Relmada Therapeutics, Inc. (RLMD) Earnings Call Transcript & Summary

Thanks, everyone, for joining us. Andrea Tan, biotech analyst at Goldman Sachs. With me, I have the team from Relmada, Sergio Traversa, CEO; and Cedric Gorman, CMO. Thank you so much for joining us.

Well, thank you for inviting us in this beautiful environment, and thanks for everyone that is listening to our webcast.

Sergio, maybe I'll start with you. A lot has happened over the course of the last year, where we are now versus where we were last year when we were having this conversation. Maybe to level set here, maybe just provide an overview of what did happen across 2 of your clinical trials where you've had data readouts. And maybe what are the learnings that you took from that with respect to trial execution?

Right. Well, thanks, Andrea, for the question. And it's a very good way to start. And I'll give you the top-down and then Cedric he is our new, just came a few months ago, Chief Medical Officer, will give you more details on what we learn and how we are planning to move forward. As everybody knows, 2022 was a little bit of a challenging year for Relmada. And -- but if I have to make like a general comment, even the -- what happened is widely reflected in the value of the company, as you can tell from the share price that is justifiable and understandable. But if you look at Relmada as a business, Relmada never been so strong. While we have done a lot -- we have made a lot of progresses as we are planning to file the NDA this year so everything else that is except the 2 critical and important Phase III trial has been done, all the Phase I, the special populations, the 12-month safety study and manufacturing is all being completed. So the gap between us now and the NDA with the FDA is 2 positive clinical studies. But that said, it's not a trivial program, but it's definitely feasible. So what didn't go right of last year? And the reason that we feel that the company has never been so strong is that we completed 2 Phase III trial that, yes, they did not meet the primary endpoint. But the data overall and the slice in different patient populations, they look very compelling. And although they are not enough for filing an NDA to get an approval, but they provide a large amount of information. And then we also completed just now the 12-month long-term safety study that provides also some information on efficacy that altogether, they accounted well north of 1,000 patients. So what the drug is the -- has always happened, it's not only 1 factor. There were different factor that contributed to the 2 trials not meeting the primary endpoint. And COVID was a contributor, although I personally don't think was the only reason, but it contributed to the non-positive results. And the -- we outsource a lot of services from outside suppliers and vendors, and that is not the same thing that we have internal efforts in terms of control and the patients enrolled, we are not required to provide medical records. We had more than half of the patients coming from social media. So on this, if I -- if I can summarize in what the results of all this was that we had a very high placebo response, and Cedric can go through that and on the causes and what we can do to mitigate that. And we do believe that there was a certain percentage, probably a large percent of the patients that were not affected by the typical biological-based depression. So all that was the results of different factors. We -- all this is fixable, and we have been working hard over the last few months to -- we made a lot of changes, including a brand-new R&D team led by Cedric. We do believe that especially the second trial in a per protocol was a p-value of 0.051. So we barely missed by a hair to be statistically significant. So with all these changes they have been implementing, we are hopeful and we believe that we can get to positive results. At this point, I always tend to talk about. I'll let Cedric to talk more in detail.

I think you still all my talking points, but everything Sergio said is right. The way I just try and think of it to summarize is there's no denying that the results of the Phase III study were in the context of a very high placebo response, almost 13 points, an MDD that's pretty unheard of. And so it shows you that you didn't control the study as well for measuring. But the drug, nonetheless, outperformed placebo even in the context of a high placebo response by 2.3 points. So then you think about -- I'd like to think of what we can do moving forward in 3 categories or 3 pillars, the 3 Ss, if you like, the study protocol, the sites and the subjects. And the study protocol, the original protocol was attempting to answer a lot of questions in 1 study. And I think that what happened was that led to a very long duration of study of subject visits at site. So we've significantly cut that down with the amended protocol in the new 304 protocol. And so it's much more streamlined subjects in and out, focus on the primary endpoint and not a lot of interaction with the staff. The subjects, as Sergio pointed out, we learned a lot from the completed studies about which subjects were more prone to placebo response. Definitely those who didn't have verified medical records, which we now have as obligatory component of the protocol. And other things, too, about severity of illness that actually turned out that those patients who are the most severe did very well on drug and placebo response was very well controlled. So that's a very encouraging finding that normally you think the more difficult-to-treat patients might be the ones that your drug doesn't separate. We found the opposite. So it's really encouraging from an efficacy perspective. And then finally, the sites. Despite having failed studies, we're an enormously beneficial position of being able to identify the sites who know what they're doing, know how to control placebo response. So we preferentially go back to them. And as you know, 2 sites, in particular, who run unfettered, uncapped really made the difference between a significant study and a nonsignificant study. So now we have measures in place that carefully monitor the sites and don't have arbitrary caps, but take a careful thought to look to those who are very active in enrolling patients.

Maybe a couple of follow-ups there on this idea that the more diseased patients had a better response, does that suggest that the inclusion criteria or maybe the HAM-D baseline score is being modified as you look at patients coming into the trial now? Or how are you thinking about that?

Can I just start and then you -- the real answer. The -- that's a great question. Looking at the data, we do not believe that the drug works better in patients that are more depressed as well. It doesn't -- we don't believe it works better for patients that are about 50 years old or that's what the data show. But all these factors just only point to one direction. Patients that are really affected by biological depression, they do respond well to the drug. So it's not a question of more depressed, I respond better. I'm more depressed more likely to be affected by depression. Please Cedric, go.

I totally agree with that. I think that more severely depressed you are, the less chances you are that it's being -- it's a fake or that it's somehow a situational depression or an adjustment disorder. It's kind of like you're certain of the diagnosis, the more severe you are. You have depression written all over your face. We've had some PIs described. You just look at the patient, you know. In other situations, you have patients who maybe feel they are depressed, but it's a transient thing. And that's the gray area whereby to be absolutely sure that they're truly depressed, you need these other things. You need these other measures, these safeguards, the medical records, the deep probing of symptoms and medical history to truly get at them being a true patient. So again, as Sergio said, it's not that severe depression is the sweet spot, it's that those patients with a true diagnosis of MDD is the target.

And then as you think about your trial sites and you've mentioned here 2 of them were slight outliers. But maybe walk us through the steps you're taking to have better oversight over those particular trial sites that you now have.

I had to let Cedric to answer this.

Well, when you work in MDD for a long time, decades, right, in my case, it's maybe 2 decades now, you become very familiar with sites, site networks that have a number of different sites within their network, which is becoming much more and more common. And you become very familiar with the PIs by name. And so like in any industry or in any career, you know who have reputations as being better performers, more conscientious, more professional, focused on true patient type as opposed to those, which it's a business after all, and we can understand motivations. But you want to be very careful about those that have kind of a reputation to be a research mill. And they'll throw any amount of money on advertising just to get patients into the study. So the oversight that goes into that is making sure that things like radars are consistent over time at that site that there isn't a lot of staff turnover, that the radar that is assigned to a subject over the course of the study remains with that subject over the course of the study because MDD and symptom measurements are very subjective. And if you go changing your radar over the course of a short 4- or 6-week study, you can really throw off the evaluations. You also want to look at site staff credentials, education. Again, we have the benefit of being able to look back at their ability to control the placebo response. You also want to see the timely entry of data into the electronic data capture system. Or are there delays in their ability to put it in? Why are there delays? What are the reasons? Are they responsive and answering to queries in a timely manner? And are the red flags that have been associated in the past in terms of lack of concordance between scales that you're using in the trial or either too fast or too slow to complete a given symptom scale, right? So -- and the oversight is all about being able to contact the site in real time and say this seems off. Can you explain this? Not waiting until the end of the month or the end of a quarter to kind of then collate all your red flags and then try to address them. And so that's where having an internal team within R&D at the company, really make sure that we're on top of everything.

And out of those 3 assets or your 3 verticals or pillars that you've talked about making improvements across, do you have a sense of which one will have the most impact?

Well, I think that if you believe, as I do, that placebo response first and foremost, is the issue with a 14-point change or a 13-point change in the 301 study. You then have to ask yourself, which do I think is going to have the best control of placebo. And I think personally, it's the improvements that we've made to the study protocol. And everything else is incremental because we know now what patients, what subjects are damaging to data, and we know which sites not to go back to. But it all comes back to the cookbook, the recipe of the protocol. And I think the streamlined version both in the amended 302 and the new 304 should be able to address that.

Right. And if you look at from a little bit of a different angle, a good protocol, simple protocol is implemented by a good site, a good site and real good patients. So there is a lot of correlation, right?

Yes. Maybe characterize then your confidence that the issues that plagued RELIANCE I and RELIANCE 3 will not carry over to II and IV.

Yes. Look, we clearly we have the advantage that we understood very well what did not go right. And we have confidence and Cedric has done it before, at least twice that I know. And so yes, we are very confident that we have addressed the issues in the right way. And again, in a per protocol, the p-value was 0.051. So with all the changes, we do believe and -- but also believe that it will make a big difference.

Maybe we can speak a little bit about RELIANCE II, which has been ongoing as RELIANCE I was reading out here. Remind us how many patients had been enrolled prior to that data read. What does that really imply for the patients that have been treated up to that point? And what does it imply then for the outcome of RELIANCE II as you look to complete enrollment and complete treating the remainder of the patients?

I'll give you the top down. And then as always, Cedric, will comment more in detail. But before we implemented the changes in the protocol, I believe was about around 80 patients. So -- and now they're about slightly north of 100 patients. So the old trial will go to 300. So we're not planning to do any early stop or -- and so we will enroll another 200 patients, about 2/3 of the study with implementing the -- all the changes that we have discussed. So the key question is how these 80 patients that have been enrolled before they look like? Well, we have no reason to believe that they will look definitely worse than the 301 study. The 2 sites that generate the issues that have never been in that study. And so that's already is an advantage. And most of these patients have been enrolled after the restriction for COVID were lifted and we do know from 301 that there is a very considerable difference on pre-COVID or during COVID and post COVID in terms of separation for placebo that is mostly due to the nature of the patient. During COVID, the patients, the -- even they presented symptoms of depression, it was -- in many cases, it was a reaction to the COVID situation that usually does not respond well to drug or let's put it in the other way, it responds very well to placebo because you are not biologically depressed. And so as soon as your situation changes, you respond to anything you're taking. So there is a good premises by these 80 patients or 100 patients that have been enrolled until now. They could look better than 301. And we also count a lot of the 200 patients that will be enrolled that have been enrolled from now on. And Cedric, please go ahead.

Yes. And as a reminder, in 301, drug outperformed placebo by 2.3 points. And I think that going to a larger and in that study, which we're planning to do in 302 could very well likely have resulted in a positive study. And so as Sergio said, there's no reason to believe that the first 100 in the 302 study are worse than 301. There's actually some reason to actually believe they may be better. But of course, MDD trials are risky. We don't know, it's blinded. But we're certainly much more cautious and careful in monitoring everything now. So I think that we ran very statistical computations, which justified and made sense that we would see this study through to completion because again, 301 would likely have been positive, it has gone to about 300 patients.

Maybe remind us what the assumptions are on the profile for REL-1017 as you think about the first 1/3 of the patients being coming from this initial population, the second 2/3 coming with this amended protocol, what were the underlying assumptions on that profile that led you to think let's keep going with the study?

Yes. Well, the -- myabe Cedric, do you want to answer that at this point?

Yes. I think that we obviously understood how 227 subjects in a sister study 301 performed, and so we had that data. And we also have the protocol, which is a prespecified analysis where the drug further separated from placebo and the p-value was just shy of 0.05. So we plugged in those assumptions in terms of change from baseline, right, and drug placebo difference into the first 100 and worked out calculations what would be required of the next 200 in order to bring about a positive study. And I think when you think of all the measures and safeguards that we're putting in place for a more tightly controlled trial, that we were very confident and comfortable in believing that we can bring about the required placebo difference in the subsequent 200 even to assume the first 100 are performing pretty poorly. So we looked at all those different assumptions for the first 100 and feel that it's eminently doable based on our post tox, based on the changes we're implementing on the ongoing 302. So for the perspective for the next 200, we can very prospectively control for the subjects that we feel have a better chance of separating.

Can you put a number on that, what you maybe are assuming in terms of a delta? Should we assume it's more similar to what you saw in the Phase II trial?

Well, I think that the Phase II effect size was very high. There were like 20 subjects per arm, and it was of the order of 0.7. And I just think that going into Phase III, you have to be willing to realize that you can't have the same degree of control in a larger outpatient Phase III study. And so I think that certainly close to Phase II, but being willing to take a little bit of a haircut in order for the study to come in positive, just realizing that we don't have the same control.

Perfect. Maybe remind us on RELIANCE IV, which is the newly announced Phase III program that you're initiating.

So it's more similar than different. That's the most important thing because you run a trial in order to replicate the first one. So the primary endpoint, primary analysis all are the same. But what you can do is you can reduce many of the exploratory academic-like investigations. And if you're opting for a clinical global of impression scale or you're opting for a patient-reported measure or a quality of life measure or a functional impairment measure that you opt for ones that are much more discrete and short, just to really reduce the time. So again, 4-week study, 2-arm, same primary outcome of the MADRS total score at 4 weeks. And just a few changes on the other exploratory endpoints, so that it's not an exhaustive time spent at the site.

And then remind us where you stand in terms of initiation? And when can we start to maybe expect some of the data from RELIANCE II and IV.

So given that we're about 1/3 of the way in with RELIANCE II, and we've got another 200 to go and given that we're just starting 304 with a target of 300, you can think of the completion of RELIANCE II as being the middle of next year and the completion of 304 as being 6 months after that or towards the end of 2024.

Perfect. You've alluded to this that post-hoc analysis and subgroup analysis from 301 or RELIANCE I has confirmed the profile or the benefit and the clinical benefit of REL-1017. Maybe share with us the learnings that you've found from that data set that continue to make you so constructive on this asset moving forward.

Yes. We have a lot of data now from -- especially from 301 is 227 patients, of which the per product were 198. So there were 29 patients that did not pretty much complete the trial, so that they are called major protocol violation. And on the per protocol, it was very close to be significant. And we analyze this data in as many ways as possible as a prespecified endpoint like a female versus man, age. And there was an interesting one I will touch in a second lastly affected patients. And no matter how we slice this data, the drug has always performed the same, so very, very well with the delta from placebo in the high double digit. So there is only not one factor that suggest that the drug has an effect very similar or the same as in Phase II. It's a multiple. And the -- again, I take the risk to repeat myself, but all these indicators, they only show one thing that if the patient is really affected by biological major depression, it responds well to the drug. I mentioned about Tachyphylaxis. And many trials, the design tend to exclude patients with Tachyphylaxis. A patient with Tachyphylaxis is a patient that taken antidepressant, they could respond initially and then the response rate up or disappears after some time. Usually, these patients are more difficult to treat because they already responded to one drug and then the same drug doesn't respond anymore. Well, the in patient effect by Tachyphylaxis, the p-value was high statistically significant. We did not exclude them and the results are very, very good. They only show you one thing. It's not the patient with Tachyphylaxis respond better. A patient that has taken a drug and tells you that the same drug doesn't work anymore, it's much more likely to be affected by depression. And so it's a real basis. So you put all of this together, it's pretty clear not just to us. We have a little bias because we have been working on this for well over 10 years. But we asked also the KOLs to avoid to have the bias, and they all confirm that the signal is there, and they usually say, "Please don't give up. We need something new."

Maybe on that consistent response that you see in the Tachyphylaxis patient, how much do you attribute that to the mechanism of action of REL-1017?

Well, that's the -- if you ask me personally, I think none. It's purely due to the fact that the patient is really a real patient. I don't think there is any synergies between an antidepressant, even -- especially an antidepressant that doesn't work anymore and REL-1017. The angle is that it's easier to show an effect because you do believe you do know that, that patient is real. And then you see the effect. We don't think there is a mechanistic reason that works better.

I think that people will acknowledge that it's not understood truly how antidepressants work, right? It's all hypothesis and we've had for decades the monoamine deficit hypothesis. I think that it's very interesting and tantalizing that whatever REL-1017 is doing as an NMD receptor antagonism -- antagonist, the true effects lies pretty close to that, which we saw in the Phase II in a very controlled inpatient, very sick, depressed patient population. So I think that it remains to be seen. But there's no reason that REL-1017 couldn't potentially be a more effective oral once-daily agent. You need head-to-head to really uncover that. But I think we have an opportunity in a really tightly controlled MDD program to -- it's on us really to allow 1017 to elucidate it's greater effect size. And I think it's exciting to try and do that now.

Cedric, maybe from your experience speak to the importance of the rapidity of action for REL-1017.

Yes. I think it's very important. I think that with the monoamine, with the serotonin, dopamine and Ephedrine reuptake inhibitors, you're waiting on average 6 to 8 weeks for a treatment effect. People are clamoring for faster, faster, faster. But I don't think that faster necessarily means you keep dialing back one day to put in a time point that's like one day faster than the next agent. I think if we get to a point whereby within 1 or 2 weeks, you're getting depression and control, I think that is fast enough. You know what I mean, unless you're actively suicidal or your hospitalized and you're now eating where you've got a psychotic depression. But otherwise, I think at some point, there is a time point that's fast enough. And I think that our drug with the NMDA receptor antagonism and is headed in that direction. It's important to be able to sell for that with an oral agent. And again, I think the once a day is an attractive option as well.

Maybe in that context then, help us understand how you see REL-1017 fitting into a competitive landscape where you do have stages alone, also benefit occurring very, very quickly and potentially a once or twice annual treatment, you have axon with. Where does REL-1017 fit in amongst these agents?

I'll give you more of the commercial perspective and Cedric will give you the more clinical perspective. And from a conversion point of view, right, the NMDA doctors, they want new mechanisms of action, right? They have like 20-plus monoamine reuptake in SSRI, SNRI. So the NMDA that is pretty much a reality is esketamine and this REL-1017 is a new mechanism. And as you can see from the performance of the competition, I made the name. And the uptake is pretty nice. So there is demand for new mechanism of action. And you go from like one is an in-clinic nasal administration that you have to go the first month twice a week in a clinic and with restriction about driving and the ramps and so on. And at the other side, you have a combination, fixed combination that it can, but it's very unlikely would be used a lot in combination with the term drug, right? And you go to REL-1017, it is once a day extremely well tolerated. And we do believe it's rapid acting and effective. So of the NMDA antagonist currently available, and I'm not aware of anything late stage that is in development in this category. It chose to be like the best-in-class. And on other classes in the market is so large, so demanding for, right? It's very highly unsatisfied. I mean half of the patients don't respond to everything available and the 1/3 respond well. So you have 2/3 of the patients at something better. So there is space with different mechanisms of action. And we see the competition that you mentioned, how well the -- if it will get approved, that consensus that it will. And I will perform in the marketplace, but there is a lot of space for new mechanism. And from a clinical perspective, Cedric?

From a clinical a perspective, say we -- patients need more options. We know that patients respond differently to different treatments. There's a whole dilution of the SSRIs and SNRIs. We're relatively new into the available options for NMDA receptor antagonist. But bear in mind that if REL-1017 is approved, it would be the only NMDA receptor antagonist approved as an adjunctive agent MDD, basically opening up an entire treatment landscape regardless of what background antidepressant the patient is on.

Maybe I'll throw one at you, Sergio. Axsome this put out morning, guidance for Auvelity amongst other products with the idea that in major depression disorder, $1 billion to $3 billion peak sales for Auvelity.

Yes, we do like that.

How is your internal expectations for REL-1017 compared to that?

The final answer of the internal expectation will be after we see the data and the profile and the label. And we have done some initial work on the commercial side. But look at the -- I would avoid the number for now. Let me state that with -- I mean, if our colleagues and competitors they are right on the numbers, but we don't see why have the best-in-class should not be somewhere around. Let me leave it at that.

Sure. Cedric, maybe we can touch on the open-label safety study that is wrapping up now. Help frame expectations for the data that you'll be sharing with the community.

Yes, I think the open-label data sets are always a great insight into the real world and how patients may respond when they're not quite in that controlled environment, you also have to realize that they're on other medications. They're coming in less frequently for assessments. So it's in that context. But I think what it gives you is a nice snapshot over time for a longer duration than the short-term efficacy trials, which can show out to 12 months how well tolerated the drug is, how patients continue to stay well over time. And we look forward to sharing those data with you as soon as we have them.

Do we have an expected time frame for that data set?

I think over the summer, we should be able to. The study shut down, we're cleaning the data, and we should be able to generate reports in the next couple of months.

It's over 600 patients, so there is a lot of data.

Maybe to that point, we've spoken that the patients that are within this study come from multiple different RELIANCE I, RELIANCE III, some de novo patients. I guess maybe what will you be looking for from this broad data set as you look forward?

Yes. In the 12-month study, including the de novo patients, the patient, it's a real-world experience. So the patient do anything he wants. Take out the drug, combined, not combine. So the -- I would say the 2 things that we expect to learn is, one, safety. And I mean, we do believe that there was an issue, we would know it already. So we assume that safety is good. And a signal of efficacy because the -- ultimately, that's the way the drug will be used, and we expect to have similar data compared to all the other trial, the drug has always looked the same profile. And maybe one information that is very useful, especially in light of the competition is duration of efficacy, right? The drug continue to work over time for 12 months.

We're right up on time, but let me just ask you one question on safety. Do you believe the question about abuse liability has been answered at this point?

We do believe so. Yes. We have performed the largest abuse potential study or program ever done and the data are very clear. It's a profile placebo-like. So in terms of a view and so forth.

Perfect. With that, Sergio, Cedric, thank you so much. Thanks, everyone.

Thank you, Andrea. Thank you, everyone.