Thanks so much, everyone, for continuing to join us on the CF Virtual Oncology Days. I'm happy to have for the next fireside Brian Sullivan, the CEO of Celcuity. Brian, thanks so much for joining us. If you could just kick off, please, with an intro to the company.
Sure.
Give us a status update of the gedatolisib phase III trials in second-line HR positive metastatic breast cancer?
Sure. So our company is focused on treating cancers that involve the PI3K AKT mTOR pathway, or PAM pathway. You know, it's the most important oncogenic pathway. It's the most highly altered pathway, also one of the most difficult pathways to drug effectively because of its complexity. And as a result, you know, we think it's one of the largest untapped drug development opportunities in solid tumors. You know, gedatolisib, our lead candidate, is a PAM PI3K mTOR inhibitor, essentially as it equipotently inhibits all six of the Class I nodes and mTORC1 and mTORC2. Highly differentiated potency and cytotoxicity profile compared to drugs that only inhibit single nodes of this pathway. We currently have two programs underway: second-line breast cancer study in patients who progressed on their prior CDK. We are essentially conducting two separate cohort studies.
You know, one in patients who have PIK3CA wild type, who lack PIK3CA mutations. We expect to report that data second half of 2024. And then a separate cohort analyzed separately of patients who have PIK3CA mutations. And that data we expect to be available first half of 2025. We have a second program in prostate cancer. We just began enrolling patients in that study in the first quarter this year. And we're evaluating men who've progressed on a prior next generation androgen receptor inhibitor. And we're treating them in a metastatic castration-resistant setting.
Right. Great. And relative to the other PI3K mTOR inhibitors, can you overview the properties of gedatolisib that have produced a better safety profile in early clinical data so far, even as a triplet therapy and without careful patient selection?
Sure. Right. No, I think, you know, just to give a perspective on the tolerability and put it in context. The mTORC1 inhibitor, everolimus, which, you know, is approved for use in this setting, reported a 24% discontinuation rate due to AEs in its registration study. The PI3K alpha inhibitor, alpelisib, reported a 26% discontinuation rate due to AEs. The AKT inhibitor that was recently approved, capivasertib, reported a 10% discontinuation rate. And those drugs have been, you know, widely used. You know, gedatolisib is part of a triplet, palbociclib fulvestrant. With our phase 3 dose reported a 4% discontinuation rate, which we think highlights the general high level of tolerability of the drug and certainly compares favorably to the reported data for those other drugs. You know, we think this the reason for the differentiated safety profile reflects two important factors.
One is that the drug is administered intravenously, whereas the other drugs are administered orally. And that's relevant because the primary, or one of the primary AEs associated with the pathway is hyperglycemia. The PAM pathway regulates the glycemic system, and so when it's inhibited, it can lead to hyperglycemia. This activity primarily takes place in the liver. So an oral PAM inhibitor is gonna, you know, go through on a first pass the liver, and the result has been pretty consistent, reporting, you know, high levels of hyperglycemia, increase in glucose levels. So gedatolisib avoids that on first pass. Second property that's very important, though, for an IV-administered drug is that it has a favorable volume of distribution. Essentially, you know, if a drug gets overly retained in the liver, it can have the same effect as an orally administered drug.
Gedatolisib, though, has a very balanced volume of distribution, so it doesn't get overly concentrated in the liver. And we think, you know, those two factors, IV-administered balanced volume of distribution, allow Gedatolisib to avoid initiation of, you know, the hyperglycemia that has been a barrier. We've had no patients. We've had 500 patients evaluated with Gedatolisib. No patients have discontinued any of those studies due to hyperglycemia.
Yeah. And now, hyperglycemia, the way it's defined, has changed over time based on the clinical trial criteria. But simply, what happens to serum glucose levels for patients on alpelisib, or gedatolisib, in terms of the increase relative to alpelisib capivasertib?
Sure. So if you look at the, you know, if you look on the labels of these studies, of these drugs, you'll see that the lab assessments, which have an objective measure of glucose, and there's some predefined thresholds that determine whether an adverse reaction, you know, has been induced. And so capivasertib reported overall 58% of patients had a Grade I or greater increase in glucose. 9% of those patients had a Grade III or greater glucose increase. Alpelisib reported 79% of patients had an overall increase in glucose levels, 39% of which had Grade III or greater glucose increases. Whereas gedatolisib reported 23% overall, so a fraction of those other 2 drugs, and 10% Grade III or greater increase in glucose levels. And that's probably the most objective metric, because the CTCAE definition of hyperglycemia doesn't include an objective metric. It essentially involves, it's a more subjective evaluation.
If you're just wanting to assess the effect on glucose, you know, we think the lab assessments are probably the best way to do a head-to-head comparison.
Yeah. And now, with the drug, we did see a lot higher response rate than is normally anticipated from inhibitors of this pathway. What is driving that higher cytotoxicity with the gedatolisib combinations?
So that's really a reflection, we think, of the differentiated mechanism of action. The PAM pathway, very complex pathway, unlike many pathways that, you know, have a target, you hit it, you address downstream signaling activity. In the case of gedatolisib, of the PI3K AKT mTOR pathway, there are six nodes. I referred to those earlier: four Class I isoforms, mTORC1 and mTORC2. These nodes essentially serve as redundancy mechanisms, and they induce compensatory resistance. And so inhibiting one node can essentially activate other nodes and essentially offset the benefit you may get with, let's say, a single node inhibitor. Gedatolisib comprehensively blockades the pathway. And our data suggests that this pathway is intrinsically involved in breast cancer, that it's interdependent with the estrogen receptor pathway and CDK4/6 pathway.
As a result, if you comprehensively blockade this, you can get, you know, very significant, very potent, and very cytotoxic activity. We've published data or presented data at various conferences in breast cell lines, prostate cell lines, gynecological cell lines, I think 60 overall, various PIK3CA status. Two findings, you know, stand out. One, you know, gedatolisib is 300-fold more potent than these other drugs. We think that's a direct reflection of the mechanism of action, the benefit you get from comprehensively blockading that pathway. The other interesting finding is that gedatolisib was the only drug that was cytotoxic in these cell lines.
Interestingly, when you isolate the activity of cell lines that have PIK3CA mutations versus those that don't have PIK3CA mutations, gedatolisib was essentially equally active, equally potent, equally cytotoxic, which, again, provides further confirmation that in tumors that involve PI3K pathway, it's not relevant simply because of a mutation. It's relevant because it's dysregulated overall. There may be amplification of that activity in mutated cell lines, but that the background level of activity that is associated with the pathway is, we think, more important. That's why gedatolisib is effective independently in comparable levels of the mutational status of the pathway.
Yeah. And now, I'll point investors to your corporate slide deck for a table of this. But in your phase I and the second-line plus patients, you did have 2 different cohorts. And you nicely outlined the different activity seen in that. And there was a delta between those. And I think it's been a topic of debate of what might be driving that delta. So can you discuss the differences between those 2 arms and the push back?
Sure. Exactly. So you're referring to the difference in results between Arm C and Arm D of our phase 1B study. There are 2 key differences between the arms that we think provide an explanation for the difference in results. First is that the patient populations were different. The second is a different dosing schedule. The Arm C patients, which had lower median PFS or were more heavily pretreated, had greater levels of visceral metastases and were on their prior therapy for a much shorter time period than the Arm D patients. So essentially, they had much worse prognosis. You wouldn't expect them to do as well, all things being equal, as the Arm D patients. The second important factor was that the Arm C patients were dosed once weekly, while Arm D patients were dosed 3 weeks on and 1 week off.
Based on various analyses that we did, we think it was very clearly important to have an intermittent schedule with this path, with this drug, versus an every week administration schedule. You know, as far as why would that be, we think there are two factors that play a role. One is ribociclib, which is, you know, one of the drugs that was used in combination, CDK4/6 inhibitor. Cross-activates inhibition of CDK4/6 can activate PAM pathway and vice versa if it's inhibited. We think it's important, then, in this setting to synchronize inhibition of these pathways, the blockade, to avoid inducing, in effect, resistance by reactivating the other pathway when it's not inhibited. Palbo is administered on a 21-day, seven-day off schedule. So essentially equivalent to a three-week on, one-week off schedule. And the second, you know, is related to potential impact, favorable impact on the tumor microenvironment.
You know, PI3K, the PAM pathway, when it's dysregulated, can actually serve as an immunosuppressant. And when you inhibit it, you can improve the TME and potentially improve access of antitumor immune cells to the tumor. Hard to tease out which factor is more important. We think likely it's a bit of both. But regardless, you know, analysis of the results very clearly indicates that the dosing schedule is relevant and it's a meaningful part of the difference.
Yeah. And you mentioned the heavier pretreatment, prior pretreatment of the arm that it's slightly worse. Can you talk about how prior chemotherapy impacts PFS from what you've seen from other second-line trials in HR positive breast cancer?
Right. I think probably the best data is from a Novartis study that evaluated alpelisib, which is the PI3K alpha inhibitor, plus fulvestrant in various patient cohorts. You know, these were all patients with prior CDK4/6. So one cohort included second, third-line patients who had no prior chemo, just endocrine prior therapy. The other cohort, roughly 50%, had prior chemotherapy. There's roughly a 30% delta, you know, essentially a 30% reduction in, or 30% better median PFS in patients that did not have prior chemo. And I think that's fairly consistent with what's been reported in general, that patients who've had prior chemo will tend to have less favorable response to targeted therapy than patients who haven't had prior chemo.
Okay. And as we think about the change to the phase III, are there any major differences between the patient characteristics that you would expect in the phase III versus the phase I? And will that help or hurt probability of success?
Sure. No, I think that's always one of the questions, right? Are you changing the baseline characteristics with your eligibility criteria in a way that could potentially add an unpleasant variable? In our case, we actually think the baseline characteristics of the phase III patients will be more favorable. And there are 2 important differences between the patients in phase III that we're enrolling versus the early phase. The phase III study is not enrolling patients who receive prior chemo in the advanced setting. You know, Arm C and Arm D, you know, the patients with prior chemo had between 20%-50% of them had prior chemo in the advanced setting. As we just discussed, you know, chemo can, in fact, essentially depress potential results on the targeted therapy.
So by not including patients who have prior chemo, we think, you know, then that effect, you know, could be to shift your Kaplan-Meier curve to the right a bit. The other important difference is that the phase III study is allowing patients with bone-only disease to enroll. If you look at other study reports for, in this setting, phase III studies, you'll see that bone-only patients comprise about 20% of the overall population. Our early phase study only enrolled patients with visceral metastases. And again, the data is fairly clear. If you dig into study results, you can find that bone-only patients tend to have a better prognosis overall. And, for instance, in the palbociclib letrozole registrational study, found that bone-only patients had a PFS, median PFS, that was 50% higher than patients with visceral metastases.
So, in effect, our early phase studies had a tougher group of patients to treat. Phase III study patients, you know, will, again, we think, have a bias to the good relative to the patients that were studied earlier.
Right. Okay. And now, would you point to any other investigational triple therapies in HR positive breast cancer that target these 3 nodes, even if it's not as potent or optimized to gedatolisib, that provide support for this approach that you're going through?
Sure. So that was a study done a few years ago with everolimus. Everolimus is the approved mTORC1 inhibitor. And they combined everolimus with CDK4/6 inhibitor ribociclib and exemestane, which is an aromatase inhibitor. And these were all patients who had progressed on prior CDK. So, you know, comparable population to what we're evaluating. Data overall was favorable. In one cohort, median PFS was about 8 months, not as favorable as what we reported, but certainly better than what you'd expect with current, what's been reported for current standard of care in that population. So we think it provides essentially further proof of concept, or further evidence that, you know, combined inhibition of these 3 pathways is, you know, highly likely to be beneficial to patients.
Okay. Now in the PIK3CA wild type cohort, which you'll have data the second half of this year, we tend to get two questions on this from investors. I think the first one is fulvestrant control arm and expectations for PFS there and how you've thought about what needs to be successful for the triplet.
Sure. So fulvestrant's been the control for almost every phase III study that's been run recently. And we confirmed the use of that control, you know, during several meetings with the FDA to review our protocol and get aligned on the study design. I think if you look at studies that use fulvestrant as a control, you'd see roughly 2-3 months has been reported, potentially up to 3.5 months. But I think there's a cap, right? There's a limit of what fulvestrant as a single agent will do in this setting. And, you know, we have 2 investigational arms in our wild type study, in both actually cohorts of study cohorts, where we're comparing gedatolisib fulvestrant to fulvestrant. That's the second primary endpoint.
As far as what the expectations are, I think in general, you know, KOLs would like to see, you know, 2-3 months benefit if you have a single drug addition to a therapy. And that's about what capivasertib reported in the mutational. It was about 3.5 months in that setting. So 3 months is probably the floor. Kind of depends what your control does. Somewhat proportional is important. And then with the triplet, I would say if you talk to KOLs, their desire would just see a regimen of, you know, ours potentially would offer, you know, 5 months or so incremental benefit relative to, in this case, control of fulvestrant.
If that was the case, hey, they, you know, the feedback we've received, that would be considered a big win and would position that regimen as a new standard of care, or certainly one that they would go to and motivate them to want to continue treatment with an endocrine backbone therapy.
Right. And then the second topic is the inclusion of the doublet arm of gedatolisib and fulvestrant. Why that was included, and then also, how are the FDA going to look at these data, and does it have to be successful in its own right?
Right. So we thought it was important to include the doublet arm for a couple of reasons. One, it helps isolate the treatment effect of gedatolisib. Two, it potentially could, because it's a primary endpoint, support registration of that as an indication itself. So we could potentially get from this study in the wild type population two different indications. As far as, you know, what the expectations are, I think, you know, the agency would like to see a favorable trend in Arm B, which is gedatolisib fulvestrant versus fulvestrant. You know, the bar to beat, you know, from a statistical standpoint is relatively low if you do the math, you know, and make some assumptions about what to expect with fulvestrant relative to gedatolisib fulvestrant. I think, you know, you can achieve statistical significance with, you know, a couple of months effect size difference.
The bar is not very high to demonstrate statistical significance with that combination.
Got it. And then if the wild type cohort does work, we look at it as a help to de-risk the mutant cohort, which we'll read out next year. And why? Because the control arm will become different and a bit difficult, more difficult to beat with palbociclib involved.
Sure. Well, what's interesting about this setting is that, you know, what's been the most difficult patient population to treat has been patients who lacked mutations. The drugs alpelisib and capivasertib both evaluated, have both been evaluated in wild type patients and have shown no clinical activity in those patients. And so I guess we think of it this way. If we can demonstrate activity that shows a meaningful benefit in the wild type, that augurs very well for the potential to demonstrate activity in a mutated setting. You know, the data we have, again, it's non-clinical, but compares the impact or importance of the mechanism, certainly suggests that you would expect to see greater activity when you comprehensively blockade the pathway versus only inhibiting the pathway partially.
And then, you know, if you compare the data we have in the early phase studies to the data in the mutant population that's been generated for alpelisib or capivasertib, you see a wide delta, you know, 3x delta in objective response rate reported and also in the percentage of patients who are progression free at 12 months. So we think there's a very significant delta, at least if we just compare, obviously, if all the cross-talk comparisons to be careful about. But that certainly, we think, provides a strong rationale to expect favorable results in not just wild type, but also the mutated population.
Okay. And then in the mutated population, how do you empower for that cohort with a palbociclib when that drug was never robustly studied in a global population that progressed on a CDK4/6 inhibitor, like is the enrollment criteria for your study?
Sure. Well, yeah, as Novartis, you know, because their registrational study didn't include patients with prior CDK4/6, I don't know if they were forced to or what the requirement was. But nonetheless, they did a fairly robust study called BYLieve, which was a single-arm study. So there was no comparator. But they evaluated different cohorts of patients, which include patients who've had prior, you know, they all had prior CDK4/6. And, you know, those results, we think, are supportive of making an assumption for your control. And capivasertib reported results in a similar population. And so those results vary between 5.5-7 months. And so that range, 7 months, for instance, would be a reasonable assumption to think about as a control median PFS rate.
Right. And then can I ask about market opportunity? Because I think we all see the big sales for CDK4/6 inhibitors, primarily in the front line, though. For second line plus, it's a lot of generics. It's chemotherapy. So reported sales aren't that large. How big do you think this market could be for a branded drug?
Well, you know, if our results are favorable and we, you know, meet the expectations that, or rather, our results are consistent with what we reported in the early phase study, we think the market potential, you know, if you just do the math in terms of, you know, pricing and patient population, is more than $5 billion. Now, that doesn't mean we would suggest we can generate $5 billion in the second line setting. But we certainly think, you know, it'd be reasonable to expect peak revenues in the $1.5 billion-$2 billion range in that setting. And what's that assumption or that expectation, I think, would be reasonable to have is buttressed by the revenue that alpelisib was generating. You know, and just the mutated subgroup, alpelisib was generating $500 million of revenue.
That revenue is almost all, from what we can gather, from the US because they haven't really gotten much payer approval ex US. If you normalize that for the overall population, right? They're only in this small, in this 35% group. It's basically about $1 billion run rate, which, you know, that drug only has about five months median duration of treatment, at least based on the results reported in their clinical studies. You know, it is one that is challenging to administer. I think that would be the feedback from most physicians. So despite, you know, those obstacles, it seems to, you know, it has generated a very, very significant amount of revenue.
So, we think if that drug, it's not unreasonable to think that if that drug with relatively short duration of treatment, its toxicity profile can generate, you know, essentially $1 billion for revenue in this setting, normalized, that, you know, wouldn't be ridiculous to think we could at least do that well if our results pan out the way our early phase study panned out as a floor, essentially.
Okay. Makes sense. Well, Brian, we're running up on end of time here. So I want to thank you so much for joining us at this fireside . Thank you for attending the conference, and I look forward to tracking company for the rest of the year. Thank you, Brian. It was nice to chat.
Thanks, everyone. Bye now.