Read the Full Video Transcript

Oliver Sartor: Hi, I am Dr. Oliver Sartor with UroToday and delighted to be able to have a guest from Memorial Sloan Kettering, Dana Rathkopf, who's going to be talking about some very provocative data that she presented at ESMO 2024 in Barcelona. Dana is a professor at Memorial Sloan Kettering, and I'm delighted to have you here today, Dana.

Dana Rathkopf: Thank you so much, Oliver, for having me. I'm delighted to present some of the data that we shared at ESMO 2024 of BMS-986365, which is an androgen receptor ligand-directed degrader. BMS-986365, or 365 for short, is an orally bioavailable, first-in-class, heterobifunctional molecule, and what that means essentially is that it is a cereblon binding moiety that facilitates targeted ubiquitination and subsequent proteasome system-mediated AR degradation. It also has an AR binding moiety, and this binds and antagonizes the AR ligand-binding domain with low agonism potential. So it has a dual mechanism of action that both degrades and competitively inhibits the androgen receptor. In preclinical studies, this 365 compound has demonstrated activity against both wild-type and mutant AR.

In the first-in-human phase I study that we recently presented and which was published simultaneously in Annals of Oncology, there was a dose escalation phase, including men who had at least one prior ARPi for castration-resistant prostate cancer, and they also needed to have had prior taxane or be unfit for, intolerant to, or declined. And so in this study, approximately 50% of men, a little bit less than 50% of men, had had at least two prior ARPis and at least one prior taxane.

For the dose expansion cohort, we enrolled four cohorts and we expanded the inclusion criteria so that patients with no prior platinum chemotherapy, no history of brain or liver metastasis, and a PSA greater than or equal to 2 were included. And in this cohort, what we saw was that at the 600 milligram daily cohort—these eight patients—we closed early due to lack of durable efficacy, and so the efficacy data that I'm about to show you is only on the BID doses in this part B dose expansion cohort. So this was the safety data. In the part A dose escalation, the MTD was not reached. There was one DLT of asymptomatic prolonged QTC at the 900 milligram BID cohort.

And then in the table what we're looking at is the part B dose expansion. So the most common any-grade treatment-related adverse event in part B was prolonged QTC, and this happened in 47% of patients. For grade 3 treatment-related events, these occurred in eight patients, or 12% of patients, and six of these were asymptomatic prolonged QTC. And just to remind you, a prolonged QTC or a grade 3 QTC would be a change in QTC interval of 60 milliseconds from baseline or a QTC that's greater than 500. All of the prolonged QTC events occurred within the first two cycles and were managed with dose modification. There were no grade 4 or 5 treatment-related adverse events. Serious treatment-related adverse events occurred in three patients. All of these were at the 900 milligram BID dose cohort, and there were no treatment-related adverse events that led to drug discontinuation.

These are the PSA responses by prior treatment and dose, and what you can see is there appears to be a dose-related PSA response so that at the highest value, the 900 milligram BID cohort, the PSA30 was 70% and the PSA50 was 50%. When we look at the radiographic progression-free survival, again, this is in the part B BID cohorts, for all patients, the median rPFS was 6.3 months, and 42% of these patients were free of radiographic progression or death at six months. If we look at a post-hoc analysis by prior chemotherapy, what we saw is that patients who had no prior chemotherapy had a median rPFS of 16.5 months relative to patients who had prior chemotherapy with 5.5 months.

Now looking at the PSA response and median rPFS in patients based on the AR ligand-binding domain mutational status, about 72% of patients were AR wild-type, and 22% had a ligand-binding domain mutation. What you can see here is that there appears to be clinical benefit observed in patients with wild-type and mutant CRPC. And I should mention that the wild-type included AR amplification.

That is what we presented at ESMO this year. Essentially, we felt that this is a first-in-class, dual AR degrader and antagonist. It has the potential to overcome resistance to ARPi regardless of AR ligand-binding domain mutational status, and it was well tolerated and easy to manage in the clinic.

Oliver Sartor: Great. Well, thank you, Dana. One of the things that caught my eye a little bit was the 16.5 months on the rPFS with no prior chemo. That's a pretty impressive number, actually. We've been running a lot of phase IIIs with the RP switch and we kind of know what to expect. It's coming in anywhere from 4.5 to maybe six months. So 16.5 seems to be a bit of an outlier there. Just your initial thoughts.

Dana Rathkopf: Yeah, we were excited by that data. I think the difference with this particular compound is that it has a dual mechanism of action. So when we look historically at ARPi switch, I think often we'll see around a six-month period of time in this setting, pre-chemotherapy. But remember, this drug not only degrades the AR, but also is a potent AR antagonist and it has low agonist potential. So we think that this dual mechanism of action may actually potentiate the response to the compound, and this will be the basis for a phase III exploration in this setting.

Oliver Sartor: I'm a little bit curious. You've got some ligand-binding domain alterations. I wonder if you can drill down into that just a little bit more and talk about which mutations might've been included. Did you do 702? 742? 878?

Dana Rathkopf: Those are the ones. We used, I think it was called the Predicine Care cell-free DNA targeted panel, and it was looking at the most common AR ligand-binding domain mutations.

Oliver Sartor: You mentioned that the AR amplification went into the wild-type. Do you know how you did in the amplified setting, because that's been a little bit of a bugaboo for the degraders as a whole? When AR gets amplified, it doesn't seem to work very well.

Dana Rathkopf: Yeah, the numbers were too small, really, for us to drill down or say anything conclusive. Mostly what we saw in the clinic, though, is that it matched what we saw in the preclinical data, that there was activity regardless of the AR status of these patients. Again, we are hypothesizing that it has something to do with this really unique part of this compound, which is both an AR antagonist with low agonism potential and an AR degrader as well.

Oliver Sartor: One of the things from other reading that has attracted my attention a little bit, and this has come out from multiple groups, about the AR enhancer, the upstream enhancer that has high prognostic importance for these patients. Has anybody looked at the enhancer and how that might interact with this particular moiety? Probably not, but just raising the issue.

Dana Rathkopf: No, that's a good question, Oliver. We're still looking. This is a phase I/II trial, and we're still looking at the data. We have this ctDNA, we have biopsies at baseline. We haven't looked at that currently, but certainly something that we can bring back to the company to ask about.

Oliver Sartor: One of the things that I think has been a little bit interesting is really trying to figure out which one of these tumors or which patients might be actually AR-driven. And I don't know if you've looked at some of the other mutational patterns and see if that might play a role. Probably not, again, but things like RB loss, PTEN loss, et cetera, may sort of take patients away from the AR-driven pathway. But I don't know if you've kind of looked at non-AR elements within these analyses. If not, it's a good thing to look at.

Dana Rathkopf: The numbers are so small that to make any conclusive statement would be potentially misleading, and I don't want to do that. But we certainly looked at the typical predictors of aggressive cancer—PTEN loss, RB loss, TP53—and as you might expect with any AR-directed compound, the sense is that perhaps these patients don't do as well with AR-directed therapies. But again, the numbers are small, such that I don't think we could say that definitively. And I think that's similar to what we see in clinic. We sometimes see patients with these alterations or these mutations, and we expect they're not going to do well necessarily on an AR-targeted therapy, and sometimes these treatments still work. So we want to be really careful about how we're characterizing these patients, but we are looking at IHC on those tumor biopsies that we had pre- and post-treatment, actually, in some patients as well.

Oliver Sartor: Got it. The next question, I don't want to put you in any awkward position, so please don't answer it if it's not appropriate, but going forward, do you look at this sort of as an all-comer type of drug or do you look at selectivity in terms of patient selection?

Dana Rathkopf: No, that's a great question, Oliver, and we think about this a lot because I think, as a field, it's important for us to tailor our treatments to patients most likely to respond. We have all of this information and all this technology that the goal I think we all have is to make sure that we're using treatments for appropriate patients.

For this particular compound, though, in looking at the data, it's very difficult to pull out a population that is least or most likely to respond, except for patients that in the part B cohort, we did limit the PSA to greater than or equal to two. We tried to make it such that it wasn't a small-cell or neuroendocrine type of population. That tends to be atypical in general and not always as AR-driven. But in terms of selecting a population for the phase III, aside from selecting patients who haven't had chemotherapy yet, based on some, again, very small numbers, but some compelling numbers with the 16.5 rPFS, I don't think that the company is planning to necessarily select patients based on any particular AR status. And I think that's just because we don't have enough data at this point to say that the drug does or doesn't work in any specific population.

Oliver Sartor: Yeah. Well, I hate to say it from my perspective, the LBDs look pretty good—ligand-binding domain mutations look like they were particularly provocative—but of course we didn't get a full breakdown. And when you start looking at it, those are not extremely common. It's certainly a minority. But you've got activity. Again, to me, the outstanding feature was the 16.5.

Let me switch over just a little bit to the QTC because I'm not very well versed in cardiology stuff, but I remember QTC used to be considered to be, "Oh, watch out. You might have arrhythmias," and stuff like that. I'm sure you have cardiologists and a team that sort of looked at that pretty carefully. So from a safety perspective, at least so far, not much in the way of actual morbidity from the QTC prolongation. Seems like most were asymptomatic, and then you just managed it with a dose adjustment.

Dana Rathkopf: Yeah, absolutely. Because this was a first-in-human phase I trial, we did a tremendous number of EKGs. I don't know if tremendous is a medical term, but we were doing EKGs in triplicate at multiple time points and being very cautious. And again, a 60 millisecond change in QTC is what drove the majority of these QTC prolongations. But that doesn't mean that the QTC was greater than 500, it just means that there was a change of 60 milliseconds, and a lot of these patients were on multiple medications. So many medications can cause QTC prolongations, as you know, that it didn't really feel uncomfortable in the clinic. We just were extra cautious about it in the clinic. We just wanted to make sure we weren't missing anything. It was very easy to manage. It looked to be proportional with dose. So when we looked back at the PK and the QTC prolongations, it looked like there might be some relationship between that. And all of the QTC prolongations responded to dose modification. There were no discontinuations of treatment due to QTC prolongation, so it was very easy to manage in the clinic and a completely asymptomatic finding.

Oliver Sartor: Got it. Well, that's quite provocative. We wanted to have you on the show because it caught my eye when the presentations came out and it still catches my eye. So first of all, congratulations on helping to lead a really nice new drug and early in development, so pleasure to be able to see that. Before we wrap up, were there any other comments you think our listeners would like to hear? Unique mechanism, being able to move it forward, manageable toxicity, interesting rPFS in the pre-taxane space particularly. Those are my highlights, but any summations or things you want our readers to listen to before we sign off?

Dana Rathkopf: I appreciate that, Oliver. I think I'd just like to say, in the clinic when we're looking at our patients and they come in for treatments, in general, AR-targeted therapies as a class of treatments are relatively well tolerated, right? Patients don't have to come in frequently. They can get a prescription and travel and carry on their daily activities. There's not a lot of issues that interfere with their daily routines. And so I think a lot of our patients, even in the setting of progressing on a prior ARPi or prior treatments, appreciate an opportunity to try another AR-directed therapy in this setting. And so this compound really offers a novel mechanism for doing that.

Again, it's not just an AR degrader, it's also an AR antagonist first-in-class, and so it's pretty exciting in terms of how tolerable it was for patients to manage and to stay on treatment, especially patients who—imagine having had a prior ARPi and then having this as an option for another year plus—I just think it's an exciting new drug class that offers additional opportunities for our patients.

Oliver Sartor: Well, thank you. Very provocative. Great presentation. Thank you, Dana, for being on UroToday.

Dana Rathkopf: It's a pleasure. Thanks for having me, Oliver. I appreciate it.