Read the Full Video Transcript

Elena Castro: Hello, my name is Elena Castro. I'm a medical oncologist in Madrid, Spain, and today I'm joined by Dr. Alan Bryce. This discussion is part of an educational series highlighting the state-of-the-art, interdisciplinary management of prostate cancer. Today's discussion will be focused on metastatic castration-resistant prostate cancer, real-world evidence regarding patterns of care, the framework in which we make our decisions, the outcomes with some common sequencing options, and the results from recent phase three trials. Alan, I'm very happy that you could join me today. I know you've been involved in some of the most relevant clinical trials that have been conducted in the mCRPC space in the last year, and I'm looking forward to your presentation.

Alan Bryce: Well, thank you, Elena. It's my pleasure to be here today and thank you to UroToday for setting this up. So, what I want to talk about today in this series is the management in the metastatic castration-resistant space. Just as a way of introduction, I'm Alan Bryce. I'm the Chief Clinical Officer here at City of Hope in Phoenix, Arizona, and also a professor of molecular medicine at TGen, our Molecular Research Institute. Here are my disclosures. So, I think we all know that this has become a broad space now with lots of options. And so, a lot of questions arise in terms of clinical management. For the purposes of today's talk, I want to talk about the current options and then real-world evidence that goes to patterns of care because I think it is very important to think about the reality of the patient course.

When we then think about answering the questions, how do we select which treatment to give? How do we sequence treatment? How do we think about the timing of treatments? And because this is a dynamic space that's ever-changing, I want to at least discuss what I think of as my theoretical framework in which to place all these decisions, because the dynamics will change over time with every new drug approval and new data release. And then we'll talk a little bit about some different outcomes with common sequencing options. And then in closing, talk about the recent phase three trial results that perhaps some of the listeners haven't seen yet or thought about yet.

All right, this is mCRPC in 2024. We will focus on the mCRPC part of the slide, but I think about our treatment options in broad categories: chemotherapy, ARPIs, radioisotope therapies, of course, the PARP inhibitors, and Sip-T in a category of its own, and never want to forget anti-resorptive therapy. These are important adjunctive therapies in the mCRPC space that are probably underutilized in modern management. So, as I say, it's a complex space. There's a lot going on. We have opportunities, we have conundrums, but we can talk about the core principles that stay the same.

So, we're blessed with many treatment options. And the fact of the matter is, there's a lot in the pipeline, there's a lot more to come. This will be a dynamic space moving forward. But the consequence of that is that the treatment path is no longer straightforward with new biomarkers and with all the different treatments that we can give in mHSPC, because of course, what we do in the hormone-sensitive space will impact what we might decide to do in the castration-resistant space. So, all this leaves us with a lot of questions about how to make the best treatment recommendations for our patients.

I would always put forth to the clinician that patient management should be thought of as a unified arc from the beginning to the end. And so, we're really not just thinking about what's best for the patient right now, but we're also thinking about how the decisions we make today affect every other decision we need to make tomorrow and the day after. Because we want to give the patient the best overall outcome, not just the best short-term outcome for the next six months, but across the entire treatment arc. We want the highest quality of life, the lowest toxicities, and the most longevity possible from the cumulative treatment course that we provide to the patient, across multiple lines of therapy.

So when we think about that, I always say as oncologists, we go back to biology. And I always love this quote from Dobzhansky, now decades old, but saying, "Nothing in biology makes sense except in light of evolution." And of course, what we do as medical oncologists or as urologists treating advanced prostate cancer is we witness evolution in real time. I mean, what we induce in the patient as we see the cancer evolve over the course of the patient's care, but every line of therapy that we induce is in itself a form of natural selection. It will force the cancer to evolve in certain directions, and that evolution has implications as we think about what treatments should follow which other treatments and how do the decisions we make today affect where the cancer is going to go at the next step.

This real-world evidence from Dan George and his team in 2020, I think, is still very relevant. There are a few things I'd have everyone take away from it. This is looking at patients who went through multiple lines of therapy, the same patients walking through, and one part of the data set just shows you what treatments are being selected at each line, which is straightforward enough. But another part of this data set is looking at the number of patients that made it through each line. So we start with nearly 2,000 patients, but in second-line therapy for mCRPC, we've already lost half the patients. Half the patients do not move on to second-line therapy, and then you go to third-line therapy, more than half again have dropped off. So there's a steep drop-off as we move line to line. And then the other point here is, still in the first two lines, sequencing of the ARPI tends to be the dominant treatment pattern. We know this anecdotally, but we see it here in data, and chemotherapy tends to come in later in the third-line setting.

This is another graphical way of showing the drop-off in therapy from line to line. So again, you see that 20%, 30% of patients dropping off line to line, not ever receiving subsequent therapy. So part of the point I make here is, when we're thinking about treating patients and the clinician is making decisions, we have to keep in mind, you only get so many shots on goal. You only get so many opportunities to treat the patient, to intervene. Disease progresses, comorbidities accumulate. At a certain point, the patient is not going to be able to receive more therapy. So I think we have to resist the thought that I can save something for my third or fourth or fifth line. In all likelihood, you won't get to the third, fourth, or fifth. So part of my message here for the practitioner is really think about using your best drugs at the first opportunity, as early as you can because you really are only going to get so many shots on goal.

All right, so then we've seen this data multiple times. This is the CARD data looking at Cabazitaxel versus a second ARPI in patients who have had both Docetaxel and an ARPI, and I think we all are very familiar with this by now, but unequivocally, Cabazitaxel is more effective than going to the second ARPI. Overall survival of 13.6 months versus 11. PFS of 4.4 months versus 2.7. The curves show early separation and maintain separation throughout. So this is just one data set, but we have multiple data sets like this now where really the second ARPI doesn't give you much. Median PFS is more or less the first follow-up point. So when I talk about the emphasis on tumor evolution, you also recognize or think about the impact of cross-resistance between various drugs. Cross-resistance within classes of drugs and why switching within a class has limited efficacy as a general rule unless there's a distinct difference between the drugs in that class.

One of the things we suffer from in this space, though, is if we move it one step forward and we say in the pre-Docetaxel setting, what about comparing Docetaxel to an ARPI? And the fact of the matter is we have very little data in this setting. We have nothing equivalent to the CARD study. And so the one data set I would refer people to, and something I'd have them keep in mind when you think about future studies, is this data from the TRITON3 study, which I presented at GU ASCO 2023. And to remind people, this was a study of Rucaparib in BRCA or ATM mutant mCRPC where patients were randomized either to Rucaparib or physician's choice, which could have been Docetaxel or an ARPI. And the way this played out, of the 135 patients on the physician's choice, 75 went on Docetaxel, 60 went on an ARPI.

And so we have a randomized clinical trial showing a difference between Docetaxel and an ARPI in this setting, and this is also the only study we've had, and after several dozen studies comparing novel drugs to Docetaxel in mCRPC, this is the first one where anybody beat Docetaxel, right. Docetaxel can be beaten in this space. And Rucaparib was the first drug to do that. But for the purposes of your daily decision-making, I'd have you focus here. So in this setting, ARPI gave a PFS of four and a half months, Docetaxel eight and a half months. So much like the CARD study, the chemotherapy was more effective than the second ARPI after a patient had already had one. I'd also have you think about these numbers as we look at the next couple of studies where in this space, most of our randomized clinical trials are going to be randomized versus a second ARPI.

And a constant question is going to be should it have been Docetaxel, or should that have been included? And what implications would it have had had Docetaxel been the comparator? Okay. So carrying that forward, a very important data set we were all waiting for this year, PSMAfore, which is looking at Lutetium PSMA-617 in the pre-Docetaxel setting. So remember, VISION and THERAPY were post-Docetaxel. This is one of the two studies looking at a Lutetium PSMA therapeutic in the pre-Docetaxel setting. This was randomized Lutetium versus an ARPI, so crossover to the one or using the one the patient hadn't had before. And importantly, all the patients were given the option to crossover from the ARPI to the Lutetium at the time of progression. This is consistent in the Lutetium PSMA studies, this study design, and this is what patients and regulators wanted because we all know patients really want to get on Lutetium, and if they hadn't been given the option to crossover, it probably would've been very difficult to accrue. So I regard this as a patient-friendly and patient-centric design.

The primary endpoint here was rPFS. And you could see this is very positive, unequivocally positive, right. Hazard ratio is 0.43, median PFS 12 months versus 5.6, unequivocally positive. I think we could feel very good about this result. Here's some of the secondary data. You could see the objective response rates strongly favoring Lutetium PSMA, almost 50% or over 50% complete and partial responses versus only 15%. PSA change favoring Lutetium as well. But probably the biggest concern about this study right now is the overall survival data in the interim analysis. So not enough events for the final analysis yet, you see that there's really no advantage. The median here is 19.2 months versus 19.5, in fact a little bit favoring the ARPI. But importantly here and highlighted in red, the crossover is 84.2%. So 84% of the patients crossed over from the ARPI to the Lutetium.

And what that means in practice is that this was really a clinical trial of Lutetium right after one ARPI or Lutetium eight months later. Because I asked Oliver Sartor about this, I said, "What was the median time from progression to a patient crossing over?" And it was very tight. It was only several weeks. So this is really a clinical trial that delayed Lutetium by only several months. And what you see is that the outcome's not much different. So this leaves open a question, will this be approved by regulators? Is this result going to be enough? I don't know the answer.

I won't be shocked by either outcome. We'll see what regulators do, but in my opinion, the crossover design almost should set an expectation that overall survival's not going to be much different. So we'll see. We'll see. We need further follow-up on this. And then we have the SPLASH trial. So the SPLASH trial is a different PSMA, Lutetium radioligand therapy. It uses PNT2002 rather than PSMA-617 as the targeting agent. But the same radioisotope, almost the same design as PSMAfore. And what you see is in the press release, which is all we have so far, very similar results, which is yes, it meets its primary endpoint of a PFS advantage, but the interim overall survival showed no difference. So again, similar to PSMAfore, we haven't had the full data release yet. So not much more to say at this point. So stay tuned. We'll see this at a future meeting.

So one of the most highly anticipated data releases for GU ASCO was the CONTACT-02 study. This was a study looking at Cabo and Atezolizumab versus a second ARPI or NHT in patients with MCRPC who had already progressed on one ARPI. They selected for high-risk patients, needing patients with soft tissue measurable disease. And so ultimately, there was enrichment for patients with liver metastasis and visceral disease. Primary endpoints here are PFS and OS, nothing really remarkable in the baseline characteristics, but again, emphasizing the high risk. So the visceral disease, 39%, liver metastasis, 23%, de novo metastatic disease, 52%. So again, a very high-risk population. So the primary endpoint of PFS, 6.3 months versus 4.2. This is technically positive, hazard ratio 0.65, so the study met its primary endpoint, but I think there's a lot of concern.

When you see this curve, there are a few things you take out of it. One is that the progression events in a stair-step like this are happening at the time of assessment. When you see the stair-step, it really means patients are not progressing symptomatically; the progressions are being caught at the scheduled imaging intervals. And the gap between the two arms is really just one scheduling interval. So not to get too into the weeds, but what this means if you're familiar with clinical trial design, is that the schedule of assessments is driving the difference in the median PFS here. And if you were to model something like this, scanning patients at different times, 10 weeks or 12 weeks or 15 weeks instead of nine, you'd get significant variation in the difference here. The gap is almost an artifact, in some sense, of the study design. It's hard to get excited about this.

And the other point, I think we all know, emphasizing the results from the Rucaparib study, the second NHT is the inferior potential treatment choice here. The truth is the patients had the choice of going on to Docetaxel. And one wonders, would this be positive if Docetaxel was a comparator instead of Cabo Atezo? We don't know. This is a very high-risk population, but it raises a significant concern. So this is, I think, in my opinion, a little bit of a disappointing data release. We'd like to have seen more of an advantage here. Probably the most remarkable and interesting result is the result in the liver metastasis, really a very strong hazard ratio, PFS of six months versus two. I think this is an intriguing result. The rest of this is very median, not perhaps as exciting. I think the prior docetaxel probably reflects the fact that the patients who got prior Docetaxel were probably the ones that had liver metastasis. This is probably the same population. So a multivariate analysis here would be useful.

But here's the interim overall survival, no difference. I mean, really, hazard ratio 0.79, 16.7 months versus 14.6. Interim analysis, we still can wait for maturity. We'll see how this changes with time. So in summary, the management of mCRPC has become increasingly complex as new treatment paradigms have developed and new drugs have been approved, which I always say is a great problem to have. This is exactly the position we want to be in with lots of options. But with this many drugs in the toolbox, it's likely that many of your patients will reach the end of the road before they exhaust all treatment options. So I would emphasize for the listener, the key is not just to make the best decision at each step but to make the series of decisions that lead to the best cumulative result for your patient.

So a key operational principle is to use your best drugs as early as possible and don't hold back. I recommend that you think about classes of drugs and consider how switching or combining classes can have advantages from the perspective of disease evolution and not just sequencing. And then finally, I'd say after several years of continuous success, the development of new classes of drugs for prostate cancer has hit maybe a bit of a lull, but this is temporary. The pipeline is still full. And since we have yet to cure prostate cancer, you can count on a lot of ongoing development. And please, please continue to accrue to clinical trials so that we can get better options for our patients. Thank you very much.

Elena Castro: Thank you, Alan. This has been a very, very nice presentation, and I would like to ask you about something you mentioned and is that it is very clear that the benefits that our patients get from sequential hormones depend of course on the characteristics of the disease. We saw in the PSMAfore in which perhaps patients had better outcomes, that the median duration of the response until time to rPFS was five months and a half, whilst in the contact two it was about four months. And for those with liver metastasis, it was even shorter. So it seems that we are going to be in a range between five months and two months, depending on the patient's characteristics. But then all the new trials that we have assessing new therapies continue to use a second hormonal agent as control therapy. So on one hand, we have evidence that says we shouldn't be doing this, but the new therapies are being tested against a second hormonal agent as well. So what is your take on this?

Alan Bryce: Yeah, well, I think there are a couple of pieces there. Number one, when you talk about the range of results, I think we are going to get increasingly sophisticated in terms of really categorizing patients more effectively, clinically in terms of say, high-risk features, liver metastases, etc. But also genomically I think we'll get to the point of DDR not being the only biomarker we have. We'll have an increasing panel of them. In terms of what is the right comparator in a clinical trial context, I mean, that really is a question for regulators. I mean, we as a field can say we want to see more active comparators. And yet, even while we say that, you look at the real-world evidence, and I think there is an absolute legitimate argument for somebody designing a study to say, "Well, this is the standard of care that's being used."

So to call it a standard is accurate, and I think that's a fair argument to make. Having said that, I hope we will get to a point where we'll be able to say, "All right, we know there's a quarter of patients in whom an ARPI switch works really well. We can define who those are. Let's define those patients where we know it's not going to work well." And as you know, we did a study of physician's choice, and when given a choice, a majority of patients chose Docetaxel. So I think we do a little bit of a disservice to the field when we say, "Well, nobody wants Docetaxel, nobody wants chemo. Everyone's going to choose the oral therapy." I actually think with adequate advisement and adequate input from their physician, a lot of patients ultimately are okay choosing chemo. And we went through this before in RCC, didn't we? All those clinical trials that were randomized versus Sorafenib, even though who was using Sorafenib? But it's just a characteristic of the regulatory pathway.

Elena Castro: Thank you, Alan, and thank you to all of you for listening.