary, we believe these studies, along with several others not shown, are consistent with the hypothesis that the NPI score shown in patients may be related to the changes in agitated behaviors shown in the preclinical studies. Next slide, please.
One criticism of the latrepirdine studies is that the potential mechanism of action of the drug was not clear. Many possible mechanisms of actions have been proposed, including 5-HT7 receptor antagonism. In this figure, in work done by BioXcel, the mechanism of action is potentially 5-HT7 receptor antagonism. In these studies, BioXcel transfected human 5-HT7 receptors into cells and used the cells for binding and second messenger studies. In the first panel, 5-HT stimulated adenyl cyclase activity was measured, and the percentage inhibition of the 5-HT effect was measured. In the second study, the binding of a 5-HT7 selective radiolabeled tracer was measured in the presence of latrepirdine. In both studies, the potency of latrepirdine was calculated to be about 10 nM or 4 ng/mL .
The number 4 ng/mL is of interest because it is also the calculated free brain concentration in the preclinical studies at doses that met the endpoint. It is also the plasma concentration achieved by the three times 20 mg a day doses used by Pfizer that may have produced the NPI score change. Next slide, please. Because the intended indication, chronic treatment of agitation in patients with Alzheimer's disease, requires chronic dosing, it is important to establish whether latrepirdine can be dosed in a chronic manner. In this study, latrepirdine demonstrated both continued positive activity after repeat dosing and also potential reversibility, meaning the effect was lessened when drug was stopped. Stress causes the mice to remain immobile, meaning that they don't swim to try to escape the swim test.
Improvement is measured by an increase in mobility of the mice as they attempt to escape the chamber. As can be seen in the figure, during dosing from day one to day five, when the animals experience stress alone, there's a decrease in their swimming behaviors, and they become immobile. Drug, in this case, latrepirdine, decreased the immobility, causing improvement from days five to day 12. When the drug was washed out, meaning they were no longer dosed, performance returned back to vehicle levels. When drug was reinstated on day 20, the animals again performed better, and it reinstated the beneficial effect. These are important features for a chronic drug, maintaining efficacy after repeat dosing and also reversibility, allowing effects to be washed out if necessary. Next slide, please.
So although, latrepirdine exhibited a promising safety profile along with a potential early sign of efficacy, BioXcel determined it would be difficult to develop the drug, given the large population variability in plasma exposures after dosing. A solution to this problem was necessary for latrepirdine to be successful. In recent years, we have seen that different companies have taken different approaches, take a promising compound with limitations, combine it with a second compound to make it a successful drug. Examples include dextromethorphan with bupropion to block its metabolism, xanomeline with trospium to block peripheral side effects, and BioXcel used an innovative approach with dexmedetomidine. Overcome the poor bioavailability, about 20%, by administering it on a sublingual film and directly absorbing it into the blood, bioavailability increases to 80%.
For latrepirdine, BioXcel took the approach to test compounds that could stabilize the metabolism and thereby support the developability of latrepirdine. To do this, we started screening compounds known to affect metabolism. Next slide, please. In this graph, we show one representative experiment. Latrepirdine was incubated with human liver microsomes and a set of metabolic inhibitors. After one hour, some of these compounds were able to stop the metabolism of latrepirdine and were chosen for further analysis. The goal is to find one of these compounds would be suitable to be co-formulated with latrepirdine and reduce the variability in plasma exposures after dosing. This will allow once-a-day dosing, rather than the three times a day dosing used by Pfizer.
BioXcel is now working with a set of prioritized compounds that should be able to be co-administered with latrepirdine and stabilize plasma concentrations after dosing. These compounds have been retested and the findings are consistent. Preliminary planning on a formulation strategy has begun with the goal of co-administering latrepirdine with one of these compounds. Next slide, please. In summary, BioXcel believes that BXCL502, which will be latrepirdine co-administered with a metabolic stabilizer to be chosen, has attributes that could potentially be useful for chronic treatment of agitation in patients with Alzheimer's disease. BioXcel has done studies that provide new insight into the mechanism of action of latrepirdine, and that it likely affects both serotonin and norepinephrine signaling pathways. BioXcel has identified compounds that potentially could improve the PK profile of latrepirdine, potentially allowing once-a-day dosing.
The newly formulated potential drug, BXCL502, will be designed to complement existing efforts with BXCL501, a drug that is being developed for treatment of acute symptoms. BXCL502 has clinical safety results and early efficacy signals that support further development. Next slide, please. With that, I'd like to turn it over to Dr. Jeffrey Cummings, one of the leaders in the field of quantifying neuropsychiatric symptoms. He will discuss the Neuropsychiatric Inventory, measuring agitation relief in Alzheimer's disease. Dr. Cummings?
Thank you, Mike. Great to be with you today. Great to be part of this presentation. I serve as a consultant for BioXcel, and I'll be talking about the NPI, as Mike said. Next slide. I want to establish the magnitude of the problem of Alzheimer's disease and our progress in understanding it. In the United States today, there are 6.5 million individuals with Alzheimer's dementia, and by 2050, this will become 19.5 million. So this is an enormous problem. Almost everyone knows someone who has this disease. We've become increasingly able to define the pathology and understand the biology of Alzheimer's disease. On the right, for example, I'm showing you some of the dramatic findings with PET scanning that show the amyloid protein in the brain of the Alzheimer's disease patient.
In addition to advances in diagnosis, there's been substantial advances in treatment. For the first time in 17 years, we have new drugs for the treatment of Alzheimer's disease. Aducanumab and lecanemab have been approved as the first disease-modifying treatments for Alzheimer's disease, and brexpiprazole was approved this year for the treatment of agitation in Alzheimer's disease. This is the first drug for any neuropsychiatric syndrome in Alzheimer's disease, and the first drug for agitation in any neurodegenerative disorder. There's heightened public awareness about the occurrence of Alzheimer's disease, and these successes and this awareness has increased and improved the investment environment for drug development in Alzheimer's disease. Next slide.
To give you a sense of the magnitude of the problem of agitation, about 70% of Alzheimer's disease patients will have agitation during the course of their disease, and in cross-sectional studies, about 40% will be symptomatic with agitation. 60% of these patients will receive pharmacotherapy, and that might be antipsychotics, antidepressants, or benzodiazepines. All of these drugs have different differences, and they have drawbacks in the sense that antipsychotics, for example, have black box warnings, which are a hindrance to prescribing these agents. The etiology of agitation in Alzheimer's disease includes biological changes of the type that Mike described and environmental influences. The consequences of agitation are many. They include reduced patient and caregiver quality of life, more rapid progression of cognitive decline, and earlier nursing home placement. Next slide.
There's been substantial progress in establishing the parameters necessary to conduct clinical trials of agitation in Alzheimer's disease. An important milestone was the definition of agitation, comprised by a consensus conference of the International Psychogeriatric Association. So a neurocognitive disorder, such as Alzheimer's disease, is present, and one of three agitated behaviors must have been present for at least two weeks: verbal aggression, such as shouting and cursing, physical aggression, such as hitting and kicking, or motor hyperactivity, such as hyperactive and excessive pacing. These cannot be due to pain or delirium or some sort of environmental provocation. And then rating scales are used to establish the severity of the agitation at the baseline. So as you see in the figure below, first, the patient is diagnosed with Alzheimer's disease, then it's determined if they meet the IPA criteria for agitation.
Then the severity of the agitation is determined with scales, and then the patient is randomized, if eligible. Next slide. The Neuropsychiatric Inventory is the most commonly used instrument to measure behaviors in clinical trials in Alzheimer's disease and other neurodegenerative disorders. I developed this scale, and it has 12 neuropsychiatric syndromes, including agitation, hallucinations, delusions, irritability, disinhibition, elation, apathy, depression, anxiety, appetite and sleep changes, and aberrant motor behavior. It's based on a caregiver interview, who is asked to recall these behaviors over the past month. The caregiver rates both the frequency and the severity of the behavior, as well as their own distress. Next slide. Here are the examples again that Mike showed of the latrepirdine results using the NPI. There were three trials, and you see the characteristics of the trial there in the blue panels.
You can see that there was no meaningful effect on cognition or function, but there was a reproducible effect on the NPI score, whether you're looking at the mean difference or the change from baseline at 26 weeks, or the change from baseline at 52 weeks. There was a similar effect size across the published studies. The low baseline scores, as Mike mentioned, limit our ability to show a drug-placebo difference and minimize the difference that's likely to be seen. Next slide, next slide. There's one additional study interrogating these same three studies, was based on a meta-analysis of 127 articles and found five trials and three Alzheimer's trials.
Again, no cognitive or functional benefit was found in this analysis, and again, the NPI showed a significant benefit in terms of the reduction of the total score in the treated patients compared to the placebo patients. Next slide. There are two primary measures that are used in Alzheimer's disease clinical trials. One is the Cohen-Mansfield Agitation Inventory. This is 29 agitated behaviors, such as kicking and shouting and hitting, those things that are noted in the NPI in the IPA definition. And then the NPI, you have heard about, with 12 neuropsychiatric syndromes. In clinical trials, it's common to use the CMAI as an outcome assessment and the NPI to characterize the severity of the syndrome at baseline. It's also common to use the Clinical Global Scale of Severity to further assess the agitation changes in the clinical trial. Next slide.
There are some risk mitigation activities that have to go on in terms of trials like this. These can be mitigated with planning. These include slow recruitment and a placebo response that frequently occurs. In terms of latrepirdine itself, we need to know which items on the NPI responded most robustly, and we need to explore more whether patients who are selected for a behavior like agitation show a robust response. Next slide. So in summary, Alzheimer's disease is very common, and Alzheimer's disease with agitation is common. The treatment of agitation is a large unmet need in the market, despite progress with the approval of brexpiprazole. Trial foundations have been established, including a definition of agitation, experience with the clinical trial, outcome assessments, and a regulatory pathway has been established. The preliminary observations with latrepirdine and the NPI are highly encouraging.
Thank you.
Thank you, Jeff. What I'd like to do right now is turn it over to Friso Postma, who's our head of AI drug discovery, and he's gonna discuss a new concept that we've been investigating, which is treatments for apathy and the BXCL503 opportunity. Friso?
Thank you, Frank. Good afternoon. My name is Friso Postma. I am the VP of AI Drug Discovery at BioXcel Therapeutics.
… Today, I will be providing insights into BioXcel's emerging pipeline, particularly focusing on our pursuit of compounds for nomination as BXCL503, and how it may treat apathy as associated with dementia. Next slide, please. At BTAI, we refer to AI as augmented intelligence. We do so because our platform is augmenting the decision process involved in drug re-innovation along the entire value chain. Thus, AI is not a black box providing answers, but an ecosystem of AI tools also referred to as composite AI. Our platform, which we named Novare AI, is tuned to address very specifically questions linked to the drug re-innovation process. The central component of our platform is a knowledge graph, visually represented by the large circle in the middle of the slide. This graph interlinks compounds with their canonical targets, neural circuitries, indications, and defines the relationships between them.
For example, a compound might inhibit a receptor that activates a neural circuit relevant to stress behavior, which we then can test in our preclinical platforms. This approach allows us to deliver mid- to late-stage clinical assets, either by identifying new opportunities through our models or by targeting specific symptoms like agitation, or by linking compounds that have been in the clinic and were tolerated well to new targets or indications. As you heard from Frank, we identified apathy associated with dementia as a promising opportunity through the indication side of the knowledge graph and opportunity prioritization models, and initiated a strategy to identify a putative BXCL503. Next slide, please. Our strategy makes use of an observation that many receptors can allosterically affect each other because they form heteromeric complexes. This can result in modified receptor affinities for the ligands or affect receptor stabilization at the membrane.
Some examples of heteromeric receptors are shown on the right. The advantage of targeting hetero receptors is that it may confer a level of selectivity, and that heteromeric receptors often to those circuitries where only both are expressed. In addition, rather than directly stimulating receptors or increasing levels of neurotransmitters globally through reuptake inhibitors, modulating receptor activity in response to the endogenous ligands would be more specific. We deployed Novare AI to find targets associated with circuitry that governs goal-directed behavior. Then we used Novare AI to select compounds with a published clinical safety profile and predicted potential efficacy in treating apathy. Lastly, we proceeded to test these compounds in our preclinical platforms. Next slide, please. This slide visualizes the process and is analogous to the BXCL502 pathway.
We identified the indication apathy, looked for potential neural pathways regulating the behavior, then looked for potential targets in the circuitry, dorsal striatum, which led to the identification of late-stage clinical compounds that are predicted to lessen apathy. In this case, putatively through modulation of dopaminergic receptors. Next slide, please. Apathy in Alzheimer's disease is an emerging opportunity, potentially larger than agitation, given its prevalence. It's characterized by a lack of initiative, interest, and emotional responsiveness, which manifests as reduced voluntary, goal-directed behavior. This represents a high unmet medical need, with about 5 million patients within Alzheimer's disease alone, and currently, there are no FDA-approved drugs. Apathy contributes significantly to the disease burden, is associated with an increased dependence on caregivers, decreased quality of life, cognitive and functional impairment, and increased mortality.
As an important side note, one of the underlying mechanisms leading to apathy is very similar to anhedonia, and it involves similar neural circuitry. Anhedonia, or the inability to experience pleasure, is a key component and large unmet medical need associated with depression. We consider our BXCL503 compound candidate promising for addressing apathy and potentially anhedonia. This data from two phase II clinical trials in an indication outside of apathy indicated that the compound was well tolerated. Despite being deprioritized for commercial reasons, it exhibited safety results that support development to treat apathy symptoms. As mentioned, we hypothesized BXCL503 may specifically modify dopaminergic signaling in goal-directed behavior circuitry, in contrast to direct receptor activation. Next slide, please. The BXCL503 compound candidate was tested in SmartCube, which is a system that uses computer vision to monitor and extract behavioral features from mice behaving both passively and reactively.
The left side shows a mouse in a cage, looked at from three different angles by the computer. The animal is challenged with different stimuli, such as the cage floor moving up or a puff of air from the nozzle on the right. On the bottom right, you see the real-time assessments and classification of the behavioral features by the algorithms. Over 1,000 features are extracted by the system and compared to a database of behaviors obtained with marketed drugs for known indications. The deep learning classifiers that were trained on this database are then used to assign signatures to the test compounds. On the right, you see that when we tested BXCL503 compound candidate identified by Novare AI, it was classified as a stimulant by SmartCube. Next slide, please. We next tested the BXCL503 compound candidate in a preclinical model that measures goal-directed behavior.
A reduction in goal-directed behavior is a model for apathy. Rats are trained to press a lever to perform a task to receive a reward. However, the rats have to incrementally increase their efforts to get the reward. The point at which the animal gives up pressing the lever is recorded. Thus, a baseline is established for goal-directed behavior, represented by the blue bar under Vehicle. Depleting monoamines using the vesicular monoamine transport inhibitor drug tetrabenazine, reduced the effort rats were willing to put in, and the rats gave up more easily, as represented by the orange bar under Vehicle. Next, treating the rats with the BXCL503 stimulated animals and increased their effort, while in the tetrabenazine-treated rats, the effect of monoamine depletion was reversed.
The effort the rats are willing to put in was equivalent, if not better, than when rats were treated with bupropion, a norepinephrine dopamine reuptake inhibitor. From this data, we concluded that in a preclinical model of apathy, a putative BXCL503 demonstrated positive activity. Next slide, please. Taken together, we believe that our BXCL503 compound candidate represents a clinical asset that could potentially be developed in an accelerated manner. The compound has been in phase II in over 500 subjects for up to 26 weeks and was well tolerated. We have tested our lead compound in two preclinical models, where it showed robust activity, both as a stimulant and improving goal-directed behavior. Lastly, the BXCL503 compound candidate has a proposed mechanism of action that confers increased specificity in modulating dopaminergic signaling and circuit selectivity. Next slide, please.
In conclusion, we believe that BXCL503 concept for the treatment of apathy presents a compelling value. It engages a proposed mechanism that would be preferred over overt stimulation of dopamine receptors or reuptake inhibitors, such as methylphenidate, also known as Ritalin, which has recently been tested in the phase III ADMET trials. The BXCL503 compound candidate was well tolerated in published studies and importantly, a recent consortium, the International Society of CNS Clinical Trials Methodology Apathy Work Group, that included regulatory bodies, came out with diagnostic criteria for apathy, which together with the recent ADMET phase III trials, indicate that trial foundations may have been established. Finally, apathy associated with Alzheimer's disease aligns well with the company's BXCL501 and BXCL502 development programs for treating Alzheimer's disease-related symptoms. Next slide, please. And with that, I'll-- we'll turn it over to Dr.
Sandra Comer, who will discuss BXCL501 for the potential treatment of acute opioid withdrawal. Dr. Comer?
Thank you. So next slide. As many of us are probably aware, the U.S. is in the midst of an opioid crisis. It was actually declared a public health emergency in 2017. And just earlier this year, the White House Office of National Drug Control Policy declared fentanyl, either adulterated or associated with xylazine, an emergent threat. So, you know, this has led to a tremendous increase in the number of opioid-related overdose deaths, which I'll show you in just a minute. But what we're concerned about is the treatment gap. There are no effective or no treatments for opioid withdrawal that have been tested against either fentanyl or xylazine.
A recent publication has estimated that about 1.25 million opioid-related deaths will have occurred in the U.S. between 2020 and 2029. Next slide. So this is the timeline of, you know, this crisis. So it has occurred in about 3 waves, starting in the late 1990s, with the rise in prescribing of opioids for chronic pain. There was an associated increase in prescription opioid-related overdose deaths. And then starting in about 2010, there was a rise in heroin-related overdose deaths. And then just a few years after that, illicitly manufactured fentanyl began to infiltrate the illicit opioid market, and this is really, you know, responsible for this enormous escalation in opioid-related deaths.
So as of June 2023, well over 100,000 drug-related overdose deaths occurred in the previous 12 months. And of these, the large majority, as you can see in that purple line on the figure, is related to synthetic opioids, and it's mostly due to fentanyl. Next slide. So the reason that, you know, these drug dealers are adding xylazine to the fentanyl supply is that fentanyl is pretty short-acting, in terms of producing a high. And so, xylazine is a sedative, and so it's, you know, thought to prolong the effects of fentanyl. There's also a thought that xylazine also can boost the high that users feel when they use fentanyl.
So the problem with adding fentanyl or xylazine to the fentanyl supply is that it can increase the respiratory depressant effects of fentanyl, making it more lethal. And even, you know, another effect of xylazine is that it can cause these really horrific skin ulcers and abscesses and other complications. So this is what has led to the designation of fentanyl combined with xylazine as an emerging threat in the U.S. Next slide. So xylazine is a adulteration of opioids, actually began in Puerto Rico about 20 years ago. But much more recently, it has emerged in the mainland U.S., and it started in the Mid-Atlantic states, so it was really concentrated in Philadelphia. It has since rapidly increased to other states in the East Coast and is now rapidly spreading across the country.
Next slide. A couple of years ago, we started to study the effects of BXCL501, the sublingual formulation of dexmedetomidine, as a potential treatment for opioid withdrawal symptoms. This was... I'll show you the data in just a minute. The results of a phase Ib randomized, double-blind, placebo-controlled, ascending multiple ascending dose study, which we published just this year. Jermaine Jones in my lab wrote up the results of this study. Next slide. The purpose of this study was to evaluate the safety and tolerability of ascending doses of BXCL501 relative to placebo in people with opioid use disorder who were physically dependent on opioids. They were maintained on morphine, given orally in four doses throughout the day during a stabilization phase, so days one through five.
They also received placebo sublingual strips of BXCL501 twice a day. So starting on day 6, we abruptly switched the active morphine to placebo morphine for the remaining days of the study. And then also, participants were randomized to either placebo or active BXCL501. So the doses that we tested, you can see here, ranged from 30-240 mcg, given twice daily. And at each of the active dose cohorts, we randomized them in a 4/1 ratio to active or placebo medication. And there were 15-25 participants per group. Next slide. So what we observed was a decrease in both clinician-rated and subjective ratings of withdrawal symptoms.
So just to orient you to this, these figures on the right here, these are ratings, mean change from pre-dose ratings of, the clinician-rated on the top and then the subjective rating, ratings on the bottom, across the six days when they were switched over to the placebo morphine. So as you can see from the top in that open circles, when participants received placebo medication, the withdrawal symptoms peaked at around day 9. When they were randomized to the active BXCL501 dose, there were significant reductions in withdrawal symptoms based on clinician ratings. These effects were mirrored in the subjective ratings of opioid withdrawal. So the 240 micrograms twice daily dose of 501 was the most effective.
And these effects occurred at doses that produced no serious or severe adverse effects, and the ones that we were most paying attention to were orthostatic hypotension, bradycardia, dizziness, and somnolence, 'cause these are the major symptoms that are produced by lofexidine, which is currently the only non-opioid medication for treating opioid withdrawal symptoms that's on the market. So the mechanism of action of lofexidine and dexmedetomidine are somewhat similar, but I'll show you some differences in their effects. Next slide. And one thing that I wanted to point out is that the studies that were conducted with lofexidine occurred many years ago before the introduction of fentanyl.
In the study that I just showed you, for the MAD study, about 76%-92% of the urine drug screens that we collected at the beginning of the study were positive for fentanyl. We think that fentanyl is changing the way that withdrawal symptoms are exhibited. So in our study, 40% of the participants dropped out during the morphine stabilization phase, whereas in the lofexidine study, where people were mostly using heroin, none of the participants dropped out during this morphine stabilization phase. So, because the characteristics of fentanyl-related you know physical dependence are very different than heroin, we don't really know how to optimally manage the symptoms.
So we're running a follow-up study to the MAD study right now, and this really represents a unique opportunity to address this problem. Next slide. So I just wanted to show you a couple of... Can you go back? Yeah, there we go. A couple of additional endpoints from the MAD study. So what we observed in that study was that subjective ratings of, you know, insomnia and problem sleeping were significantly reduced with the 240-microgram dose of 501 compared to placebo. Similarly, the ratings of anxiety or irritability that were rated by the clinicians were also significantly reduced. So both of these symptoms are very commonly reported by people who are going through opioid withdrawal, and these symptoms are not really affected by lofexidine.
So this is kind of a unique-- we're thinking it might be a unique feature of dexmedetomidine. Next slide. Another important point that we observed in the MAD study was that retention in the study was superior with active BXCL501 compared to placebo. And this is important because it suggests that we may improve the successful transition of patients onto another medication for treating opioid use disorder, such as buprenorphine or sustained-release naltrexone. Next slide. So these are the adverse events that I mentioned earlier. Just want to point out that we have to be cautious in interpreting these results because these were done in two different studies that were done at very different points in time.
So the placebo column on the very far left and all the columns with the dark blue were from the MAD study that we published recently. Then all the three columns on the right were from the lofexidine study that was done before the introduction of fentanyl. But I just wanted to show you sort of the pattern of adverse events that we saw. Again, this is very preliminary. The study with the multiple ascending dose study was done in far fewer participants, but at least so far, the incidence of these side effects appear to be lower. Next slide. So this is the design of the ongoing study. We have just added a fourth site to the study to accelerate enrollment. So we've shortened the treatment period.
So in the other, the MAD study, it was 12 days, and now we're keeping people inpatient for seven days, and we're randomizing them to either placebo, 180 or 240 micrograms twice daily of BXCL501 or lofexidine. We're using a triple dummy dosing procedure, so everyone at every dosing period will get a sublingual strip, oral tablet, and oral solution. And every dosing administration is identical, so they won't. They don't know what it is they're getting. Another difference between this study and the MAD study that I described is that, well, in addition to randomizing completely the order of dose administration, in the MAD study, obviously, we gave it in ascending doses. But rather than using methadone, we...
morphine in the MAD study, we switched over to using methadone because for a couple of reasons. Methadone—morphine was not really effective in, you know, controlling the withdrawal symptoms, and I talked to multiple colleagues in the field who observed, you know, very similar things in the fentanyl-dependent people. Methadone seems to be holding them a little bit better, and methadone is currently approved as a treatment for, you know, for detox from opioids. So we're starting them out on 30 mg a day of morphine and then tapering them off over the next several days.... We're also stratifying participants into each of the groups as a function of xylazine positive or negative urine drug screens at admission.
So this will allow us to get a little bit of a handle on what the withdrawal symptoms might look like as a function of, you know, the presence or absence of xylazine. And, we're thinking that this is a little bit closer to what real-world use of the dexmedetomidine might look like in a detox clinic, for example, that uses a methadone taper. Next slide. So we are, you know, obviously really interested in the clinician-rated and subjective ratings of withdrawal relative to placebo and how they relate to the adverse effects that we, you know, are likely to see. We're very carefully monitoring both subjective and objective measures of sleep and clinician-reported ratings of irritability and anxiety.
Because the MAD study showed that retention was better with 501, we're using that as an endpoint in this study as well. And, you know, as I mentioned earlier, this study really offers us a really unique opportunity to examine fentanyl-related withdrawal and even more importantly now, the presence or absence of xylazine and how that affects withdrawal. Next slide. So there are several clinical scenarios that I think 501 might be useful for. So one is in the emergency department. So people who have overdosed often when they are taken to the emergency department, they are often in severe opioid withdrawal because they've received naloxone. So, you know, I think 501 would be helpful in that situation.
And increasingly, emergency department staff are directly transitioning patients onto something like buprenorphine, which also can precipitate withdrawal. So, you know, I think the 501 might be helpful in that context. And then I guess a little bit more traditionally, in an inpatient treatment facility, where patients are detoxing or transitioning to another medication for opioid use disorder like buprenorphine or naltrexone, they very frequently require treatment for withdrawal symptoms. So this is a perfect situation for giving 501. And then finally, there's been a shift in the field to, you know, transitioning patients onto MOUD on an outpatient basis. So, often what's done is, you know, clonidine is co-prescribed or lofexidine for managing withdrawal symptoms on an outpatient basis. So I think 501 could be helpful here.
And then another patient group that I think could really benefit from 501 are patients who are prescribed opioids for pain. So if they've taken it for a couple of weeks, they can often experience withdrawal symptoms when they, you know, either no longer need it or switched over to another medication. And so, you know, there's not a really clear, you know, clinical guidelines for how to manage people who need to be tapered off of opioids when they're taking them for pain. So, that's pretty much it for my talk. And the next slide just shows you, you know, the references that we cited in the study. And I just want to, you know, thank the National Institute on Drug Abuse for funding our current trial as well as BioXcel.
They're co-funding this study, and we're really excited about this, this medication as a potential new treatment for opioid withdrawal symptoms. Thank you.
Thank you, Sandy. This brings us now to our Q&A session. So operator?
Thank you. We will now be conducting a question and answer session. If you would like to ask a question, please press star one on your telephone keypad. A confirmation tone will indicate your line is in the question queue. You may press star two if you would 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. One moment please, while we poll for your questions. Our first questions come from the line of Yatin Suneja with Guggenheim. Please proceed with your questions.
Hey, guys, can you hear me?
Yes.
Hi. Thank you so much for taking my questions. Very exciting news about the two new assets, the disclosure that you've made today. So I have two questions. One is more around the funding front, and then I have a question about the plan study. So the first question I have is, so you have the minutes from the FDA. You're about to start studies for 501 , and then you have these two exciting assets. So how are you thinking about funding the development of these three assets, and what is the funding situation now? And then the second question I have is, are there specific studies that you need to do in order to file an IND for 502 , and when can we expect that?
Hi, this is Vimal. Very good question. So, we are very pleased with the two recent meeting with the FDA and having a clarity to initiate those trials. As I mentioned in my prepared remarks, our higher priority for capital allocation is TRANQUILITY At- Home study. That's where we are focused, and we believe that's one of the biggest drivers and a catalyst for the company. So we're working on that, putting the protocol together, sending it to the FDA, and then working with our CRO partners to open the site. So that will be our preference. Recently, last week only, we received a meeting minute from the FDA for SERENITY At- Home study, and as you noticed, their primary focus is safety and then the exploratory endpoints are efficacy.
So these trial, we are working to put together the protocol, and we will be in a similar position like TRANQUILITY, maybe just a month behind on that one. But we'll continue to allocate our capital for TRANQUILITY At- Home study. And regarding the 502, it's a great question. Why are we presenting? The reason we are presenting is it's such an exciting asset, and we have a very good understanding now of the mechanism for the latrepirdine. We have such a large safety data behind it. We know how we can improve the PK profile of this drug, and we know what the development path is. So 2024, for 502, will be primarily focused on the formulation strategy, so they are not that capital intensive.
503 is more in the even earlier stage, where we are trying to validate our hypothesis. So 502, 503 will not require that much capital, while they can be a big value driver as long term, but our focus will continue to be doing the capital allocation for our TRANQUILITY At-Home study.
Got it. One more question, if I may. So is it clear in the... and this is for the broader team. Is it clear in the literature whether the overactivation or hypoactivity of 5-HT7 receptor occurs in agitation? And do we know if there are other possible targets for latrepirdine?
Hi, this is Mike De Vivo. So many possible targets have been postulated for latrepirdine, including histamine H1, 5-HT7, and others. In our view, because of the potency for the 5-HT7 receptors and the plasma exposures at the doses that Pfizer used, we think there's a good chance that it's related to 5-HT7 receptor occupancy, the effects that we're seeing on the NPI score change. So we're trying to gain further evidence that that's true using these preclinical models. Now, as far as the mechanism, what is known in the literature is that 5-HT7 is a prominent post... It's also located on the raphe, but also postsynaptically in many brain regions.
Based on the results that we saw, it modulates locus coeruleus-dependent activities, which are related to hyperarousal and agitation.
Very good. Thank you so much.
Welcome.
Thank you. Our next questions come from the line of Sumant Kulkarni with Canaccord. Please proceed with your questions.
Good afternoon. Thanks for taking my questions. My question is for Dr. Cummings. So, why is latrepirdine's mechanism of action potentially better suited to tackle chronic agitation, and how would you compare it to other agents in this space or candidates in development? And separately, in your experience with patients, how would you conceptualize what constitutes acute versus chronic agitation in the Alzheimer's context?
Yes, thank you. Both good questions. Acute versus chronic, of course, it's arbitrary. And, we've used two weeks, in the IPA criteria in the IPA consensus development, as a, as a, an artificial, change in a, in what is really a seamless continuum. Some people have very acute agitation that can be managed in a very short period of time. And, so that's the role of a, of an acute intervention. And then many develop chronic agitation over time, and so there's a, a, a larger market, I think, of the, interventions that would control chronic agitation. But these definitions are arbitrary, and they'll be specific to the clinical trial. And as long as we maintain fidelity with our own definitions, I think the FDA will be okay with that.
We'll have to test, of course. In terms of mechanism, you know, we're really struggling to understand the mechanisms of any of these behavioral features. We have hypotheses that are being tested. We know, of course, that brexpiprazole, which is a mixed receptor agent, was effective in reducing agitation. We've seen good preclinical evidence of reduction of agitation with the 5-HT7 agent, the latrepirdine, but there was no inclusion of brexpiprazole in those trials. So I don't really see a way right now of comparing agents unless they were included in the same assay. I think we have good grounds for hypothesizing a 5-HT7 mechanism, and we have good non-clinical evidence. We do not have comparative evidence.
All right. Thanks.
Thank you. Our next questions come from the line of Elliott Bosco with UBS. Please proceed with your questions.
Hi, all. Elliott Bosco on for Colin Bristow, UBS. Could you just outline some of the expectations for milestones for both the TRANQUILITY and SERENITY At- Home studies this year? And then on Dimebon, I think there were elevated rates of depression in some of the prior studies. And so can you frame how you think about that in terms of chronic dosing? And then, sorry, lastly, can you just qualify why we should be treating clinically the neuropsychiatric symptoms along with agitation? Thank you.
So Elliot, I can take the first question. This is Vimal. Regarding the TRANQUILITY and SERENITY, SERENITY program, key milestones, as I mentioned in my pre-prepared remarks, we have developed the protocol based on the FDA's feedback. Now we are getting ready and working with our partners to initiate the trial. TRANQUILITY At- Home is given the preference, and we are marching forward with that. I mentioned that we will be initiating a initial portion of the trial where you are looking for information for use for TRANQUILITY, so that the caregivers and patients can follow the instruction. That part is starting very soon, maybe this year, and that will complete shortly, and then we will be in a position to start our TRANQUILITY At- Home study. That will be the key focus.
As we get alignment with the FDA on the protocol, we will be able to provide the initiation date for TRANQUILITY program and the data readout. We expect this trial to be about 100 patients, and as I mentioned, that there are large pool of patients available, so recruitment, we have a very good idea based on our past studies and other, studies in this space, that how fast we can recruit. We will be able to provide when the data catalyst will be. So, all of that work is being worked upon, and we are very actively pursuing to initiate the TRANQUILITY program. SERENITY, we just got the meeting minutes last week. We have absorbed the meeting minutes. We have a full clarity based on our prior assessment, what the, clinical trial program will look like. So we are in a good position.
We also have a protocol built there, and then we will be going back to the FDA with the protocol, and as you know, once we had a chance to get the blessing from the FDA on the protocol, we will be in a position to initiate this trial. Company, in general, has a lot of experience in working in pivotal trials for Alzheimer's, schizophrenia, or bipolar, and we will continue to expedite the initiation of the trial as fast as we can. So we will be providing a more guidance to you as we make progress. Second question was related to the 502?
About depression.
Yeah. So that's. This is Mike De Vivo again. So in the meta-analysis of all the trials and all the pooled data, there was no adverse event or serious adverse event that was different from placebo. Depression is a very common symptom and occurs generally in Alzheimer's patients. And unless you have data that I'm not aware of, there was no difference in the rate of depression between the placebo group and the latrepirdine treated group. And then your final question is about why treat neuropsychiatric symptoms. Maybe Dr. Cummings can address that.
Yeah, I was thinking what an interesting question that is. One thing that happens in Alzheimer's disease is that they get multiple simultaneous neuropsychiatric symptoms. So although we choose, say, agitation, that patient, statistically, that population of patients will have psychosis, they'll have substantial irritability, they'll have a sleep disturbance with nocturnal agitation, they'll have some motor behavior. So, when you choose an intervention for a major symptom like agitation, the downstream effects are usually pretty marked, and the general NPI score will decline, as well as a decline in the symptom of choice itself, agitation. And we have some new network analyses that are pretty cool, that look at exactly this effect.
So that it's all, it's essentially inadvertent, and you cannot avoid treating multiple symptoms at the same time.
Thank you. Our next questions come from the line of Graig Suvannavejh with Mizuho. Please proceed with your questions.
Thanks. Good afternoon. I got two quick questions. One, just on, maybe I'm looking at slide 48 in your deck, and I'm just trying to ask about dose response, with respect to what you're seeing in the opioid setting and perhaps a lack of dose response. So if you could comment on that. And then maybe a second question is just stepping back, very big picture, just, you know, BioXcel has always positioned the technology as kind of AI-based. And I'm wondering if you could just remind us simply how you consider your AI similar or different to other companies' AI-based approaches? That'd be helpful. Thanks so much.
Thanks, Graig. Sandy, can you take that first question?
Sure. So I think the lack of dose response in that MAD study is mainly due to the fact that the sample sizes were relatively small. You know, it's related to the lack of information that we have about, you know, the time course and the, you know, severity of withdrawal when people are using fentanyl so frequently. So, you know, that's why in the current trial we're running, the sample size is much larger. You know, the MAD study included, you know, around 15-20 people per group, and we doubled that sample size with the hope that we can, you know, see a better separation in doses.
Hi, Graig, this is Vimal. I will answer the AI question at a strategic level, and then I will pass it on to Friso to provide more depth into our platform. I think what I would like to say is that BioXcel Therapeutics was established to use AI-based approaches, and we were very fortunate that with our first BXCL501, that got approved in about 3.5 years from first in human, all the way to market data, which is now IGALMI and helping lots of patients and caregivers. We have built on that success in our platform, and as Friso presented, Novare AI is a new version that we have created.
I call this as a version 2.0 of our AI platform, and it has taken to a next level where we are looking at not only like in a knowledge graph, there is a, like, you know, a lot of AI-based analytical tools have been built to do these analyses very rapidly. I think we are at a point that our platform can continue to generate more assets like 502 and 503. We just presented you two examples of it. In terms of the 502, and this is very paramount to the company's strategy, we try to come up with an IP strategy. There are a lot of product concepts get that generated, but we take only those product concept where we have a...
We can build a strong intellectual property, as we are doing for 502, with combining with a metabolic stabilizer. We look for probability of regulatory and technical success. And, as you can see, we have a lot of safety data to support and also some of the efficacy data that we have observed in the secondary endpoint, combined with our strategy of looking at the preclinical and translational work. So it's no different, the way we discovered 502 than what we did with the 501. We start investing our capital on a human proof of concept. That's what we plan to do for 502. That's not due in 2024 because we are still optimizing the formulation plans for 502.
Once we have that, then we will look for a rapid human proof of concept using the combination of nalbuphine plus stabilizer that we have come up with. So that, and then if you think about 503, that's another agent and company has chosen to focus on Alzheimer's-related symptoms. The reason is, that we can do it with our platform is because it can produce a lot of candidates, and we don't need to be looking for licensing for a drug that may be sitting somewhere else. We can customize what therapeutic area we want to focus on and what drugs we can work on, what is available, and we are not limited by one drug choice. We have multiple product concepts when our team comes, then strategic teams make a decision, which one we are going to take forward.
So that's the kind of output of the platform. It speaks for the system we have created. I would say the biggest success factor is having the AI platform and having the AI scientists and neuroscientists, clinical, medical, and commercial teams under one roof is biggest part of our success. And we will continue to maintain that to build a sustainable R&D pipeline. I know, we can only prosecute certain things in our pipeline, and at some point we will think of exploring partnerships for these assets as we see that our pipeline is like, you know, full, and we can work with outside partners. So that's the kind of a company's strategy to build a sustainable R&D pipeline, and I will pass it on to Friso to provide you more context around the AI.
Yes. So this is Friso Postma. I think that the turning element here that it's not necessarily the type of AI we're using, it's about the questions that we're asking. So we use a series of different types of AI tools, composite AI, to specifically address questions in a holistic manner. And for this, you really truly need a multidisciplinary team that consists of neuroscientists that know about AI-
... and AI engineers that know about neuroscience. So it's about making the AI actionable and applied. And, this is specific to the drug re-innovation process. Does that help, Graig?
Yes. Thanks so much.
Okay. I think we have time for one more question.
Thank you. Our next questions come from the line of Ram Selvaraju with H.C. Wainwright. Please proceed with your questions.
Hi, thanks for taking my questions. These are for Dr. Comer. I was wondering if you could shed some light on the following two aspects of the opioid withdrawal program with potentially 501 . Firstly, you talked a little bit about how the advent of fentanyl and fentanyl plus xylazine is complicating the overall opioid withdrawal landscape because patients typically on these medications don't appear to behave the same way as, for example, people addicted to heroin. And in particular, I was wondering if you could talk a little bit about what challenges, if any, this may pose to the patient enrollment and also the patient dropout rate that you expect to see in the ongoing proof of concept study with 501 .
Then secondly, if you could also give us a sense of, you know, you talked about these three potential pools of patients who could benefit from 501 as an approved therapy for opioid withdrawal. And give us a sense of whether that would contrast in any meaningful way with the manner in which you see lofexidine currently being used. Thank you.
Sure. These are all really good questions. So there are emerging studies that have demonstrated pretty clearly, in my mind anyways, that fentanyl-related withdrawal is just different in a number of ways than heroin withdrawal. So we have seen both in my lab and in other labs around the country that, you know, as I mentioned during the presentation, that, you know, studies where, you know, I would routinely bring in people and maintain them on morphine, and they were... You know, the dropout rates in my studies tend to be about 10%, in that scenario. And now, like, you know, the dropout rate is much higher. And there's some really nice studies coming out of Johns Hopkins to further drill down and characterize these differences.
You know, the patients tend to be a lot more impulsive, the withdrawal symptoms are much more severe, and we're really struggling as, you know, a field of researchers and clinicians as well on how to, you know, how to manage this. So I think something like 501 has a lot of advantages over lofexidine. So, you know, the doses of lofexidine that appear to be, you know, helpful in treating some of the subset symptoms of opioid withdrawal, are also the doses that are producing a lot of, you know, adverse events, especially related to, you know, reduced blood pressure and heart rate and things like that, that can be dose-limiting.
You know, at least so far, we're seeing that, that's less, that seems to be less of a problem with 501. So I think we're getting doses that are effectively controlling withdrawal that are not producing the same level of withdrawal symptoms. And you know, I don't want to underplay the importance of the insomnia and the anxiety and irritability. You know, I think that's a common characteristic across the schizophrenia and the opioid withdrawal platforms, is that, you know, those kinds of symptoms seem to be well controlled by 501, and they are not by lofexidine.
To answer your question about how, you know, the changing symptomatology of opioid withdrawal in people who are physically dependent on opioid fentanyl, compared to, you know, previously, patients who were physically dependent on heroin. We designed our study to kind of take that into consideration. Anybody who takes at least a single dose of the study medication is considered to be valuable. You know, we're primarily interested in safety. This is still, you know, a phase I study, but we are, you know, we're thinking that the latency to drop out of the study is an important endpoint, as I discussed during my presentation.
Based on our data, we're expecting to see that, you know, active 501 will increase the retention in the trial compared to placebo for sure. We're not 100% sure if it'll distinguish from lofexidine. That, you know, that remains to be seen. But we should be able to evaluate... You know, the data will be usable across all the sub-participants, regardless of when they drop out.
Thank you very much.
So, this is Frank Yocca again. Are there any other burning questions?
There are currently no further questions in the queue.
Okay, very good. Well, I'd like to take the opportunity here to thank everyone for joining us for our Neuroscience R&D Day. We've been pleased to share with you a number of important and exciting updates to our AI-driven clinical pipeline. This included an in-depth look at BXCL502, what we believe will be a very promising chronic treatment for neuropsychiatric symptoms in dementia, as well as a summary of a new concept with BXCL503. We're also very excited with the prospects of BXCL501, having an effect here on the OUD situation, as well as the work that's being done on PTSD, which is another huge problem. So again, thank you for your questions, and we look forward to keeping you apprised of our continued progress going forward.
Have a great day, everyone.
Thank you. This does conclude today's teleconference. We appreciate your participation. You may disconnect at this time. Enjoy the rest of your day.