At Jefferies, and thank you for joining me today. It is my pleasure to have Remy Luthringer, CEO of Minerva, joining me. Welcome, Remy.
Thank you so much for having me. Yes.
Of course.
Always a great pleasure to be here.
Great. Remy, for those in the audience less familiar with the Minerva story, can you please give us an overview of what you're working on and what you're trying to achieve milestones over the next 12 months?
Yeah. We are working on probably one of the highest unmet medical needs in the space of psychiatry, which are negative symptoms in patients diagnosed with schizophrenia. As you know, schizophrenia is mostly three types of symptoms. The positive symptoms, hallucination, agitation, and this is what is currently treated with existing treatments, with the antipsychotics. You have some cognitive impairment, and you have negative symptoms, and it's very clear that the negative symptoms are the symptoms which are impairing the functioning of these patients. A lot of people have tried. I think we have been very successful with one study. The second study showed that the highest dose we are testing now in a confirmatory study was also successful. We are trying really to address one of the highest unmet medical needs in the space of psychiatry.
Mm-hmm. You're running a confirmatory study right now. What's the timeline for that?
This study will be in two parts. Part A, which is a 12-week study, treatment versus placebo. This will read out during the second half of next year. Obviously, more we are advancing in terms of recruitment, more closer or more precise can we be. But for the moment, we are really focusing on second half of next year. This will be followed by a second part of the study, which is called phase B, and this is focusing on safety, mostly looking to relapses, because obviously what we are proposing is a paradigm shift. A paradigm shift being that you don't continuously treat patients with antipsychotics, because there is a population out there, which is probably 60% of the patients with a diagnostic of schizophrenia, who can benefit from our drug in monotherapy.
This is part B, is looking to a direct comparison in terms of relapses of positive symptoms between antipsychotics and our drug.
Mm-hmm. To be clear, do any of these antipsychotics out there help with negative symptoms? Or is it. Yeah.
I think the short answer is no, they don't help. They even have a tendency to worsen negative symptoms, because when we speak about negative symptoms, we can speak about primary or disease-related negative symptoms, and we have what we call secondary negative symptoms, which can be induced, for example, by antipsychotics. They don't help, yeah.
I think is a fair answer.
Okay. Then remind us how big the population is in the U.S.
In the U.S., people who have the, how to say, the disease is around 3.2 million people. It is a very, very large population. It is more than 1% of the population. When you think about patients who are treated or seen by the medical system, it is about 2.2 million people. When you are going according to our criteria, having negative symptoms or impairing negative symptoms and having some stability in their symptoms, it is around 900,000 patients. Our target, once the drug will get approved, I mean, it is about 900,000 patients.
I see.
Quite large population.
Right. And so, you kind of hinted at it earlier is, should you be approved as a monotherapy, how do you use this exactly in relation to the antipsychotics?
I think what will happen here is that people will remember why antipsychotics have been developed. Antipsychotics have been developed as major tranquilizers, so to take care about an acute episode of psychosis. Why, I mean, you are prescribing antipsychotics, and I am coming from the clinical side, yes, I mean, I have prescribed antipsychotics as well, is because, I mean, it is reducing somehow the level of relapses, the number of relapses, and the hospitalization. But it is not treating negative symptoms, it is worsening negative symptoms. So, what will happen is that our target population, where we have shown over one year stability of the positive symptoms, an extremely low relapse rate, and I mean, if people are interested, I can show even the slide, is to treat the patients, monotherapy with our drug.
If there is a relapse, you can use the antipsychotic to treat the acute episode of relapse. So, we are going back to the basics of the disease and of how an antipsychotic is working, or where it is the most useful, let me put it like this.
Very clear. Explain the mechanism. Is this a traditional antipsychotic? In the label, when I think antipsychotics, there is black box for increased mortality. Would this get a black box, for instance?
I mean, this is the FDA who will decide this, but what I can tell you is that, I think one of the main reasons why antipsychotics have a black box warning is indeed, I mean, the mortality is increasing when you are treated with these molecules. I think one of the main reasons, or the pharmacological reason, is probably because they are blocking the dopaminergic pathways, and particularly post-synaptic D2 receptors. In our case, roluperidone is not blocking post-synaptic D2 receptors. This was done on purpose, because think one second, when you are blocking dopamine, you are blocking dopamine, which is an important pathway in the limbic structures, in the prefrontal areas of the brain. You are giving zero chance to these patients to improve in terms of negative symptoms.
The alternative to this, in patients who have more stability, is to use a 5-HT2A, so serotonergic 5-HT2A activity, which is definitely has been shown to have an effect also on keeping positive symptoms stable. We get rid of the dopamine blockade and keep the patient stable with the 5-HT2A activity.
Mm. As I think about the regulatory pathway that you have had, there was, like you said, a phase II, phase III. You tried to file initially, it did get rejected. What exactly happened where the FDA ultimately became aligned with your pathway?
Yeah. I think we would have been a me-too molecule. We would have been approved immediately with the two studies, yes, because the second study is definitely somehow confirmatory evidence of what we have seen, yes? Because of paradigm shift, because of the population, definitely, the FDA wanted to have more certitude, and I can completely understand this. What really changed the story was that the FDA organized a public meeting about how to develop molecules for negative symptoms, and guess what? Most of what we did, KOLs, regulators, or experts were confirming that you have to go with stable patients, the treatment duration has to be long enough, it has to be done versus placebo because you have no drug approved.
All this was somehow confirmed during this meeting and afterwards, as I have to say, the very, very helpful and constructive dialogue with the FDA started, yes.
Mm. Very good. Like you said, part A data will be, or phase A data within the phase III confirmatory will be second half 2027. Can you remind us what you saw exactly on the negative schizophrenia scale of PANSS Marder in the prior studies, drug versus placebo?
Yeah. First of all, to be clear, our data cannot be compared to any other data because it is the first one. First time we have this study design has been used. If you want to compare it to when you're treating someone who is acutely relapsed in terms of positive symptoms, when they're coming down to the baseline of negative symptoms, this is a pseudo-improvement. You cannot use this as a reference. You can only use the reference we have generated. There was a company, Lundbeck, you probably are familiar with, who basically did exactly the same study design as we did because they were hoping that the drug has an effect on negative symptoms. It did not materialize, but they have used the same study design.
You cannot at this stage decide how many points do you need to have in order to have a clinically meaningful response. But what I can tell you is that after 12 weeks, we had a decrease of about 4 points on the Marder negative score. When you're looking to the open-label extension, the improvement is continuing over the complete year we have analyzed. Now, what is important, what matters is, does this mean something to the patients, do these patients function better, yes? We have done recently analysis, and maybe I can show them to you, which are here. When you want to demonstrate that your drug is useful, you're going with patients improving at a certain percentage their symptoms, and here you can see an analysis with a 20% improvement, and you see that it is highly in favor of roluperidone compared to placebo.
But we went further down the road, and it is well described that if you are anchoring your primary endpoint to CGI-S, Clinical Global Impression-Severity, again, you can claim that the drug is doing something meaningful to patients. And again, you see in the middle chart, it is really helping patients. And last but not least, in the two studies, the previous study and this study, the sole key secondary endpoint is PSP, which is looking to daily functioning. And when you are doing this analysis with PSP, again, more than 7 points improvement is really demonstrating clinically meaningfulness, and you can see that the drug is doing a better job than placebo. So, we are slowly but surely accumulating data showing that this is definitely transforming into a functional improvement of patients.
Mm. And so your phase III, that said, the primary endpoint is going to be that score change, drug versus placebo. So, what is the statistical assumption or power assumption?
So, the statistical assumption here is obviously based on a delta in terms of placebo and standard deviation, dropout rate, and so on, yes? But so, it will be between 1 and 2 points, so basically. And the study with 380 patients is powered more than 90%, yes? And we have used assumptions in terms of variability, standard deviation, which are above what we have seen in the two previous studies, and the dropout rate is also very similar to what we have seen, yes? And as far as we are in the current study, the things are extremely similar between this study and the two previous studies in terms of dropout rate and variability.
Mm. And as I look at your prior phase II, prior data sets, like you said, the drug effect on an absolute basis was -4 on PANSS Marder. The placebo was a little more variable.
Yeah.
How have you controlled for placebo in this phase III?
This I think is a key question here, yes? And I think we have a drug here, yes? Because when you reproduce a signal in two studies, I think you have a signal. In addition, it is transforming, as I just demonstrated, in a functional improvement. What can you do? First thing we have done here is that we are only testing 64 mg versus placebo, and not 64 mg and 32 mg versus placebo. The expectation to get active drug is going down, and this is well described in the literature. The placebo will be also more under control. The second thing we have implemented in this study is that we have a small committee who is helping the PI when he is selecting the patient.
Before randomizing the patient, the small committee is discussing with the PI just to help him identify the patients who can really benefit from our treatment. So you are reducing the variability at the entry. During the study, completely blinded, we are analyzing the data from the primary endpoint, which is based on a scale called the PANSS scale, and for PSP. And in the two previous studies, we were keeping this at the level of each clinical site, and at each clinical site, you have two, three, four persons doing the scoring. Here in this study, we are going down to the level of each person doing the scoring. It is just to pick up each time someone is deviating from the scoring, in order to retrain the person in order to really reduce the variability of the data.
Another aspect we have implemented in this study, which was not the case in the two previous studies, is that we are keeping the contact with the patients as minimal as possible. The patient is only seen by the person doing the scoring when the scales have to be done week two, week four, week eight, week 12. Why? Because think one second about these patients. These patients have negative symptoms, they have avolition, they have amotivation, they have no longer pleasure to do anything, so they are not bothering anybody. And suddenly you say, okay, now you are the center of a study. You will see a nurse, you will have to see a physician. This has what I call a nursing effect. So you are helping the patient to have some social interactions again, but this is only, obviously, the time the social interaction take place.
In our study, in this confirmatory study, when we interact with the patient, it is not directly with the patient, it is via the caregiver. To avoid this nursing effect and also to keep placebo as low as possible. I think when you put all these bits and pieces together, having really a strong signal in the two studies, having 64 mg showing a p-value, transforming into a PSP improvement, so functional improvement, I think we have all the recipes here to really magnify the delta between placebo and treatment.
Mm-hmm. Speaking of execution, what has the screen failure rate been like? Why are patients failing screening?
They are failing screening for several reasons, but the main reason is about the stability of the patients. Because these patients need to show stability over the last six months in terms of positive symptoms, they need to have stability also in terms of negative symptoms, and they need to have a minimum score of negative symptoms. When you have the patient coming in, you have to check all this, and sometimes you discover that the negative score is at 19 and not at 20. We have several cases like this, and this is, I think, very, very reassuring about the quality of the sites, because between 19 and 20, there is not a big difference.
But I think we have told the sites, be careful. Do your job. Do not really do your job. We had several cases where patients were not selected because the score was 19 and not 20. I think this is really good sign. This is one reason. Some other reasons is that, for example, these patients should not be cannabis positive at selection, because as you know, it has a psychotropic effect, and we do not want to have this as a confounding factor. This is a reason, for example, of patients not selected. Afterwards, there are several reasons. Some people have elevated liver enzymes, some people are not willing to take completely off their antipsychotics. But I think the main reasons are really to stick to the inclusion/exclusion criterion.
Mm-hmm. In terms of dropout rates at the moment, remind us the dropout rate in the prior studies, and how is that looking right now?
So in the previous study, over the first 12 weeks, the dropout rate was around 30%. Just keep in mind that if you compare this to the studies in acutely relapsed patients, the four, six weeks studies, the dropout rate is also 30%. So, it seems high, but it's very low at the end of the day. What is even more interesting, the dropout rate of, when you go into the details of this 30%, more than half of the reasons of dropout is for personal reasons. The family who is no longer able to take care of the son, the patient, as you know, by definition, sometimes he can decide from one day to the other that he's no longer willing to be part of the study. Nothing related to the drug, more for personal reasons.
Mm-hmm. Relapse rates, remind us what that was. Relapse for positive symptoms. Yeah.
Yeah. So, maybe here on the right side of the slide, you have the relapse rates of the two studies. That's in green. The first study, C03 study, was the first study. The second study, the C07, is the other graph in green. So you see that the relapse rate, and this is over one year of treatment in one case for the C07 study, and nine months of treatment in the C03. It's very low, yes. It's between 10% and 15%. Keep in mind that all the meta-analysis which have been done on antipsychotics for one year, these randomized withdrawal studies, the relapse rate is between 27% and 30% in the overall patient population. So, we are much lower, yes. By the way, this was a very, very important discussion with the FDA, because why are these patients less keen to relapses?
I think we have already published recently, and we will continue to publish. What we can demonstrate is that more you have negative symptoms, when you are entering the study, less you are keen to relapse. There is really an inverse relationship between level of negative symptoms and keen to relapse. What I am showing you here, I think, is interesting because this can be very comparable to what we are doing in phase B of our trial. We took the CATIE database. CATIE database is the largest database in the U.S. about studies done with atypical antipsychotics. We applied our eligibility criteria to this database, and we have taken out from this database the patients who have the same criteria as the patients from the roluperidone study.
You can see very nicely that the relapse rate with antipsychotics is very similar to what you see with roluperidone. I think we have already some answers to what we hopefully will see in phase B. So, low relapse rate in this patient population due to high negative symptoms, and also to the stability of the positive symptoms. But obviously, also because the drug is controlling positive symptoms or keeping them at minimum stable.
Like you said, 10%-15% in year one relapse rate for your drug. What was that like in the first 12 weeks?
It was below 10%, yes. Definitely very low, yeah.
Okay.
Very low.
The part A data of week 12, again, it's the MADRS endpoint, or not MADRS, but-
Not MADRS. It's for depression, yeah.
Marder endpoint in second half, that second half of 2027. Then comes the part B data. Remind me when that dataset could be.
It is a one-year part, yeah. It is one year later.
Sure. Second half 2028.
Yeah.
What exactly does the FDA want to see on the relapse rate? Because you are comparing this to three other antipsychotics.
Again, it is clear this has nothing to see with efficacy. Efficacy is week 12, yes? A p-value between placebo and treatment at week 12 is what they want to see to confirm efficacy. Part B is purely for safety and relapses. Safety, when I speak safety, is relapses basically. Because the relapse rate is so low, yes, you cannot power the study for a non-inferiority study, for example, because this would be 2,500 patients. It will be purely descriptive. As long as the relapse rate in the two conditions, antipsychotics or roluperidone, is similar, this is what they want to see, yes.
And obviously, if it stays at the low level we have seen previously, definitely, this will help for the labeling because this will be if you have a patient who is not functioning well, has negative symptoms, you should not worry because you take him off antipsychotics because the risk of relapse is the same, yes, or is low and the same as with antipsychotics.
Right. Because if the FDA wants similar, that is an easy bar because, like we just said, 10%-15% in your studies and out to year one, I would imagine antipsychotics, the relapse rates are higher.
So, as I was saying before, maybe I was not clear. The meta-analysis of the last 10 antipsychotics, it tells you that it is around 27%-30%, yeah. So, when you go with the 100% of the patients, we are going with the 60% of the patients.
Okay. And so, you have been working on roluperidone for a while, so remind us the latest on the patent estate.
Yeah. We have no longer composition of matter. We have seven families of patents, and three of these families can go into the Orange Book. The most obvious one is what is called a formulation patent, which is not really a formulation patent because it is a real recipe to follow, yes. This is going to, and I have no time to go into the details, but it is definitely not a formulation patent. It is a patent where the maximum exposure of the parent should be at a certain level. The delta between parent and metabolite should be at a certain. So, it is a really, really difficult recipe here. This is going to 2038 without the five years extensions and the six months of pediatric in the U.S. and the 10 years in Europe.
As we speak, we have also other good ideas, I think, in terms of extending IP. We are not rushing because I think this idea is good, so we can wait a little bit. But we have some additional ideas in terms of patent extension.
Mm-hmm. If roluperidone is approved, do you intend to market this drug yourselves, or are you open to a BD partner?
We are working seriously on it. We are working really seriously at it. Because this is a paradigm shift as well in terms of marketing, we are really working on how to implement this in terms of promoting the drug, educating the patient, the caregiver, and the prescriber. So, we are working on it. Obviously, the big event which will be the readout of phase A, because we have, as you know, we have raised $200 million in October last year. This will not be enough in order to do all the work you need to do in order to commercialize the drug, yes? So, we are really working hard on it in order to be ready.
Okay. Very good. I think that's all the time we have. Thank you for the update, Remy.
Thank you.
Thanks everyone for listening.
Thank you, Andrew.
Thank you.
Thank you.
Excellent.