All right, we're going to go ahead and get started. I'm Stephen Willey, one of the senior biotech analysts here at Stifel. I'm glad to have with us for this next discussion the CEO of Celcuity, Brian Sullivan. Brian's had, I guess, a relatively quiet year, so we're going to have some things to talk about. Obviously a really transformational last three to four months.
Are you going to be standing up the whole time?
No, I'm not. If you just want to start off with some intro comments, and then we can just jump into Q&A.
If you're not familiar with the company, essentially I started the company to develop a platform that could quantify signaling activity in live tumor cells. That involved assessing different pathways important in breast cancer. Initially, one of the pathways we eventually began doing research on was the PI3K/AKT/mTOR pathway. Part of our evaluation of pathways involves evaluation of a variety of drugs in the class. The conundrum that I was trying to answer was, why is a pathway that's so important as this pathway so under-drugged? What is it about the structure of the pathway or the drugs? What might account for that? We did an evaluation of a lot of different drugs, most of the drugs in the class that had ever gotten to the clinic. One drug popped out from that analysis, gedatolisib.
I thought, wow, gedatolisib was owned by Pfizer. That's a company we had a relationship with. We had different partnerships in place with different companies at that time, including Roche and Novartis. That is when we found out that they had made the decision to out-license the drug. I thought, wow. We were essentially clinical development adjacent, sponsor adjacent, and felt that if there was ever a drug, or rather if there was ever a pathway to focus on, it would be this pathway. If there was ever a drug to address this pathway, it would be this drug. It worked out for us. Our focus has been developing the drug initially in breast cancer. We have two studies ongoing in breast cancer, a second-line study. We will talk about this, I am sure, as well as a first-line study in meta static disease.
We also started a study last year in prostate cancer, similar biology, or rather role the pathway plays in conjunction with the hormone pathway, in this case, the androgen receptor pathway. There is a strong hypothesis, a lot of other data that suggests this pathway is relevant. We've reported preliminary data for early phase data. That's very encouraging. We're focused on very important tumor types, prostate and breast cancer. We don't think it's a coincidence that both are hormonally driven diseases. There is a lot of support for the idea that these two pathways cooperate and that combining inhibition of those pathways can yield a very meaningful benefit for these patients.
OK. Yeah, so a pretty transformational stretch here, right? Top-line data in July. You're now trying to submit a regulatory filing under an accelerated review to enable commercialization. Just how has your day-to-day kind of changed over this stretch? What are you now having to spend most of your time on?
It has always been busy. I think a lot of what we were doing was laying the groundwork. For instance, we were doing work and planning around an NDA. We had a date at the beginning of the year that we focused on based on time after top-line data. We laid a lot of groundwork internally to get those modules done. We are laying a lot of groundwork for the commercial program, commercialization, in terms of building the organization, the processes, the systems internally, getting data research with strategic accounts and payers and docs, et cetera. Just internally preparing to scale. That is a lot of prep work. Once the top-line data becomes available, you have dates, you have deadlines, you are focused on executing. We teed up a significant number of people who essentially accepted offers contingent on favorable data.
We kind of laid a lot of groundwork. Once the rocket lifted off, even busier. Also, because we think the data validates the importance of this pathway independent of the presence of a mutation, it allows us to also maybe think a little differently about front-line settings, about how we might design the trials, et cetera. We are basically leveraging the data we got out of this study to inform how we think about future studies and also how we think about the life cycle development overall. Essentially, this was a question. I think there was a fair amount of skepticism that our mechanism would yield a result in patients who lacked mutation just because there has not been much success in that area.
To have results as favorable as they were certainly then allows you to have, I think, a high degree of confidence, at least we do, that as you prosecute development in earlier lines of therapy, you can have a high expectation of success. You want to leverage the data in a way, as well as obviously the work that just comes with specific target dates, both for NDAs and approvals and then launch.
OK. Some time has passed since the top-line, the presentation at ESMO last month. What are you hearing from your investigators and KOLs just with respect to the role of gedatolisib in this wild-type patient population?
To be frank, it's been great. I mean, I've said all along, I think if our data is what I think it will be based on our interpretation of the earlier phase data, it's reasonable to expect we could get a majority market share in the second-line setting. Our data validated our suppositions, and that was great. As we've gone out to do market research, you do demand studies. You try to isolate the thought process of these docs to really get a clean read on how they use the drug. You do that in different ways. You also segregate KOLs, people who are working at an academic center versus those who are working in the community, just so you can tease out any differences in attitudes or preferences.
The data has been—I mean, to be frank, the demand study has gone better than I think we expected. We tend to be more conservative. We are very, very encouraged by kind of what we are hearing in the field and in the community. With KOLs, I think we helped set a lot of apple carts because there have been assumptions that are in research. They have done a lot of work with drugs in this class. It has not worked out. I think there was just generally, huh, well, these guys are doing something that has not been successful before. Good luck. We were successful, and the data was very, very good. I think they almost know too much in some ways. Their involvement has been—I mean, we expect significant cooperation and engagement with them. We are already having that now.
OK. Were you surprised by the amount of attention that the Roche data was getting going into ESMO? I guess I asked the question because when I kind of look at that combo, right, it seems to me that the fact that Roche is combining kind of a "best in class," next-gen SERD with a drug like everolimus is almost kind of indicative of the value proposition of the GETA itself.
Right. I was surprised for two reasons. One, I think there was this belief, and I think that's what it was. It was not really based on any data, a belief that maybe this time a drug will show activity, or rather an oral SERD will show differential activity in patients who lack mutations. I thought, again, when you are O for four, I mean, and this drug, giredestrant, showed no activity in patients relative to fulvestrant in the ESR1 wild-type setting. Unless there is some magic from the combination, some synergy that has not really ever been revealed, that is not going to be successful in the wild-type. At least that is what I concluded. I think there were a lot of people that thought, oh, no, maybe this will be different. That is fine. I think everolimus is a drug that has been out there for a long time.
Doctors are familiar with it. The data for that drug has been fairly clear when you look at the tolerability questions, very high, to be frank, levels of treatment discontinuation. In their registrational study, I think there's a 24% discontinuation rate. It was about 17%, I think, in the Avera study. That is a drug that can be rough on patients. I think that's how investigators would think of that drug, dosed daily. GETA, it's a feature that, again, I think people are starting to recognize is important. Targeted therapies that are dosed daily, particularly ones hitting important pathways, can be really rough on patients just because you're introducing Cmax levels of drug. Cmax concentrations are typically what would drive a lot of the adverse events. GETA is only hitting Cmax or introducing Cmax concentrations three times a month.
By the second day, you're at concentrations that are probably below your IC50 level to disrupt healthy cells. I mean, the tumor cells are more sensitive, at least in the PAM pathway, much more sensitive to inhibition of the PAM pathway than a healthy cell, 10x difference. You can quantify that. We think overall, you're just introducing lower concentrations of drug that are still very effective because that's one of the key features of the drug, is that by comprehensively shutting down this pathway, you don't need as much of it to induce a treatment effect. We've found in the non-clinical work that gedatolisib is 300-fold more potent than these other drugs in reaching a half-maximal effect on tumor cell proliferation. The toxicity associated with that drug is very, very important.
I think we've seen in this setting, breast cancer setting, that there's a threshold of toxicity that docs are really unhappy about and the patients. When Capivasertib was approved, Truqap, which they did not report better data than alpelisib, but they reported much better tolerability. There was, I think, I would estimate probably 80%+ shift away from alpelisib, if it's not more than that. Everolimus has that aspect to it that you can't get away from. Doctors know it. They've used it. We think we have a much better tolerated drug just based on the adverse event profile. That'll factor into the good. I just thought it was clever on Roche's part. I thought, OK, that's a smart thing to do, is get out there with a generic drug that's hitting the pathway, see what you see.
It is a good strategy. There have been three or four companies already advanced in the monotherapy that how is somebody going to differentiate there? Unlikely. You are fighting over table scraps. It was a smart development strategy. I give them credit for that.
I think some investors still have an outstanding question as to what the GETA triplet looks like in that 30% of ESR1 mutant patients. Is that a subgroup analysis that we should expect to see at some point?
One thing that's important to look at is actually you've got kind of four subgroups in this setting: ESR1 mutants, PIK3CA wild-type, ESR1 mutant, PIK3CA mutant, and then vice versa. The proportion of patients that are ESR1 mutant, PIK3CA wild-type, is about 20% overall. There's been a number of studies that confirm that. That's about the only overlap we have with gedatolisib in our indication. We think, at least if we report data consistent with what we saw in the phase 1b study, roughly 15 months in the mutant setting, that will not be considered an option for patients who are ESR1 wild-type versus and PIK3CA mutant. We also provide some regional analysis for gedatolisib.
US and Asian patients in particular showed meaningful differences, pronounced differences, I guess might be the best way to say it, between the intended treatment population, 19-plus months in the US, 16-plus months in Asia. There seems to be an association with kind of the general health care practice of the country and potential duration of benefit you may get from gedatolisib. When all those factors are taken into account, I don't think mutational status will be relevant when doctors are prescribing or making a decision about the second-line therapy, nor will they have to make decisions based on HbA1c levels. I think the low discontinuation rate, the low levels of hypoglycemia, the prior data that suggested we don't affect HbA1c levels from baseline over time of treatment, similar with glucose levels. It's a very safe drug to prescribe patients that may have some glucose-related issues.
Do you think that extended duration of response or PFS that was seen in the US, I know it was kind of a smallish patient number in terms of denominator, but do you think that that's a real signal to your prior statement? What kind of implication do you think that has just from a modeling perspective, right, as we try to figure out what this drug could be in the US?
You should assume in your model, 19 months in the US when you publish your research. How about that? I think it's relevant. I think then when you start combining some of these other regions, let's say you look at Asia, it was also very different relative to the intended treat. We didn't break out Western Europe from Eastern Europe when we did the initial regional analysis. That certainly would be an analysis worth doing. It'd certainly be worth looking at, let's say, the bucket of what you might call more advanced economies, Western Europe, Asia, US, Canada, versus the countries that have less developed economies, and see what you see. If you do that analysis, you'd have a fairly significant proportion of patients, more than probably about 2/3, who are treated in countries that have more advanced economies and health care systems.
Then sample size becomes less of a concern about how to interpret the data.
OK. What drove the timeline extension for the Victoria 1 PI3K mutant population?
The enrollment for that study was almost exactly on track. We finished it within a week of what we had projected. Once your enrollment is done, or rather once your enrollment has come in per your assumptions, the only variable really you're left with is the event rate. We track event rate. You have a model, but it's based on assumptions. You don't really have any way of validating your assumptions. There's a lot of it's hard to place too much stock in that. It gives you some boundaries to think about, but it's hard to use it as a pure projection because you don't know the key question, which is what's the PFS, and when will these events occur? You look at the actual event rate, number of events, and you look at that on a rolling basis.
You also look at it in time buckets, let's say Q1, Q2. We saw fairly consistent event rates. That was the basis. You can do the math. Say we're at x, we need to get to y, divide by the event rate. That's how many months we have to go. That event rate slowed, and the slope of the curve shifted. Based on that, we said, if that event rate is now the new normal, then we'll have to adjust our estimates. The question is always, is it due to the control? Is it due to the study drug? At least in the case of wild-type, it appeared that the slowing down was a function of not the control. The control was pretty much what we thought. It was a function of the study arm.
Again, we do not have any data to allow us to speculate about that. Those are kind of the two drivers, and you will see it when you see it.
Should we expect the nature of that PIK3CA mutant disclosure to look kind of very similar to what we saw in the wild-type?
It's somewhat situational. You can't really predict that because if your data is available just adjacent to the timing of a major medical conference, you'd probably hold off and allow the big reveal to occur at that conference. In the case of wild-type, we had a 3-and-1/2-month gap from when we had the data available and the next conference, which was ESMO. Given the importance of this drug to the company, we just did not feel that that was appropriate to keep under wraps. It just wasn't practical. Now, we do have data out there. There's more credibility just for gedatolisib. We have probably a little bit more flexibility. We'll have to factor in the timing and make a decision accordingly.
OK. Can you speak to, I guess, kind of what, if anything, is rate limiting right now to your ability to file an NDA before the end of this year?
Nothing.
OK.
We're on track. I mean, we got accepted into the RTOR program. That requires you to have two pre-submissions. Those are typically modules that you can have ready to go before you have data, your CMC package, et cetera. We’ve made those submissions. We were ready to roll prior to the data being available. Now you're just left with a relatively small portion, although the most important portion, of your NDA, the clinical study reports, your summaries of efficacy and safety, et cetera. You have a lot of work you can do ahead of time to get ready. It’s a fairly predictable process.
OK. Per the RTOR pathway, I know that we've seen some prior art where drugs have gotten approved pretty quickly. What kind of position will you be in from a commercialization and readiness perspective should a similar rapid approval timeline apply here as well?
There's kind of two dates you use when you're thinking about projecting an approval. One is the PDUFA date. If we're accepted for priority review, that would be six months from date of submission. That would kind of take you to, or date of acceptance. That takes you to kind of the beginning of the third quarter. RTOR kind of works on a somewhat independent timeline. I mean, they're obviously compliant with PDUFA. The history has been, and there have been, I think, 19 or 20 RTOR programs evaluated, that the review is about a four to five-month process. Internally, we've kind of bucketed those two time points and said, OK, let's be ready for the early time point just in case. Can't count on it. There's not everything you can do.
I mean, you have to be a little careful with bringing on your field sales force too prematurely. And so we have kind of a T-minus 0 plan and a T-minus 60 plan where if we get surprised to the good, we'd be ready to roll.
OK. What does your market research tell you about the kinds of practices that you're going to be kind of targeting right out of the gate? Do you skew academics, skew community, hit it all?
One of the virtues of America is you can get data from pretty much anything. You can identify the doctors who see according to how many patients they see. You can get a fairly good estimate. You'll see that the major prescribers or major treaters are in the community setting. I think the data is pretty consistent, however you cut it, that roughly 80% of these patients are seen in community settings, and they're treated there. Clearly, you would target doctors. You'd make a priority of being able to get in front of the ones who see the most patients. It's just logical. You also prioritize the academic centers because of their importance, their influence, et cetera. They also see a lot of patients.
You develop a launch plan that really is very, very specific according to kind of where doctors fall in terms of how many patients they see. Because obviously, you want to have an impact. If a doctor is seeing 100 patients a year, they're going to have 100 patients they can benefit if they start prescribing gedatolisib. You build your plans around that. You build your sales force around that, your call patterns, et cetera. You can get very specific as you're preparing your launch and who you want to see and how you go about it.
How are you thinking about xUS and prioritizing those regulatory submissions? Is it strategically important for you to maintain global ownership of this asset?
I don't think we need to make that decision now, and so we're not. We will advance the drug through the regulatory process, though, in Europe and Japan now, even prior to having a partner. We expect to submit the MAA for the European market soon after we submit the supplemental NDA, or assuming we submit the NS NDA for the mutant population. That would be a combined submission. It just makes a lot more sense in Europe to have a combined submission. That is a 12-13 month review cycle if you follow the typical timelines. We have some time, basically, to sort through how we want to approach xUS. We've already had a couple of regulatory interactions with the Japanese health authorities. We think we've defined with them what their expectations are for a data package.
We know what needs to be done. We're already moving forward with some of that work. We think we're laying the groundwork to have regulatory approval in the major countries, right? There's the five EU countries that comprise the largest component of drug spend, and then Japan. Those six countries, plus the US, will account for 90+% of most drug companies' revenues. In our case, we're going to make sure we don't have any delay in being able to reach those markets. How we get there, we've got time to figure that out. We think it's to our advantage to delay that as appropriate as long as possible because more data is available, just gives us more leverage in conversations, gives us more data about how to think about positioning the drug, et cetera.
You have another phase three trial that's also up and running, Victoria 2. Can you just remind us the types of patients that are eligible for this trial and how your version of this trial is different from Roche's version and ABO-120?
Sure. The Roche study, we'll just start with that. The Roche study evaluated women and our study, so there are commonalities. There are two groups of women who present with metastatic disease. One group is considered endocrine resistant. These are women who got nominal benefit from their adjuvant tamoxifen, their adjuvant endocrine therapy. They progressed either while they were on that adjuvant treatment or within 12 months. They are considered endocrine resistant. That's about a third of the total population. The NAVO study studied those patients and generated really the first data for a CDK4/6 fulvestrant combo that was isolated to that population. That data was very informative for us. Those patients only reported about seven months median PFS. In some ways, you can think of it as more of a second-line study, even though these are treatment-naive patients.
That helped inform how we would design a study because you had a good bead on what the control assumption should be. They had to limit the study because of the nature of their drug. In patients who have PIK3CA mutations, that's about 40% of the population. They only studied patients who, or rather, they excluded patients who were pre-diabetic or diabetic, which is about 50% of women with breast cancer. Their effective population that they can treat is only about 20% of the total of that population. Our study will be an all-comer study, independent of mutational status, independent of HbA1c status. We'll exclude patients with uncontrolled diabetes. That's a small fraction of the total. One of the features of the drug is that we don't induce high levels of hyperglycemia.
We think it's a safe drug to prescribe to patients pretty much up to HbA1c levels of 8, which is kind of where the cutoff is for defining someone as having uncontrolled diabetes. As a result, we think we'll be able to position gedatolisib in this setting as a drug. Again, we have to get the results. We have to enroll the study. If the data bears out in a way that's consistent with what we saw in the second line, we think gedatolisib would be an option that these doctors could use. Independent of, again, their PIK3CA status or HbA1c level status. That, we think, is a very important feature, particularly in the community setting.
If you can reduce the complexity associated with treating their patients, and you can offer the best-in-class therapy that's appropriate to prescribe to the bulk of their patients, we think that's going to be very effective positioning for a drug like gedatolisib.
Is there much overlap between the trial sites that you're looking to activate in Victoria 2 relative to the ones you use in Victoria 1?
There is. I mean, we know who the bad sites are, so you exclude those. I mean, whenever you have a study, there's always about 20% that don't do nothing. That is very consistent. You sort through some of that, so you have a higher hit rate of effective sites. Yes, and we're leveraging that quite a bit. The good news is that every site that we wanted to participate who were activated in the Victoria 1 study wanted to participate in Victoria 2. They had very good experience with the drug and the patients. They did not know what the data was, but they saw how well their patients were doing. They saw how well they could tolerate it. We took that as a bi-signal before we had our data. We said the doctors really like this drug. They like prescribing it.
They're watching patients respond to it because it's an open-label study. They know which patients are getting it. That kind of made us feel positive. I mean, you can't take it to the bank. Certainly, their experience is important in terms of their perspective. That’s helpful. You fill in around the blanks. The regions will probably be similar. We'll probably exclude Latin America in that study because there just seems to be something going on in Latin America that it's hard for us to tease out. Other studies show poor performance of their drugs or poorer results in Latin America. That's what we saw in this study. It doesn't make sense for us to go back.
OK. When do you think you might be able to move into the randomized portion of Victoria 2? Will that also coincide with a data update from the safety run-in?
We have to complete the safety run-in. We just enrolled our first patient in late July. We have really just begun that study. You have got a couple of factors. A, you are starting to enroll patients. You are also activating sites. That is a bit of a ramp. We would expect we have not given dates yet. Certainly, in 2026, we would be well on our way to building up enrollment in a randomized setting.
OK. Maybe just lastly, if you can just talk about the current cash runway and what that allows you to execute on.
The short answer is we think with the cash we have on hand and then the access to cash we'll have, the incremental cash we'll have from the term loan facility without even accessing much of it would take us well through 2027. I mean, through 2027. We think at that point we would hope to be generating meaningful revenue. Your burn rate is significantly where you should be in a position where you're getting significant contribution from your revenue to offset just the pure burn based on expenses. We'll constantly assess our cash position, our balance sheet, what makes sense to do, and mindful of that. We think we've got an appropriate amount of cash given where we are. Always when we're reviewing that.
OK. If there's no other questions, Brian, I appreciate the time.
Oh, you're welcome. That was my pleasure. Great.