Celcuity Inc. (CELC) Earnings Call Transcript & Summary

We're going to get started. Hi, everyone. I'm Tara Bancroft. I'm one of the senior biotech analysts here at TD Cowen, and thank you for coming to TD Cowen's 46th Annual Healthcare Conference. For our next session, we have a fireside chat with Celcuity. And from Celcuity, we have the CEO and Co-Founder, Brian Sullivan. Thank you so much for being here, Brian. It's such a privilege to have you.

That's my pleasure.

[Operator Instructions] So Brian, I guess, to start off, can you just give us some high-level thoughts and a general overview before we get into details?

So Celcuity is a clinical stage company focusing on development of drugs that address the PAM pathway, PI3K/AKT/mTOR pathway. It's one of the most important pathways in oncology. We have 3 trials underway currently, Phase III trial evaluating GEDA in combination with palbociclib and fulvestrant in women who've progressed on their prior CDK. We have data that we'll be reporting out -- a second set of data reporting out in the next few months. We have a first-line study in process with the same drug combination in women who are treatment naive, who have endocrine-resistant metastatic breast cancer. And then we have a third study at an earlier phase investigating data in combination with an androgen receptor inhibitor in men with castration-resistant prostate cancer. And a lot of other stuff going on as well. So we'll talk about that.

You sure do. Okay. So let's start with talking about launch preparation. You had the NDA acceptance, PDUFA date in July. So as that review is ongoing, I know you have the RTOR designation. So maybe you could start by talking about the various benefits of that. I know it's a somewhat uncommon thing to get. So maybe you could familiarize us with the benefits of that package.

Sure. So RTOR stands for real-time oncology review. Essentially, it's really only given or used for drugs that in the FDA's view offer a promise of changing the standard of care landscape for patients. And we weren't surprised by getting it because we had breakthrough designation and the breakthrough designation reflects results from earlier phase studies. The benefit of RTOR program is that it allows you to begin submissions prior to your completion of your formal submission of your final submission. So we started providing the FDA data sets within a month after we had our data and then provided that on a rolling basis. We have a priority review. So essentially, PDUFA will be 6 months after they accepted our NDA. And we think it's unlikely that the review will come in prior to that PDUFA date. And I say that because of the 20 or so RTOR reviews that have been done in the past 5 or 6 years since they put this program in place, drugs that have a profile similar to ours, first-time submission, priority review, full approval, not just an accelerated approval. For the most part, have gotten an approval decision along the line at the PDUFA date. Where you'll see the real advantage in terms of short time frames is for supplemental NDAs. And so if our mutant data, which we expect to report out end of this quarter, early next -- sometime next quarter, is what we hope, then we would file a supplemental NDA. We would also seek RTOR and hopefully, in that case, we would see a shorter approval time line.

Okay. Great. And maybe just briefly, we exist in a pretty volatile FDA environment. So maybe a little bit on your thoughts on confidence in approval in this kind of environment.

We haven't seen any change in our -- the type of interactions, the quality of the interactions, the timing of the interactions over the past year. We've been interacting with the same group, same people for the past 5 years. And relationship has been very collaborative and all the meetings we have regular interactions with the agency. All the interactions typically are scheduled, i.e., they're supposed to occur within a certain time frame. There has been no diminution or delays that have occurred. So from our standpoint, we wouldn't know there was an issue just based on what we're seeing. So we don't expect any anomalous outcome that could be influenced by some of this other stuff that's going on.

Okay. Great. Okay. So how about launch preparation? Where are you at with that? Maybe some numbers on sales force, MSLs, anything else that you...

So we actually began preparation for the launch about 2 years ago when we brought on our Chief Commercial Officer. And in turn, that led us to build out the senior team reporting to him that would be responsible for key functions in the commercial department. But we also developed essentially a company-wide launch plan that not only involves building out the team over time and have a plan for that and the projects associated with establishing your infrastructure commercially, but also laying out the projects required for all of your other departments, whether it's IT, a huge number of systems you need to put in place that involve IT. Your safety department has to substantially change to be ready for commercial reporting of safety events. HR, finance, every function essentially gets affected. So we had a company-wide launch plan or have one, and then we've been implementing against that. We've pretty much completed the hiring of the commercial organization by the end of last year, except for the sales force. We did hire the regional sales managers prior to that. And so now we've identified -- we know how many we need. We've already identified the great bulk of those that we would be making offers to, and we would expect to begin bringing the sales reps on second quarter. And so I would say we're directly absolutely on track to being able to launch effectively around the time of the PDUFA date.

Okay. And then maybe a little on launch expectations, the cadence of what use could look like in the beginning because you're in a little bit of a unique situation where you expect to get approval in one segment, but a little bit later in another. However, the data is going to be out for that. So just curious what types of patients and then how many you think it could be used right away.

Sure. So you can think of this market as having 3 fairly discrete segments. Obviously, the most common one or rather the broadest one is just to use PIK3CA status as the determinant, right? PIK3CA wild-type patients are roughly 60% of patients. So that's what we'll focus on. We'll have data in the mutant cohort. That data will actually, we think, be available, be public prior to launch of GEDA. So that will certainly be wind at our back if it's what we hope it is for the launch. And so we'll be focusing on women who are PIK3CA wild type. Our data is very favorable relative to drugs that are currently available. And we think we've done a very effective job at being able to analyze the physician health care practitioner profile, so we are able to prioritize where we should put our efforts, how to design the sales territories, how to define the call patterns that we want the reps to use and also then how to build out the messaging and how to make these individuals effective proponents of GEDA and recognize that these doctors are used to alternatives. So part of our job will be to help them understand the context of GEDA and how it relates to what they've done or what they might do in the future.

Okay. I do want to get into a little bit more on which segments and types of patients that it can be used in and that you have confidence it will be used in. But I think everyone in this room and listening probably wants to hear expectations for the mutant data. So the first question that I have is with the late Q1, Q2 guidance, we're really coming up on that very, very shortly. So what can you say about your confidence that, that time line will not be delayed? Are you very close or have reached events already?

We haven't changed our guidance, and I don't think we will be changing our guidance with our quarterly call. So we're confident that that's the time bucket that we'll have data to report.

Okay. Got it. All right. So expectations, I think let's first start with the control arm. I know we have some benchmarks with BYLieve and EPIC-B5 in the 7.3, to 7.4 range, but maybe talk about your expectations for what we could see in the control arm using those benchmarks and others.

So there's always a question of is there a risk that the control arm could overperform relative to whatever assumptions you were and how would that impact you? We think just based on data that's been reported with those 2 studies that reported almost exactly the same results. We think it's unlikely that there would be a surprise on the high side of that. Also, if you look at the data for another drug that hits this pathway called capivasertib, that drug reported 5.5 months in a similar population. And it also had a control fulvestrant that was about the same level that we would expect in the mutant population with our study. And that drug has a similar hazard ratio as alpelisib, roughly 0.5. And so that would be an example or rather, that would be a case to say, okay, well, that further buttresses the idea that it's unlikely to be outperforming what that drug has done previously. So that's one -- that's the base case -- rather that's the foundation for any assumption you make about our drug. And so there's really kind of 2 aspects of expectations. One is what do we need to be successful. So that's a statistical benchmark. And just based on how the math works, roughly 10 months or so median PFS would be Statsig. And that would only be 10% or less better than what we reported in the wild-type population. So we think from a probability standpoint, that seems like a highly probable outcome to at least hit that threshold. Now then the question is the clinical significance, -- how meaningful is that number? Well, alpelisib is no longer really being used very frequently. It's market has primarily moved towards Capi, Capi reported 5.5 months median PFS. And so if we're offering at a minimum 10 months relative to 5.5 months, we think that's a great advantage. We think our safety profile will also be an advantage. And certainly, anything better than 10 months would be better. One frame of reference is the Phase Ib data that we reported at ESMO. And this Phase Ib study evaluated a very similar patient population as what we're evaluating now. And in the wild-type population, it reported 9.1 months for wild-type. And that was almost exactly what we reported in the Phase III study, which was 9.3 months. In the mutant portion of that study, patient population, we reported 14.6 months. And so we all know what we're going to see in the Phase III, but at least it gives us confidence that the 10-month number is highly probable. Essentially, it just leads you to kind of put a high number for your probability of success and gives a sense of what your range could be. I'm not assigning probabilities to other numbers, though.

Yes, of course. I mean within that range, though, I mean, 14 to 19 months, it's quite a range. So maybe which of those different segments, do you think that the Phase III population most closely mirrors?

Well, I would love it if it mirrored the 19-month one, but we'll see, right, in a few months. But we'll see. I think it's a reasonable enough sample size that it gives a sense of what's possible. There's also one other number independent of our study that is useful to reflect on is that with all the other studies done evaluating inhibitors of this pathway where they've generated data, patients who lack mutations versus those that have mutations. They've seen significant difference in outcomes for patients, i.e., better outcomes for patients that have mutations versus those that don't. And so alpelisib in this setting reported the same results as we reported in the wild-type setting. And so it's very much the case that it's reasonable, we think, to think that the drug will do better in the mutant than wild-type. That's what our Phase Ib data showed. So all those factors, I think, give us reason for optimism about what the outcomes will be.

Okay. And how about making a comparison to mutant-specific inhibitors? I mean, do you need to be over that 11- to 12-month range to be competitive with them and gain comfort.

To be competitive with a 5.5-month Capi.

I'm more so talking about relay, [ Scorpion.]

Again, those drugs, maybe they get approved, maybe they don't 3 or 4 years from now. So it's not top of mind. I think there's -- ultimately, when you're assessing the potential for a drug, it's really a function of what the relevance is of the target, right? And typically, there's only so much biological potential for inducing antitumor effect associated with the target. And we saw with the SERDs that -- what is it the book eternal optimism or something, where every SERD coming down that was reporting data was going to be the one that reported positive wild-type result. And after 5 or 6, you'd start to think that optimism would dim and it doesn't seem to. And I'm not being critical of those SERDs other than to say, this is an example where lacking a mutation there's only so much potential you can get out of hitting that target. You can't get more than 100%. And so there's only so much potential, I think, that you can get when you hit the component of this pathway. And certainly, to optimize the impact of inhibiting this pathway, you should comprehensively inhibit it. And we've shown non clinically very clearly what the benefit is, and then we've shown clinically what the benefit is. So I think whatever our drug can do, I think it would be better than whatever a single target drug could do.

Got it. And then I still want to dig in a little bit more on what that mutant specific target is. I mean, I know we've seen Phase II data, but maybe you could tell us a little bit more about the impact of baseline with measurable, nonmeasurable disease and how much that actually impacts...

Well, I think when you're looking at data sets, particularly early phase data sets, when you're trying to project what that might imply for a future readout in the Phase III setting, you need to pay attention to the composition of that population. And one factor that can have a big impact on PFS is whether or not the patient has bone-only disease versus measurable disease. And in pretty much every study that I've seen that has an endocrine therapy component or monotherapy, the results for patients that have bone-only disease can be typically 3x to 4x better than the results in patients that have measurable disease. So for instance, PALOMA-3 broke out that data sort of wow, 3x better results for patients with bone-only disease. We saw that, although a very small sample in our own study where the baseline was 2 months, bone-only was 8 months. And so if you have a population in an early phase study that has a high proportion of patients with bone-only disease, then you need to factor in the impact that, that would have. If you've got a cohort of patients that are generating 8 months of benefit from fulvestrant. Well, that's going to goose your numbers and can potentially make them not representative of what you'd see in a Phase III study. And so again, as you -- as people think about what the biological potential is of a single component inhibitor of the PAM pathway, the prior results are probably more instructive than in those Phase III studies than anything else.

Okay. Great. So I want to move on to safety. And before I ask more specific questions, I do feel like we came out of ESMO in San Antonio with a little bit of misunderstanding. And not -- I'm not talking about from the Street and investor perspectives, but really among KOLs on the impact of the safety profile that you saw in GEDA's data. And so maybe you could give us some context there on.

Well, we were able to put some specifics. I think one of the big surprises for doctors was that GEDA induces very, very little hyperglycemia. So drugs in this class are typically associated with inducing high levels of hyperglycemia, can turn oncologists into endocrinologists. It's very tricky to manage patients. It's particularly important in this setting because roughly 50% of women with advanced breast cancer are either prediabetic or diabetic. And so they're particularly sensitive to any impact on the glucose system, metabolic system. And we reported less than 10% overall, 2% grade 3. Alpelisib induces 80% hypoglycemia. So orders of magnitude difference. So that was a surprise to the good. And I think the -- on the other side, stomatitis was higher than we had expected to be frank, we thought the dexamethasone would have a bigger impact. But then as we dug into the data and then presented some of those results at San Antonio, we saw that, okay, this kind of makes sense. If you get stomatitis, it's going to manifest quickly, first 2 doses, days 1 and 8. And then it resolves to a lower grade within a couple of weeks. And you don't really almost have any grade 3 by the end of a second cycle in your overall population. And that was -- the fact that the stomatitis was really not impacting patients' perception or experience of treatment burden was really demonstrated in the quality of life patient-reported outcomes that we were able to report. And so probably the best metric for assessing how a drug is impacting a patient's sense of well-being is this one score that essentially asks patients to rate how they're feeling in 5 different parameters associated with their health and general sense of well-being. And we showed no degradation over an 8- cycle treatment period, which was a follow-up period in that score. So essentially from baseline to end, there was no change. And so if you had an adverse event that was seriously impacting their quality of life, their sense of well-being, you see a diminution in that, and we did. And then when you see only a 2% discontinuation rate, which is what we had for this -- which we read out, that essentially says patients can stay on this drug. And then qualitatively, we've heard from investigators that have said their patients will say unprompted, I don't feel like I'm on a cancer drug. And this relates to the PK profile of the drug. We're only needing to drug these patients 3 times a month. Drug stays at sufficient concentration to induce antitumor effect during that off period. So 25 days out of 28, they're not getting the drug, which means they're only getting that hit with that Cmax concentration of drug 3 times a month. And whereas an oral drug is hitting the patient every day typically or at least 30 times -- 20 days a cycle -- 20 times a cycle. And that is what induces the sense of burden that a toxic anticancer drug can impose on a patient. And so that's why we think the quality of life data is what it is, is that essentially for 25 days out of 28, you're not on this drug, so to speak, physiologically, the concentration is below a level, below the IC50 level for healthy cells. It's above the IC80 level for tumor cells. And that's this very specific window that GEDA is able to fit in. And in turn, that's what we think leads to a very, very favorable safety profile. We think we'll absolutely win on that basis.

Okay. Great. And any reason to believe that mutant and wild-type data would look any different?

No. I mean in the early phase data, we kind of looked at that. That was part of our NDA submission, providing that analysis. And there's no reason for it to be different physiologically, but there wasn't from an actual readout standpoint.

Okay. Great. Okay. Now I want to talk about the overall market. And aside from the few months that it's going to be on the market with wild-type, potentially without mutant yet, let's maybe -- I think it's easier to talk about segmentation of patients in everybody. So there's a lot of mutations and PIK3CA, AKT, ESR1. Can you break down your assumptions for what the peak use could be based on those segments? And in particular, I'm curious what you're hearing and what you think about priority of mutations, like if a patient has PIK3CA and AKT and ESR1, like which mutation gets treated first?

Right. So we've obviously done a lot of research with doctors, quantitative and qualitative to get a feel for how they think about this and then in turn, prompt us to dig in and do some quantitative evaluation of those comments. And what's very clear is that doctors do, at least today, prioritize the PIK3CA as the treatable mutation that they'll use to make a decision. And so that's the 40% bucket that I mentioned earlier. Now with ESR1 mutations, you get a further bucketization of the segmentation of this group. ESR1 wild-type, PIK3CA wild-type patients. So essentially, they lack biomarkers. That represents about 40% of the population. And so essentially, 80% of the population is one where we think we have a very, very compelling advantage. For patients who lack mutations of any type or those mutations, there really aren't great alternatives and GEDA really does offer a very, very meaningful benefit for them relative to what's available. So where you get more potential competition is in the patients that have ESR1 mutations and lack PIK3CA mutations. That's about, again, 20%. There'll be some drugs that are likely approved or some regimens by the end of this year. And we think our data stacks up very well against their data, but there'll be more relative competition. And so I guess I would say we have clear advantage, clear advantage in 80% of the market. And in the 20% of the market, we'll have a bit of a dog fight.

And even then, any idea of potential third line use? What have you heard about...

No. I mean, obviously, doctors can decide to sequence the therapy. There's no reason why -- if they elected to use an alternative regimen, why they wouldn't necessarily consider it as an alternative in the third line. I would say one other observation that's pretty clear from the research is that doctors do want to spare patients from chemotherapy as long as possible if there's an appropriate and efficacious -- sufficiently efficacious endocrine alternative. We think our drug regimen offers that efficacious endocrine alternative. So even if it isn't used in the second line, we think the data would warrant use of it prior to, let's say, in HER2 or some other chemo.

Okay. Great. And I know you also have this frontline trial that's ongoing. So maybe you could talk about what your assumptions for use there are. And in that, there's a lot of discussion and debate about LIDERA and persevERA and how those regimens can impact potential getter use in the frontline and even potentially second line. So maybe...

Okay. Right. Okay. So you have really talked about 3 different groups of patients. There are patients that are early breast cancer getting adjuvant treatment. That was the LIDERA study. Really won't have any downstream effect on advanced breast cancer. Over time, you might see some diminution in the number of women who progress or rather who recur to metastatic disease. But you also have a countervailing factor, which is the increase in overall incidence of breast cancer. So net-net, I think it should be 10 years before you might see any effect. And if you do see any effect, it would be nominal. So then the other population, the 2 populations of women who present with metastatic disease. The larger group, it's about 60,000 women a year, present with either endocrine-sensitive disease. These are women who got their adjuvant -- rather got their adjuvant tamoxifen or whatever, and it was years before they recurred, at least more than a year or the de novo. They just were diagnosed for the first time with metastatic disease. They didn't have early breast cancer. The second group, it's about 35,000 patients, similar in size actually to the second-line population are women who have what's considered to be endocrine resistant. And these are women who didn't get much of a benefit from their adjuvant endocrine therapy. They progressed while they were either receiving it or within 12 months. That's the study that we're doing right now, VIKTORIA-2. And so we're evaluating these women who today, if they're getting a CDK4/6 and fulvestrant, which is the standard of care for most of these patients, only get about 7 months median PFS. So it's almost like a second-line study in some ways. We have data from the endocrine-sensitive population in our early phase study that was very, very promising. The women who are endocrine-sensitive, we kind of delayed recurrence from early breast cancer when they are given the standard of care, which is a CDK4/6 and letrozole, get about 25 months median PFS. We reported 48 months median PFS in that patient population. Small samples, I mean, 41 patients, not nothing, but at the same time, just a single arm, 80% response rate. So clearly, we think it demonstrated, if nothing else, that this pathway, the PAM pathway is relevant intrinsically in this disease. It's not an adaptive response to prior treatment. And that's consistent with the mutational pattern that you see. The percentage of patients with PIK3CA mutations really is no different in the first line versus the second line, i.e., it's an intrinsic mutation. And this pathway, because of its role, we think, is intrinsically involved in the disease.

Okay. Great. Kind of a tangent, but I kind of want to ask you, do you think persevERA is going to work?

So if you were placing bets, I think you take into account 2 other studies. I mean one would be a study with a SERD that had a similar hypothesis using -- combining an oral SERD with a CDK4/6 inhibitor and comparing it to CDK4/6 and letrozole. And that study didn't get through interim analysis. And the interim efficacy analysis resulted in the study being discontinued due to futility. So that obviously suggests that it's not a slam dunk. Now not all oral SERDs are the same, but they all are in the same class, and there's only so much potential that you can get out of a target, even if one is better than the other. Second data set relates to -- what was my other second data, was PARSIFAL study where they compared palbo fulvestrant versus palbo letrozole. The hypothesis was, well, more direct engagement of the target, hitting ER directly with the SERD would be more efficacious than the indirect effect on estrogen with aromatase inhibition. And numerically, the letrozole CDK palbo did better than fulvestrant, not statistically significantly different. But it suggests that just having a good oral SERD or rather good SERD doesn't do much. And the other, I guess, third set of data that you'd point to would be that in patients, tumors lacking mutations, there really is no difference between oral SERD and fulvestrant in terms of efficacy. And that's the patient population that you're treating in the first-line setting. And the question is, does the delay in onset yield a meaningful benefit. And that's what we'll see. [ SOLAR ] study didn't show it, but we'll see. But certainly, it's a bit of a bank shot.

Okay. One more question before we go, but what do you think is the most underappreciated aspect of Celcuity?

I think if people really step back and understood the size of the patient population that we think we'll eventually be able to treat, right? We have 37,000 patients that will be eligible in the second line. I think ultimately, we will progress with trials, and we think high probability of success in these trials in the first-line setting. And if you do the math on the potential impact in terms of treating that number of patients and the revenue that we generate, I mean, we have $10 billion plus higher potential peak revenue with this drug just in breast cancer. And so I think if people understood that and just fast forwarded what the implications are of having effective drug that treats 1 of the 3 critical disease drivers, that would potentially change their current view.

All right, Brian. Well, thank you so much for your time and insights, and thanks, everyone, for listening.