Great. Thanks very much, everybody, for joining. It's my pleasure to be moderating this chat with Abe Ceesay, CEO of Rapport Therapeutics, and Troy Ignelzi, who is the Chief Financial Officer. We'll keep this mostly conversational, but maybe Abe, Troy, if you guys want to give a quick background of the company and an update on the various programs through 2029, we will do Q&A. Thanks so much, and take it away.
Great. Paul, good morning, and good morning to everyone else that is joining. I really appreciate the opportunity for the discussion today. Rapport Therapeutics, a company focused on precision neuroscience, we're leveraging our foundational kind of scientific discovery, which is receptor-associated proteins. Receptor-associated proteins really allow for a new era of neuroscience drug discovery, really in two ways. One is the ability to target a receptor-associated protein directly that, in the case of our lead program, allows for really an unprecedented level of neuroanatomical specificity in targeting an AMPA receptor, but also have really allowed for a new era of drug discovery as it relates to really opening up new targets with other ion channels that is really reflected in our discovery pipeline overall. The company is based on a partnership between Third Rock and J&J to form Rapport, a true company creation model.
Not only were we able to bring out our lead program and our discovery pipeline, but we are also able to bring over some of the core scientists that were working on these programs at J&J. We really hit the ground sprinting with our efforts here at Rapport, but also our institutional knowledge around our lead program, but also our discovery pipeline is pretty significant. Just a quick note on our lead program, RAP-219. This is a TARP gamma-8 AMPA modulator program. We're really leveraging what is validated biology for epilepsy specifically with AMPA modulation, but through targeting a receptor-associated protein, it really allows for a neuroanatomic specificity that, in our data today, both preclinically as well as clinically, is really presenting what we believe is an unprecedented therapeutic index, potentially in anti-seizure medications, but also a best-in-class profile.
We're currently executing our phase IIa proof of concept study in refractory focal epilepsy patients with data expected there in the third quarter. We're also here on the ability to launch into our bipolar mania study, which we'll also initiate in the third quarter with RAP-219.
Awesome. Let's just get into it and talk about the RNS study, right, given that that's the hot topic right now. Maybe let's take a step back, talk about the design, Abe, the population you're enrolling, and what was the thought process for this being the best way to describe proof of concept from the model?
Sure. Our objective, as we thought about a proof of concept study with RAP-219, is to really be able to assess proof of concept in the patient population that we would ultimately assess in registrational studies, and that is the refractory focal epilepsy patient population. As you know, Paul, historically, for proof of concept studies with investigational anti-seizure medications, it's primarily been done in photosensitive epilepsy or external EEG models. Both are interesting models, and both have shown data that is supportive of pharmacodynamic activity, but ultimately, those data do not translate very well into understanding effect size and really translating into designing your study in the registration patient population, which is refractory focal epilepsy patients. We saw an opportunity based on our work with the community to do a study in that patient population, and that is the study that we're doing.
Specifically, our study is assessing a refractory focal epilepsy patient population that has an implantable neurostimulation device, the RNS device that is manufactured by NeuroPace. What this device leverages is what is picking up a lot of appreciation in the community, a novel biomarker by the name of long episode. Long episodes are extended EEG activity that are really referred to as subclinical seizures, and they are precursors to both electrographic seizures as well as clinical seizures. What data has shown is that this biomarker and the reduction of long episodes is highly predictable of reduction of clinical seizures. We've seen this data published prior to us starting the work, but then we've also published some data working with NeuroPace that has further supported the validity of this biomarker.
Okay. All right. I want to tackle this from a few angles. The first was, you know, it's interesting. This is a personal anecdote, but you know, I have a friend whose kid had one seizure, and the end of the story is great because he's okay now and he's off medication. After he had a seizure, to try to vet whether it was going to happen again, they did an EEG, and they looked at whether or not you were seeing this interictal pattern, right? How similar are long episodes to this concept of just high-frequency baseline EEG that people have been? I feel like the investing community views long episodes as super new, but in my mind, looking at kind of aberrant EEG activity as a component of epilepsy diagnosis is not new.
I'm just curious, how similar is this to other waveforms that people have described?
Yeah, you're right. Looking at EEG patterns is not new. As you're assessing epileptiform activity as well as just overall epilepsy in patients, it is not a new concept. What is different about the RNS device is the specificity of an interictal EEG recording. What do we know about this patient population? One, we know that this patient population has a defined seizure-focus onset zone. We know specifically that these probes and the reading of this device are reading true epileptiform activity. That is really important for us. The second aspect is the concordance, and what concordance is, is how many of these long episodes actually turn into electrographic seizures. In our study, the minimum concordance is 50%. When you think about reducing that long episode, in essence, you are reducing seizures.
For us, we think that tying those two things together are really important. It's just not aberrant EEG activity. It's aberrant EEG activity that directly correlates with an electrographic seizure.
Maybe this is a silly question, but why isn't the endpoint then electrographic seizures?
Yep. The reason that the endpoint isn't electrographic seizures is really the capacity of capturing that data. The data is most easily captured and most reliably captured on capturing long episodes and the reduction of long episodes. In our study, what we are doing is we are capturing long episodes, but we are also capturing clinical seizures in terms of diaries. We think our presentation of data and the data that we're capturing will ensure that we're really presenting that full picture.
Okay. I mean, would you have electrographic seizure data too?
Yep. We are capturing that data, and we're capturing many, I'll call it EEG-based endpoints. There are many different endpoints and biomarkers that we'll be looking at. The study itself is powered on the reduction of long episodes and that correlation of long episode reduction to clinical seizure reduction.
Yeah. Okay. I know the baseline is widely variable, but what's the typical baseline frequency of long episodes in this population?
Yep. We're waiting to see the full baseline characteristics of our study once the study is fully enrolled. What we have done to understand the baseline characteristics of this patient population is to really understand what is that minimum threshold of long episode frequency that you need to be able to detect a clinically meaningful change. That minimum is eight. We think about that as an eight over a 28-day run-in period, so an eight over a 28-day prospective baseline period. We see that clearly as the minimum. We expect that we will see patients enrolled in this study with a higher disease severity, both in terms of their long episode frequency, electrographic seizure frequency, and then our minimum threshold for clinical seizure frequency to be included in the study is a minimum of one.
Yeah. Okay. I've hypothesized just from papers, and I've asked you this on, well, your minimum is one, and one alone would not be well enough powered to show anything that in all likelihood, you'll find patients where the rate of clinical seizures is just higher because they have an RNS, they're participating in a clinical trial. If you look at some of the publications that you guys have pointed people to on long episodes, in those publications, the baseline seizure frequency naturally, again, is usually higher than one. Can you talk about what you're observing on baseline seizure frequency in this study? If you can't be specific, at the very least, do you feel like the diary seizure data and the change in that data over time is going to be interpretable in this trial?
Yep. We haven't guided specifically on what we're seeing in baseline characteristics, but I think you pointed to a couple of really important points. The first is when you look at the demographics of these patients. When you look at the demographics of these patients, they look very similar, not exactly the same, but very similar to patients that have been historically enrolled in phase III epilepsy trials, looking at the baseline demographics of the cenobamate trial or what is expected to be the baseline demographics of the Xenon trial as an example. We do believe that there's an opportunity to see a higher baseline count of clinical seizures. We also are capturing that data for a reason because we know it's important. We will see what the data tells us.
The way that we are capturing the data, both in the baseline period as well as capturing data through the treatment period, puts us in a position to be able to show movement on that secondary endpoint. We're capturing in a very rigorous way.
Yeah. Okay. Okay. Great.
One thing I want to add is just why the field, the community, is so interested in long episodes as a measure. That is thinking about clinical seizure diary data. What the field has come to appreciate, and I know, Paul, you've done some work on this too, is how inaccurate that data actually is in terms of both its capture of the actual disease severity of a given patient, but also the capture of the errors in capture and clinical trials. When you look at the disease severity of these patients, many of these patients are having seizures in their sleep. They're amnestic during their seizures. Clinical seizure diaries really are not the true picture of the disease severity and disease burden of patients. Also, it just relies on human input and also is somewhat limited by human error.
What the field really likes about long episodes as a biomarker is it is highly objective. There is no reason for human error. When you have a concordance like we have in our trial, you know that when you are reducing long episodes, you are ultimately reducing epileptiform activity.
Yeah. Yep. Makes sense.
Paul, sorry, real quickly, Paul, at the beginning, you mentioned that these types of seizures are not new to the community. It's Wall Street that's listening in on this and catching up and understanding it. I think, further extend on your point and Abe's about the community, that extends into the drug development community as well. They've been looking for something like this to overcome some of the limitations of the diary mechanism, in particular in phase II, where you're really running a proof of concept. You've got several other companies that you cover. And phase IIs can run two and a half to three years to get these results because you're doing the diaries. You've got the imprecision. We're going to have definitive results around these long episodes in less than a year from when we started the trial.
I think there's advantages from the community that has seen as well here.
Yeah. Makes sense, Troy. Thanks for adding that. Can you guys talk about the work you've done to arrive at this 30% responder bar on long episodes? I guess, does it pass the smell test that a 30% change in long episodes would actually predict a bigger change in clinical seizures? Why would that even be the case? Why wouldn't it be the same?
Yep. This work is work that has been done not only by us, but by other members in the epilepsy community. We presented data in 2024 at the AES meeting, but much to our surprise was additional data that was done by other members in the community looking at the correlation between long episodes and clinical seizures. Really, what all of the data has pointed to across multiple publications and analyses is kind of this 30% threshold or 30% cut point of reduction in long episodes predicts at least a 50% reduction in clinical seizures. In terms of why there is a discrepancy between a 30% reduction in long episodes predicting something greater than that in clinical seizures, I think, again, it comes down to how clinical seizures are captured, right? We know that there is just some level of error in that.
I think the precision is, again, around reducing direct epileptiform activity in the form of long episodes and that correlating with what is that clinically meaningful change. We believe that that 50% threshold is an important threshold. As you just look at having a percentage of patients that are hitting that level of seizure reduction is ultimately what's going to be meaningful as we think about a phase III trial.
Yeah, absolutely. All right, great. The one other question I want to ask you in the RNS study is just the concept of it being a similar population to a focal phase III study. You and I have been talking about this for a couple of years, Abe. I mean, the one inclusion, the one very well, there's multiple inclusion criteria in this study that won't apply, right? You're not going to have people with an RNS in phase III. There's one nuance, and it's that you're requiring patients to have at least one RNS lead in the hippocampus. I think that raises the natural question of whether this phase I/II study is better enriched to the mechanism of RAP-219, which might, we don't know, but might be more pronounced in the hippocampus based on where this TARP is expressed.
Can you just talk a little bit more about that? I guess, again, let's say we see a great seizure reduction in this study. Obviously, going from phase II to phase III, you never know. From a patient population perspective, do you feel like a phase III would essentially be running the same experiment?
Yep. Yep. Great question. I think there's a couple of aspects here. The first is what we're trying to show in proof of concept versus what we know about the pathology of focal epilepsy and also what we know about the biology and the target here with TARP gamma-8. If we think about that first, what we know about focal seizures is 50% of focal seizures have their onset zone in the hippocampus. The rest of these seizures, the other 50%, are really starting in the cortex and have some, and what the community and experts will say, almost all of those also have some hippocampal involvement. What we know about this target is it is highly expressed in all cortical structures and the cortex overall.
The highest expression is in the hippocampus, but that doesn't mean that it is void or devoid in other areas of the cortex. We believe that we have expression in the right areas as you think about the pathology here in focal epilepsy. What we also know is in preclinical testing, when we look at models, the many models that RAP-219, but other compounds in this category have been tested in, these are models that aren't specific to kind of hippocampal focus. These are models that are focused on general seizures, overall focal seizures such as corneal kindling. We see broad efficacy in those models, which we believe is highly translatable, and I think the community does as well.
Specifically in our study, what we want to understand is we want to make sure that we had the fastest way to proof of concept data going back to Troy's point. What we're requiring is patients have to have at least one of their leads in the hippocampus, but can have other leads in other cortical regions. We think that the data is a more homogenous population, but ultimately, by having leads in other brain regions, we'll also be able to shed some light on the broader expression of TARP gamma-8.
Okay. Could we or would you be comparing change in long episodes or electrographic seizures across different leads, or is that too much of a leap of faith than what we're trying to get at?
Yeah, I think the population will probably be too tight to do those types of comparisons. We have a whole host of data that we're going to be looking at and assessing in many different ways. We should just again remind folks that this is a proof of concept study, 20 patients. We think it's designed very well, powered on overall group change with long episodes.
Yeah. Yeah. Okay. Okay. Great. Maybe one question on safety tolerability, Abe. I mean, in general, I thought the phase I data you guys put up for 219 was better than average for an anti-epileptic drug in healthy volunteers. I mean, I probably talked to some people about this who are listening into this panel right now. I mean, if you look back at healthy volunteer studies for AEDs, like AEDs, antipsychotics, these drugs are often not that well tolerated in healthy volunteers. That being said, I thought the one adverse event of interest that we were thinking about when you guys put that data out was a case of amnesia. I know there's more color you can provide on it, but what is that added context for those that are newer to the story?
Because when we think about, again, this mechanism, right, being hippocampal targeted, you think about the hippocampus and its role in the brain. It'd be good to kind of get your level of comfort that that's not a real concerning safety signal and that you can ultimately have a pretty wide margin here.
Yep. First, we know that tolerability or AE is an on-target AE. As you mentioned, we have a target that is highly expressed in the hippocampus, which is the seat of memory. What we do know is both in terms of our preclinical as well as our clinical experience is that these are transient AEs. These are AEs that ultimately resolve on their own. Specifically, as you look at our clinical data, we're achieving receptor occupancy to the levels of 85% and above. When we do see these AEs, we have never seen anything greater than Grade 2. It is not something that we think is going to be a significant liability to the compound. We think we're going to have much more data, obviously, in a patient population with the RNS data that is going to be coming here in the third quarter.
Given the low frequency, but also the pretty mild to moderate grade, what we've saw thus far at that level of receptor occupancy, we don't think that it's a major liability by any means, especially when you look at the total tolerability profile, kind of that risk-benefit profile compared to other anti-seizure medications.
When you guys started doing the titration and starting lower, I mean, the profile clearly improved. I mean, do you feel like you could completely mitigate this by lower dosing and titration?
Yeah, it's something we're going to assess. I don't know if we can completely mitigate all tolerability in any way.
I mean, all tolerability issues. I honestly meant this one specifically, right? I mean, I think there's par for the course tolerability issues, but memory or something like that would be, I think it would be something that would be, I'm not saying it would be unprecedented for AEDs or AEDs that interfere with cognitive function, but it'd be something more notable, right? I think that's why I'm asking about it.
Yep. What we do believe is that we have a lot of opportunity with dosing. If you look at what we have learned through the human PET study as well as our multiple ascending dose study, we will obviously look a lot at the RNS data to really understand that PK-PD relationship. We have some opportunity as we think about dosing moving forward into registration studies. That is going to be balancing kind of that tolerability-efficacy trade-off. Right now, we know we have a pretty wide therapeutic index with this compound, and we are going to continue to kind of think about dosing going into phase III.
Yeah. Okay. Okay. Great. All right. Maybe to wrap things up, I mean, I wish we could spend another half hour talking about this mechanism and other indications, but I want to make sure we at least cover it. I mean, can you talk briefly, Abe, about your efforts in bipolar and pain?
Sure. We have always thought about this compound as a pipeline and a product type of opportunity that is based on the biology, but also based on the clinical precedence of anti-seizure medications being effective for indications such as bipolar as well as neuropathic pain. We are really excited about both of those areas. We will be starting the bipolar study here in the third quarter of 2025. Data expected in early 2027 for that study. What is really interesting about this is that there are three approved anti-seizure medications for bipolar, both in terms of anti-mania agents, but also mood stabilization agents. We believe that based when you look at some data, some biology, and really hyperexcitability in the hippocampus, it gives us a really nice shot as you think about this mechanism and also this mechanism working through the glutamate pathway.
If you think about having a potential anti-mania agent that has no sedative properties, that would be a real kind of transformational agent for patients. Neuropathic pain, very similar, except for the fact that the biology is even stronger here, in our opinion, and our preclinical data is pretty supportive in neuropathic pain too, given the fact that the other expression of AMPA receptors containing TARP gamma-8 is in the dorsal horn of the spinal cord. There was actually very early work done in very early compounds in this category in areas of pain, and it always has been kind of a promising indication for this mechanism in biology.
Okay. Great. Thank you very much, Abe and Troy. We look forward to your data in the third quarter. Yeah, we'll talk soon. Best of luck.
Thanks, Paul.
Great. Thanks, Paul.