Okay. All right. Great. Thanks, everyone, for continuing to join us here at the Stifel Healthcare Conference. My name is Brad Canino, Senior Biotech Analyst here. Happy to be joined on stage for the next fireside with Celcuity. We've got Brian Sullivan, CEO. Thanks, Brian, so much for joining us.
Welcome.
Can we just start off with an introduction to Celcuity and the lead drug, gedatolisib, and what investors can expect as the news flow through 2025?
Sure, so I started the company initially with the goal of developing a platform that could characterize, quantify signaling pathway activity in live patient tumor cells. We essentially evolved focus on the PI3K/AKT/mTOR pathway and quantifying activity there, and that led us to a drug called gedatolisib. The reason we were interested in this pathway is because it's one of the most important oncogenic drivers in cancer, but it's also one of the most difficult to target because of its complexity: six targets in one way to think about it, and the end result was that even though it's involved in a variety of different tumor types, it's very underdeveloped. Very few patients, relative to the proportion of patients who could benefit, are receiving treatment with a PAM pathway, so that led us to focus on this area.
We came across this drug in our analysis that clearly indicated it was superior to any other drug that we had evaluated in this space and happened to approach Pfizer at a time when they were reconsidering their development approach, and so we were able to acquire this drug, and so our focus now is on three programs: two in breast cancer, one in prostate. We have a study that will read out an initial cohort of patients in second-line breast cancer. These are women who have HR+, HER2- breast cancer that has progressed on prior CDK and letrozole. We've essentially run two studies in one: patients who are PIK3CA wild type, they don't have an activating mutation. That data we expect to read out late Q1 or sometime in Q2.
And then a separately powered study, independently analyzed cohort of patients who have PIK3CA mutations that will read out second half of 2025. And then we expect to get going, and we're currently working to activate a frontline study in HR+, HER2- breast cancer. These are women who have endocrine-resistant disease, essentially progressed quickly after or while they were on their adjuvant endocrine therapy. We expect to enroll our first patient group of patients in this study second quarter 2025. And then we began a study earlier this year in prostate cancer. We're evaluating geda, a combination with darolutamide, men who have metastatic castration-resistant prostate cancer that's progressed on their prior androgen receptor drug.
Okay. So you walked through a couple of the different design elements of the phase three in second line. But maybe one thing that struck me is Celcuity is one of the first companies that's conducting a trial with clear biomarker positive and negative cohorts. What led you to make that decision?
Sure. Well, A, common sense, but B, because of the nature of this pathway, the results to date of drugs that have been evaluated that target this pathway have really only shown activity in patients that have a PIK3CA mutation. And that in turn informed the agency, FDA, which essentially told us, and we were fully expecting this as we were meeting with them to get their feedback on our protocol, that if you want to get approval to treat patients who lack the mutation, you need to independently evaluate that patient cohort. And because there's a different standard of care in wild type versus mutant patients, in the wild type, it's really fulvestrant is the best control for that study. And then alpelisib and fulvestrant is really the best control in the mutant population.
It just really made sense to, and for a variety of scientific reasons and regulatory reasons, to break that out. I think we've seen subsequently with drugs that are targeting estrogen receptor in patients with ESR1 mutations or PI3K alpha drugs or AKT drugs that essentially have a similar effect, that the FDA is not going to accept an intent-to-treat analysis. That essentially you may have a favorable readout in your mutant population, but if you don't separately demonstrate activity in the patients lacking that mutation, they're not likely to approve you. So folding the mutated population inside an intent-to-treat analysis isn't going to work, at least in this setting. At least that's our view.
Okay, and now when you enrolled patients, you split them based on the biomarker status into the two cohorts. What happened to the proportion of patients enrolled into each versus your original expectations, and has that impacted the powering of the readouts at all?
Sure. So when we initially designed the study, you have to make some assumptions about what proportion of patients who lack mutations versus those who do have mutations will. If you were to look at various studies, you'd see that the population ranges of patients who have mutations between 30%-40%. And so we used 35%. We split the difference as our assumption. So 65/35 split. As it turns out, as we got closer to full enrollment, which we hit in the wild type cohort this quarter, it trended to be 60% wild type, 40% mutant, which essentially means that we had faster enrollment of the mutant cohorts, lower enrollment of the wild type. It has nothing to do with powering. The powering is determined by a number of events and the enrollment target. So essentially between those two, that will define your powering assumptions.
Okay. Maybe we just focus on the PIK3CA wild type, which we'll read out first. What have you commented on about the powering assumptions for what will be a fulvestrant monotherapy control arm? I think benchmark results in the past have ranged from two months to five months, and we're going to see some additional trials coming up soon to validate which one that is.
Sure. So I'll just kind of caveat your five-month number. There have been four studies done on patient populations comparable to the one we're enrolling, i.e., patients who are post-CDK who have visceral disease. And three of those studies essentially reported two months median PFS. A fourth study for capivasertib reported roughly 3.5 months. If you do an average, weighted or unweighted, it's about three months. And so we think the most likely outcome of fulvestrant, what you'd expect fulvestrant would be three months plus or minus half a month. I mean, that's the best data. There was another study that showed higher activity, but it was not a representative population, had a bunch of patients with non-visceral disease. They have a better prognosis, but they're not visceral. They don't have solid tissue that allows you to do a RECIST evaluation.
And I think most folks and oncologists understand that that data doesn't really. It's not useful as a benchmark. So we think it's a strong basis for us to have a good assumption. We were conservative in what assumption we used to do our statistical analysis plan. And so we're, to be frank, well overpowered because the results we've shown in our preliminary earlier phase, if that data is replicated, then that result will be. We'll have been well overpowered.
Yeah. Maybe just take a second to recap what you saw in the phase 1 data that suggests the potential efficacy for gedatolisib.
Sure, so in two main cohorts that were relevant, one is a group of patients who were treatment naive, and the second group, which is related to what Brad is talking about in the second line setting, patients who got our phase three dosing schedule. For that study, the objective response rate overall was 63%. Median PFS was 12.9 months. If you break out the data in the different wild type versus mutant cohorts, you'd see the numbers were comparable, and that's a very unusual result. Typically, you would see high differentiation between activity of drugs that, let's say, hit alpha only between the wild type and mutant cohort, and so that was very encouraging to us. 49% of the wild type patients were progression-free at 12 months. 60% were progression-free at 12 months who had mutations.
If you compare that to results published for our comparators, our data is very favorable. So there's a strong hypothesis for us to proceed. And that's consistent with the nonclinical data we reported. It's hard to look at nonclinical data without the context of clinical data. But what's been interesting to date is that if you look at the nonclinical data for these node-specific inhibitors, alpelisib, which hits alpha, AKT, capivasertib, mTORC1, everolimus, the nonclinical data actually is very predictive of lack of activity in the wild type cohorts, highly differentiated sensitivity or potency in mutant versus wild type. Whereas geda, in the same analysis, was a 300-fold more potent, but equally potent in both wild type and mutant data.
And so while the nonclinical data is kind of, I don't want to say predictive, but consistent with the data for those other drugs, our preliminary data at least is consistent with the nonclinical data. And obviously, we're waiting to see if phase three data is too.
Okay. So we talked about this study with the control arm generally being around three months for the wild-type patients. We discussed the phase 1b data that informed the efficacy of gedatolisib. Now, obviously, the pivotal trial has an event-driven endpoint, and that is a time-to-event endpoint that you have pushed out now about six months or so, correct me if I'm wrong, from your original assumptions. I guess, what should we take away the potential implications from that?
Sure. Well, I mean, an event-driven study, you don't control when the events occur. You develop a projection based on the assumptions you used in your statistical analysis plan. You model that. Our case is a little bit more complex because we have three arms in each of these cohorts. And so we can only look at, because we're blinded, we can only look at the data in aggregate for the number of events. And the event threshold will get triggered, the threshold that will trigger analysis is the B versus C event total and the A versus C event total. We're blind to that. And what we can do is, through our projection, see, well, what aggregate number of events correlates to the event thresholds for those two primary analyses. And so if the events are taking longer, they're taking longer.
Earlier in the year, we stretched out the timeline by about a quarter. And that's because of this shift in enrollment between the wild type versus mutant. And so even though our overall enrollment was 100% on track, it just was falling in slightly different buckets. And so obviously, if you have a lag in enrollment, that's just going to lag your event rate as well. But now that we're 100% enrolled, we have certainty on our patient population and who's there. Basically, all we can do is tally the number of events at the end of a month and see what that trend looks like and how that compares to what we were expecting and say, huh, if that trend is a little different, that suggests that the timeline of hitting the threshold will stretch out.
Obviously, the question is always, well, is that because the control arm is outperforming your expectations? We really can't comment on that other than to say we think there was a lot of data that supported our assumptions for the control. We've talked about that data. We were more conservative in our assumptions than the data that's been reported. So we'll see what we see.
Okay. Now, in this setting of second-line patients who have progressed on the CDK plus an aromatase inhibitor, what do you hear from KOLs in terms of the magnitude of PFS benefit that they demand to declare something a new standard of care?
Sure. So there's always that. Say we think will be relative. I don't want to say easy bar to beat, but it's a low threshold because the control assumption is low. So from an absolute standpoint, you can be statistically significant, but maybe not clinically relevant. We hear very consistently from regulators, KOLs, community docs, that if you're able to show a three-month benefit, three-month improvement in PFS, that is what will motivate them to use a new regimen. Now, obviously, we would hope to be better than that. We'll see. But that's in effect the endpoint to be able to, we think, get approval to be able to have an effective launch. And that's also consistent with what we've heard for the mutant subgroup as well.
Okay. And now, as a triplet, how do you think physicians will respond to the palbociclib component as the CDK4/6 partner that you've chosen when the class leaders really switch to ribociclib in the front line?
Sure, well, my short answer is I don't think it'll matter because ultimately our data will stand on its own, and the ribociclib data is for front line indication. They don't have data as a later line drug. We'll have data for our drug in that combination, but secondly, and from just a practical standpoint, doctors are used to switching therapies that hit the same target, so that's the case in breast cancer, right? They start off in most cases with an aromatase inhibitor. They switch to a SERD, again, having to disrupt estrogen signaling activity, and so in prostate cancer, you see doctors switch from, let's say, abiraterone to enzalutamide, similar target, slightly different mechanism.
There's this common understanding that, well, if you need to maintain pressure on a pathway, better to switch out and see if ultimately maybe slightly different characteristics of the drug might induce some benefit. But net-net, if we report favorable data, I don't think it'll be questioned. And the final, I guess, comment is I think most oncologists would agree that or believe that palbo is a better tolerated CDK4/6 than abema or ribociclib. And so in the second line setting, there's an advantage to the patient. It's a less rough drug, requires less monitoring. And so I think net-net it'll be viewed as an advantage.
Now, since it is a triplet as well, can you talk about how you expect to discuss contribution of components with the FDA?
Sure. So we're evaluating geda, palbo, fulvestrant, then geda , fulvestrant, then fulvestrant. And so the initial one bucket of contribution will be the comparison of geda , fulvestrant versus fulvestrant. That'll obviously be 100% assigned to geda . And then the difference between the triplet versus the doublet will essentially be split between geda and palbo because the contribution of effect really is only available by combining those two drugs in that setting, not independently available.
Yeah. I think I get questions a lot on how the discussion around, there's no palbo plus fulvestrant arm within the study. Is that going to be a part of the discussion as well? Are there other external data to help round that out?
Sure. So we've had the discussion. We've had a couple of Type C meetings with the agency to review a design. We got breakthrough designation for the indication a couple of years ago. So we've had a bunch of Type B meetings since then, partially to lay the groundwork for the NDA and making sure that the data that we have will address their requirements for an NDA, but also just to discuss that type of question. And so there is data that's available, PACE study, evaluated retreatment of patients who progressed on prior CDK with palbo, found nominal to no effect. And so I think there's a general understanding, and other companies have sponsored studies, that there's a nominal effect, less than a month potential benefit.
So the agency understands that from a practical standpoint, it would kind of penalize us to try to field that when there's limited prospect of effect and that the data is fairly consistent. And so they believe that, and the nonclinical data supports the rationale as well. Interesting nonclinical studies have shown that tumor cells that have become refractory to CDK4/6 treatment, when a PAM inhibitor is introduced, it reactivates that pathway and resensitizes those cells to treatment with CDK4/6 and you essentially get a benefit from that. And so that's the hypothesis for the addition of palbo in that setting.
Okay. Now, if you receive positive top line results in the first half of next year, what could an NDA process look like? And given the unmet need for these patients, is there a potential accelerated timeline available?
So we've already had those types of conversations with the agency. They agree that if we have favorable data in the wild type setting, that we could submit an NDA for that cohort of patients. Study is independent. Statistical analysis plan is independent. So it makes sense. We also think because we have breakthrough therapy designation, that's reasonable to think we would likely qualify for either a Real-Time Oncology Review or Priority Review. It's up to the agency. But you see a high correlation between a breakthrough therapy designation and an accelerated review. And it ultimately comes back to the data. But if the data is favorable, we're preparing ourselves for that. But you have to request it and they have to agree to it.
Okay. And when do you think is the right time to bring in ramp of sales force?
Sure. So you go in stages. I think for a biotech, big pharma can spend three years getting ready. I think biotechs in general think about 18 months. So you start laying that initial groundwork 18 months out from your expected launch. And you do some of the preliminary work. The activity for the field force is essentially the last thing that comes in place. And so you're developing, you're building out your business operations capability, your market access capability, your marketing and positioning. You bring on your senior sales leadership team, whether national sales manager, then your regional managers. And then a quarter out from launch, plus or minus, not minus, but essentially get those folks in place so you can train them about a quarter before.
Okay, and what is your calculation of the addressable market and commercial opportunity in the second line setting post-CDK4/6?
Sure. So we estimate that there are roughly 30,000 women or patients annually who receive treatment in the second-line setting post-CDK. And so if you assume a price for geda that's comparable to prices currently being received for approved PAM inhibitors like capivasertib or alpelisib, and then if you assume median duration of treatment of 10-12 months, the addressable market's roughly $5 billion. So a very significant market. Now, then you have to estimate what do you think your penetration would be. We don't think it's unreasonable to think we could get 40% penetrated if we have data that's comparable to what we reported in our early data. And if that's the case, we have a potential to fill an indication that has a peak revenue potential of $2 billion. At least that's internally what we would shoot for, all things being equal.
Okay. Now talk about the work you're also doing to advance gedatolisib to front line. And what drugs for combinations you're focused on in that setting versus second line?
Sure. So in the frontline study, VIKTORIA- 2, first study is VIKTORIA- 1. In the frontline setting, we're focused on patients who have endocrine resistant disease. These are patients who, while they're receiving adjuvant endocrine therapy for their early breast cancer, progress either while they're on those drugs or within 12 months, and the prognosis for these patients is really poor. They only get a fraction of the benefit from a CDK4/6 and fulvestrant, which is the standard of care today, seven months versus 25 months, and so the standard of care is CDK4/6 is the same ones that are approved for endocrine sensitive, so it's ribo, abema, palbo. In our study, we're giving physicians a choice. We'll be evaluating geda , physicians' choice of ribo or palbo, and then fulvestrant.
Okay. And you've obviously undertaken this strategy, at least the initial stage of it, before you've gotten the second line readout. So how did you contemplate the aspect of moving fast to front line before you had the second line results?
Sure. So there are two considerations, or actually three. One was data was reported really the first time for this patient population that allowed us to develop a plausible protocol. Up until this data reported out, this was the INAVO120 study. It wasn't clear what the outcomes endocrine resistant patients would get with CDK4/6 as fulvestrant. That would be your control. Well, this data reported out median PFS for that regimen of about seven months, which suggests that you can have a reasonably sized, reasonably long study, one that would be appropriate for a biotech at our stage to initiate. Second consideration was the strategic. A lot of different drugs under development. If you can position yourself to get to the finish line first, that's always an advantage to being a first mover.
Now, there is inavolisib that's approved, but because they're really only addressing a small fraction of the patients, we think they'll lay the groundwork for us. But if we're able to offer a broad indication independent of PIK3CA status, independent of HbA1c status, diabetic status, we think that'll ultimately be the winner. And so better to be first than second or third. It just makes life a lot more complicated. So strong strategic rationale. And then a financial rationale. This is a big setting. Potential addressable market for this patient population is roughly $3 billion. And if you do the math and evaluate the net present value that could accrue from this indication, starting the study 12 months earlier adds $1 billion to the net present value of that indication. So all those factors made us say, "Huh, we want to preserve our place in line.
We want to be first. We don't want to leave $1 billion on the table if we can avoid it." It just makes all the sense in the world to proceed now, essentially. And as it turns out, we will expect to get our wild type data about the same time that we expect to enroll our first patient. So we're kind of trying to be pretty clever in how we timed that.
Okay. Can you expand a little more on the potential competitive positioning of gedatolisib versus inavolisib in the front line? Roche has made some pretty positive comments about the strength of the label that they've seen. They are expanding broader into an endocrine-sensitive population as well with an additional study that they're running. So I think you talked about gedatolisib will have a larger opportunity, but within the opportunity that it overlaps, how do you think about those two drugs?
Well, so their indication doesn't overlap with our patient population in the second-line setting. These are really two different pools of patients. And so we don't think that that will affect us. As far as where they go in the sensitive population, that's a tricky study to do, certainly for a biotech at our stage. Your control assumption would be 25 months. And so then you would need probably just from a math standpoint, at least 33, 34 months to justify the duration of treatment that long. And so you're talking about a five- to six-year study. That's beyond what we can do. If they do it, they'll have to do it in patients who only have mutations.
And even though their label doesn't have an HbA1c threshold, from a practical standpoint, their early phase data suggests if they don't limit that drug to patients who are or exclude patients who are prediabetic, diabetic, they're likely to induce high levels of hyperglycemia. It's comparable to alpelisib. So I think doctors, oncologists have a lot of experience, not great experience working with alpelisib. They used it because it was inducing a meaningful benefit for their patients, but it was very difficult for many of these patients, very difficult to manage. They, in effect, felt like they had to become endocrinologists. So from that experience, I would expect, but I guess we'll see that oncologists would be cautious about going beyond the patient population that was evaluated in the phase 3. But maybe incrementally they can expand beyond that.
But there's quite extensive monitoring required in those first couple of cycles of treatment, glucose monitoring, which is unusual. That's not the case for any other drug in this space. So we'll see. But the FDA clearly understands that, and that's why that's on the label. And so net net, I think it'll be a drug that hits a fraction of the patients that, if our data bears out, that we would expect to be able to treat.
Okay. Now thinking beyond breast cancer, how do you articulate the thesis for geda , darolutamide and prostate cancer, which was another opportunity you mentioned at the top?
Sure. So similar to breast cancer, prostate cancer is driven by a hormonal pathway, androgen receptor pathway. A lot of nonclinical data has characterized the interaction between the PI3K/AKT mTOR pathway and the androgen receptor pathway. We have our own nonclinical data that's demonstrated that. And then secondly, there's also very compelling clinical data that has shown a benefit when you combine PAM PI3K inhibitor or an AKT inhibitor with an AR inhibitor. And so the combination of the biology of the disease, or at least as it's characterized nonclinically, and then these promising clinical results. And as it turns out, the PAM PI3K inhibitor that was evaluated, its development was discontinued primarily because of drug stability issues. At least that's what we've heard. And so it provided a good breadcrumb for us to follow. And then the AKT inhibitor was a Roche drug.
Too toxic, showed a clinically meaningful and statistically significant effect in patients who had a PTEN loss, which is the primary mutation of this pathway in prostate cancer, and so both those studies essentially to us validated or certainly provided good evidence of the benefit that could accrue if you combine treatment and blockade of both those pathways in prostate cancer.
What was your thinking behind choosing darolutamide specifically as the combo partner for the AR pathway inhibitor when there's four that you can choose from?
Sure. So that really came from the discussions we had in ad board meetings. As we were designing the study, we were reaching out to leading prostate cancer doctors from around the world. One of the questions we asked them was, "Well, which partner, which androgen receptor drug would you combine with geda ?" Almost universally, they said darolutamide. And when we dug into that, they said it was primarily because it's much better tolerated than the other drugs. And then dig into the biochemical and structure of the drug, it's more potent. And so those factors really let us to conclude that it would be the best partner, best for patients, and it seems biochemically the most active of the drugs that are approved in this setting.
Okay. All right. Well, I look forward to all the updates that are coming quite imminently. Thanks so much, Brian, for joining us.
You're welcome.
Thanks everyone for attending.
Thank you.