Good morning, everyone. I'm Tara Bancroft, one of the Senior Analysts here at TD Cowen. And so thank you for joining us for TD Cowen's 45th Annual Healthcare Conference. So our next session, we have a fireside chat with Celcuity. And from Celcuity, we have the CEO and Founder, Brian Sullivan. So thank you, Brian, for joining us.
My pleasure.
It's a pleasure to have you here. And so before I get started, I do want to mention for anyone in the audience, just please feel free to raise your hand, shout out, and we'll make sure your question is heard. And so to start off, Brian, can you give us a high-level overview of Celcuity and just a general update to begin?
I started the company initially to develop a platform with our chief science officer that could assess and quantify signaling activity in a patient's tumor cells. We eventually began development of capability to measure activity in the PI3K/AKT/mTOR pathway, which then eventually led us to our current lead agent, which is gedatolisib. Gedatolisib is a pan-PI3K/mTORC1/2 inhibitor, highly differentiated, and I'm sure we'll get into this, relative to other drugs in this class. We have three currently active clinical development programs, one in second-line HR-positive HER2-negative breast cancer. We expect to report initial data for one of the cohorts in the next few months. We'll report data for the second cohort, which will be PIK3CA mutant patients towards the end of this year. The next program we have in breast cancer is a first-line study that we're just getting going now. We announced that last year.
We expect to enroll our first patient in the next few months, and that is evaluating GetA with different CDK4/6 inhibitors and fulvestrant, similar to our second-line study, and then our third program is in castrate-resistant prostate cancer, where we're evaluating men who have progressed on their prior androgen receptor inhibitor, and that's a phase 1b study, and so we're at the dose-finding phase, and we'll expect to get initial data towards the end of the first half of this year.
Okay, thanks. And so before we get into the trial and expectations of data, I think what would be really helpful is level setting for people. What is differentiated about GetA? Because the HR-positive HER2-negative space, it's becoming more competitive, and there are various drugs with similar types of mechanisms. So what stands out about GetA?
I think I would step back to the disease itself. HR-positive disease is driven by three different pathways: estrogen receptor pathway, CDK4/6 pathway, and the PI3K/AKT/mTOR pathway, or I'll just call it the PAM pathway. So these three pathways cooperate. So essentially, they can cross-activate each other if one's inhibited in the absence of the other. So ultimately, our view is that the optimal outcome for patients involves simultaneous blockade, either in the first-line setting or second-line setting, of all three of those pathways. And that's the regimen. Those are the regimens that we're evaluating in both the first and second-line studies we have, phase III studies. Now, GetA, as it relates to other PAM inhibitors, is, as I alluded to, highly differentiated. Pathway's role as a driver of disease was discovered roughly 20 years ago, and its structure was deconvoluted. And realized how complex this pathway was.
There are essentially six components that interact, essentially allow this pathway to function if different components or nodes of the pathways are disabled. The initial development in this pathway focused on inhibiting the various class I isoforms of PI3K, as well as mTORC1/2, to comprehensively block this activity, prevent cross-resistance. That proved to be challenging. It's hard to hit six different components equally potently at sub-nanomolar or low-nanomolar concentrations. It also proved to be very toxic for oral inhibitors. You had a migration in this category, in this class of drugs, towards ones drugs that only inhibited a single node of this pathway, let's say PI3K alpha, mTORC1/AKT. That was a way to avoid some of the toxicity, but it resulted in essentially a compromise on your efficacy potential.
Our data has very clearly highlighted the benefit of comprehensively inhibiting this pathway, whether it's prostate cancer tumor cells, breast cancer tumor cells, or gynecological tumor cells. Essentially, GetA is 300 x more potent inhibiting tumor cell proliferation relative to single-node inhibitors. It does so at low nanomolar concentrations, whereas just multiply times 300, these other drugs require micromolar amounts of drug to induce a half-maximal inhibition of tumor cell proliferation. The other interesting feature of the drug that's very important clinically is that, at least non-clinically in our early preliminary data, suggests that GetA is equally active or comparably effective independent of the status of PIK3CA mutation. The other drugs that have been approved in this space have really only shown activity in patients that have PIK3CA mutations.
And that data is consistent with the non-clinical data, whereas our non-clinical data shows essentially almost identical potency and cytotoxicity, regardless of the PIK3CA status. These other drugs that clinically don't show activity in patients who lack mutations also show highly differentiated activity between those who have or don't have mutations. So far, the non-clinical data has been a preview of the clinical data, and that's obviously important. It's nice to have those all line up.
Okay. Next, the burning topic of sort of phase III VIKTORIA-1 trial, it's ongoing, wild-type, fully enrolled, data coming very soon. Before we get into the details of the nuances of the patient population, the numerical comparisons, the context of other trials that have read out recently, can you set the stage for really what we should be focusing on in that data?
Sure. Well, as you guys are probably all aware, there's a variety of drugs under development or have recently reported data. And one of the challenges for doctors when they're trying to make treatment decisions is how to compare these results across trials. And the metric that becomes most important then is the hazard ratio, essentially allows you to see what the reduction of risk is for the disease progressing for these patients relative to the control. And so the controls in most of these studies are fulvestrant, and we would expect to show a very, very favorable hazard ratio, i.e., low number. And if our early phase data comes through, we would show, we think, a meaningfully differentiated result relative to the hazard ratios reported in these other areas. So then how do you translate that to outcomes for PFS? Because that's our endpoint.
Hazard ratio ultimately is the test that is done. We would expect that the control, based on the data that's been generated in randomized studies in patients that are comparable to the population we're enrolling, that the control, which is fulvestrant, would probably be close to two and a half to three months. That's an estimate. It could vary. Three trials have been done in this population that reported two months median PFS. One reported 3.5 months. So it's in that range. The feedback we've gotten from regulators in the U.S., as well as in Europe, from KOLs and from community docs, is that to motivate and to create a new standard of care, you should offer at least three months incremental PFS benefit to your control.
And so as doctors evaluate our data, we would expect them to look at that delta, three months as the threshold for relevance, and then also look at the hazard ratio so they could get a sense to the extent there are different populations being evaluated in different studies. They can compare the outcomes with our data to these other trials' data.
Okay. And so next, I think it would be really helpful to go into the patient population. So could you describe more what types of patients you were looking to enroll in this trial and how that factors into your assumptions of what we could see in the data, especially relative to the phase 1b?
Obviously, the baseline characteristics have a huge impact on the results. In our early phase study, we enrolled patients that some of whom, 20% or so of whom had prior chemo, a third of them were on their third line of therapy, and 100% of them had visceral metastases, no bone-only patients. Our phase III study is actually enrolling a population that has less pretreatment. We expect very few third-line patients. We're not allowing patients to enroll. We're excluding patients who've had prior chemo, and we are allowing patients who have bone-only disease as long as it has a soft tissue component, like a lytic or blastic lytic lesion, and those patients tend to have a more favorable prognosis in general, and so we think we've minimized the risk of some shift in patient population inducing a change in results.
And then you have the normal sample variance that we can't control. Obviously, you control a little bit by your sample size. But all in all, we think the first-line, early phase data is representative at least of what we'll enroll here in the phase III patient population. You could argue it may have at least offset some of the variability that occurs when you're going to a larger study.
Okay. One thing I didn't hear you mention is, so in the inclusion criteria, correct me if I'm wrong, is the length of time that a patient responds in the front line in order to be enrolled in the second-line plus trial. And could you explain why that's important and what percentage of the front-line population that actually encompasses?
Sure. So in our case, we're enrolling all comers. So essentially, independent of the duration of treatment or the progression-free survival period on their first-line treatment, it doesn't matter. Now, we do have a stratification factor that will allow us to analyze patients who have been on for six months at least their prior endocrine therapy or six months or less. Now, you'll see some studies will vary their characteristics. I know one that we'll be reading out soon, VERITAC-2, is only enrolling patients that are considered to be endocrine sensitive. So these are patients who will have received at least six months benefit on their prior endocrine therapy. And so that gets to this question of how you compare these trials, different population.
You would expect patients who received at least six months prior benefit on their prior therapy to probably have a more favorable prognosis in their subsequent lines of therapy, and so that's what's going to necessitate evaluation of the hazard ratio to really tease out the relative benefit of the two different regimens, and so overall then, if we're thinking about duration of prior treatment, if you look at the phase III data for the three different CDK4/6 inhibitors, and you look at the study reports, you'd find that the median duration of treatment for these studies ranges between 18 and 22 months, so these patients were on these prior drugs for a long period of time, which is great. In our early phase study, because we had a third of the patients that were third line, the median duration of treatment of their prior treatments was about 13 months.
We would expect, just because of the inclusion criteria in this study and how it differs from the first line of early phase study, we would expect patients to be closer to that 18-22-month duration on their prior treatment than the 13-month period, so we will expect to see patients who are, again, a cleaner second-line population overall.
Okay. Yeah, that makes sense, and in the data, you will be stratifying that for us, right? Like all of these different stratifications.
Yes. No, I mean, it'll come in phases. I mean, obviously, the initial data that we present will probably just focus on the immediate primary and secondary endpoints, but over time, we'll certainly provide different subgroup analyses.
Okay. Great. And then I think the next thing that I probably get the most inbounds on is trying to think numerically about what mPFS could actually come out in, both in the GetA arm and the both arms, really, and the fulvestrant arm. So one thing I love most about you is how much thought you've put into the context that other trials in the space set. And so the first question is, what numerically do you think is possible for the fulvestrant arm and then the GetA arm? And we'll go into the context of trials after that.
To get a sense of what is likely to be reported fulvestrant, it's, I think, really helpful just to look at the studies that have been done in a randomized setting that enrolled a comparable population to ours. And there are four of them, two are phase II, two are phase III. Three of them reported median PFS of roughly two months, 1.9 or 2.1 months. One of them reported 3.5 months. If you average the three, you'd see about a 2.575 average. If you weighed it, you get a similar number. That's the basis for us believing that 2.75+ or minus a quarter to a half a month is most likely what we would see given the comparability of the patient populations that we're enrolling.
For GetA?
Oh, for GetA. Again, I can only point people to our early phase data. Early phase data, we reported 12.9 months overall PFS that included patients with and without mutations. The 12-month progression-free rate in the wild type population was 49%. It was better in the mutant cohort, which was about 60%. Providing a specific number at this point isn't really appropriate. We can only guide people to consider the data that we've reported previously.
Asked a different way. Do you think it could be better or worse than a novel?
So that's inavolisib is a PI3K alpha drug that Roche has. They've only evaluated that drug in the first-line setting. So it's not our setting. Our setting is second line. And so it's really not directly comparable. But they did show when they combined their drug with palbociclib in fulvestrant, a hazard ratio of about 0.5. So very favorable benefit. So very consistent with what we've reported. I think it provides further demonstration of the importance of inhibiting this pathway and the benefit when you combine it with CDK4/6 inhibition. But it's not really a direct comparator at all because it's a whole different population.
Okay. And the one that I hear referenced the most, I know that your expectations for fulvestrant is based on a pool of different studies, but I think people are most concerned about the post-MONARCH study where fulvestrant was 5.3 months. And so how do you say that you have a reasonable level of confidence that it won't be that?
Again, it really requires you to dig into the details about the baseline characteristics of these patient populations. In the post-MONARCH study, more than a third of the patients or over 30% of the patients had non-evaluable disease. Non-evaluable disease is relevant because, A, these patients don't have a measurable target lesion. So it's very difficult to assess progression. You would typically see, and I say typically, almost in every phase III study, you would see that the patients enrolled are ones that have at least a measurable soft tissue component in bone-only patients. The relevance of non-evaluable disease when you're evaluating or when you're considering the progression-free survival period is that those patients will tend to have a much longer progression-free survival period than patients who have a measurable lesion, essentially have some form of visceral disease.
Data from PALOMA-3 broke that out in one of their subgroup analyses that patients with non-evaluable disease, essentially no measurable lesion, reported three times greater response to fulvestrant than patients who had measurable disease. Since we're only enrolling patients with measurable disease, we think that the addition of that component to the post-MONARCH makes it really not a representative study. I think if you were to talk to oncologists, we haven't talked to any oncologists that take that data at face value. Not that they think there's anything wrong with the data. I just think that's not a representative population. It's not one we can compare to the other studies that were currently being fielded in the phase III setting.
While we're talking about other programs, very curious. Well, actually, I want to ask you about EMBER-3, but I don't want to jump around very much. So let me first ask about doublet versus triplet mPFS because I think people are trying to set the context for what we expect in the triplet and then kind of come back afterwards and say, "Wait, what about the doublet?" And so how should we set up the case?
Sure. Well, we would expect the triplet, GetA, Palbo, fulvestrant to report more favorable results than the doublet of GetA, fulvestrant. And that's why we set the statistical hierarchical analysis to be testing the triplet first against fulvestrant and then subsequently, if that's positive, test the doublet versus fulvestrant. As far as the expectations, we expect it to be a positive study. But beyond that, again, we're just not going to speculate. But we would expect the triplet to be more favorable than the doublet.
Okay. And regarding the outcome, so if you had to put a percentage to it, what would you say that your confidence in the probability of success?
That projection is sitting in the upper right-hand corner, upper right-hand drawer of my desk, and it's only going to be opened the day before. No, it's not something we can really speculate on because we don't know. We can only go off what the data is. Certainly have high hopes, and we'll see soon.
Okay. And then I guess it would be helpful maybe to get into the context of the overall space and the market in the last 10 minutes or so that we have here. And so first, before we talk about specifically the market for GetA, I think one thing that people would really appreciate is some context from the most recent San Antonio conference and in particular, like the EMBER-3 data, the sentiment on SERDs right now, where Orserdu is being used. And so generally, your takeaways from the datasets like EMBER-3 that were at San Antonio and.
And this gets back to one of the earlier comments I made, which is ultimately what matters, what is most correlative to outcomes is how effectively a regimen is controlling the disease drivers. And in this case, there are three. And so EMBER-3 reported monotherapy data for their oral SERD, very, very similar, almost identical to elacestrant or Orserdu, where they showed about a month, 1.7 months incremental benefit relative to fulvestrant as a control, no benefit in wild type. And they did evaluate. And the study was a little interesting because they used their own drug as a control. So that kind of is challenging. But regardless, I think what they showed requires a lot of teasing out because their analysis initially for the control combined the first-line and second-line patients.
And so you have to tease out the effect that the first-line patients have on that overall number. And once you do that, then you can get a sense of how well the drug did, that combo did in patients who are ESR1 wild type. Our estimate based on this interpolation is that it's unlikely that it's more than four to four and a half months. So then you have a subgroup of patients that are ESR1 mutant but PIK3CA wild type. And that represents 15%-20%. That drug may or may not be relevant as we launch our drug. If it is relevant, it's only going to be relevant in a relatively small fraction of the total population. Obviously, in this market, if you just do the math, you'd see that it's a $4-$5 billion SERD market potential. So very significant patient population.
If you have extended duration of treatment for these patients, it can result in a very significant market. We're not projecting, and I don't think anybody would reasonably project to get 80% share of any market. But we can build a $2 billion indication if we can achieve 40% penetration, which we think is very substantial. It's a blockbuster indication. And so I think people have to be careful about thinking any of this is an all-or-nothing game. It's really one where specific populations treated will matter. The relative contribution hazard ratio comparison will be very important because of these variances in patient population. And again, we'll see soon how we do. But if our early phase data is replicated, we'll be in a great spot.
Okay. And next, I think it'd be helpful to get your thoughts on the competitiveness of GetA as an IV therapy versus other oral options that are in development.
Sure. Again, it's important to step back and think about what's most important in any drug that doctors are considering. And the factors that weigh most heavily in their decision are efficacy and safety. Ultimately, their preference will be guided, and the guidelines lay this out by the drugs that offer superior efficacy that balance that with a safety profile that's reasonable, so essentially benefit-risk. The consideration of a route of administration really isn't included in the guidelines. Now, it's relevant potentially to some patients, depending on the logistics that they may face in coming in. But I think if you were to spend any time with a community doctor or doctors in general, I mean, they're trying to find the drugs that'll offer their patients the most efficacy, the longest period of time they can hold their tumors at bay.
And so we don't think the IV route of administration will really have much impact at all on our ability to penetrate the market. And it's important to note that in this space, in the breast cancer space, the largest drugs by far are infused drugs, whether it's in HER2, Herceptin, Perjeta, in the HER2-positive space or in HER2 in the overall space. And so the doctors, the clinics, the whole infrastructure serving these patients is wired around delivering infused therapies. So again, we're not going to introduce anything that requires an adjustment logistically for the facilities of the doctors or nurses who are treating these patients.
Okay. Great. And so we only have a few minutes left. So I want to talk about the other half of the pipeline, which is totally off the radar for investors. Makes sense with a phase III readout. But I think it's coming upon us before people even realize. The prostate cancer data that's expected in Q2, I believe. Can you tell us more about the context of that trial and what our expectations should be for those data?
Right. So we're at an earlier phase in development with the prostate cancer study than we are in breast cancer. This is our first study in prostate cancer. The focus now is in selecting the optimal dose. We're evaluating two doses. One is the same dose that we're using in breast cancer. Another is roughly a third less than that. Our analysis will compare the relative safety benefit-risk of those two. And then depending on the results, we may decide to do additional dose-finding work. It just really is going to be data-driven as far as next steps. As far as what we're expecting, men who receive an androgen receptor switch, essentially a retreatment with a different androgen receptor inhibitor than they received previously, get between three and a half to five and a half months if you look at the data that's out there. So call it 4.75 months.
We would expect the statistical test that we're using to evaluate the data is to look at the landmark six-month PFS data, and so we'll compare the percentage of patients that are progression-free at six months in our study to this historical data.
PFS or rPFS?
rPFS. Well, it'll be rPFS, but it'll be landmark PFS percentage.
Yep. Yep. Okay. Very helpful. I'm sure that people are going to start doing work on that. And so that'll be exciting to see. But I guess in the last two minutes, there is one bigger question that I'd really like to ask. And it's, what do you think is the most underappreciated aspect of Celcuity at this point in time?
We have two phase III studies in one of the largest breast cancer markets. The SERD potential patient population is as large as any population out there that are potentially eligible for treatment by GetA. Our hypothesis really is founded on the ultimate or the underlying drivers of the disease in ways that position us very favorably relative to the regimens currently being developed or either currently available or currently being developed. So I think we should be, every CEO says this, should give it any more credit for the potential value we can create because of, A, our ability to affect standard of care in tens of thousands of patients, and then the financial impact that'll have on the company as we launch those drugs. To do the math.
As always, Brian, this was a fantastic and intelligent conversation. I really appreciate you being here. Thanks, everyone, for listening.
You're welcome. Thank you.