Read the Full Video Transcript

Matthew Rettig: Okay. Good morning, everybody. I want to thank Jeremie and Tom and Johannes and the sponsors for inviting me today. We're going to go over the treatment landscape of metastatic castration-resistant prostate cancer. Obviously, this is a massive topic. We won't go over too many individual studies. I'll focus on some of the studies that led to regulatory approvals over the last couple of years. So the incidence and mortality from prostate cancer have been increasing year over year in the United States. Prostate cancer is the second leading cause of cancer-related mortality among American men. It's the number one cause of cancer-related mortality in male non-smokers. There were about 34,500 deaths in 2022, and by comparison, there were 42,500 deaths attributable to breast cancer at the same time. The NCI funds breast cancer by about 2 1/2 times greater funds compared to prostate cancer. So when we're talking about mortality, almost all of the mortality is attributable to this, metastatic castration-resistant prostate cancer, mCRPC, which is what we're going to be talking about.

Almost all patients, about 90 plus percent of patients in the U.S., are diagnosed with localized disease, typically, by an elevated PSA and make it through these various disease states until they get to mCRPC, and mCRPC is really the lethal form of the disease. Throughout the course of prostate cancer, interestingly, the androgen receptor is driving the growth of the disease even in patients who have mCRPC. There are some patients after multiple lines of treatment, especially AR-targeting therapies that develop a non-AR-driven disease, sometimes called treatment-emergent neuroendocrine carcinoma. We'll spend a couple of minutes on that towards the end of the talk. So these are the drugs that have been approved for mCRPC by the FDA. It wasn't until 2004 that we had our first drug that demonstrated an overall survival benefit in the form of docetaxel, with about a 2 1/2 month median overall survival benefit. Subsequently, we had several other drugs with cabazitaxel, another taxane closely related to docetaxel, sipuleucel-T, a cellular vaccine. Abiraterone acetate with prednisone was approved in the post-chemotherapy space in 2011.

The approval of abiraterone with the associated overall survival benefit was really establishing the notion that the AR still drives the growth of metastatic castration-resistant prostate cancer. Enzalutamide, an androgen receptor antagonist, was approved in 2012, radium-223, which you're all familiar with, in 2013. Then we started to have some approvals based upon biomarker-driven clinical trials, either genetic biomarkers or ultimately imaging biomarkers. So we have the PARP inhibitors for patients with homologous recombination deficiency, most commonly due to a BRCA2 mutation. Mismatch repair deficiency manifested by microsatellite instability or high-tumor mutational burden can sensitize to checkpoint inhibitor immunotherapy. Now we have Pluvicto in the post-taxane space. Finally, these doublets of an AR-targeting agent and a PARP inhibitor based upon the notion that these two types of treatment can actually synergize for their antitumor effect. I suspect we'll get an approval for Pluvicto in the pre-taxane space in the near future.

When you think about these drugs, this is the categorization that one can consider. There is overlap, of course, some of these drugs like pembrolizumab, which is biomarker-directed based on mismatch repair deficiency, is also, of course, an immunotherapeutic. So this summarizes the median overall survival benefit. On the right there in the rectangle, you can see the overall survival benefit from the various agents. Overall, there's about a 2 1/2 to 4 1/2 month overall survival benefit for these drugs, although some of them do not have an overall survival benefit and rPFS or radiographic progression-free survival led to the regulatory approval. Over here on the left, the indication most of these drugs can be given both pre and post-taxane with some exceptions. There are, of course, some that are used only in selected patients based on a biomarker, be it PSMA imaging or next-generation sequencing. So when it comes to precision oncology, we need a biomarker, typically, we think of this as a genetic biomarker, but now with PSMA, we have an imaging biomarker that we can use to select patients for therapy.

So you're going to hear a lot about these studies, the VISION study and the TheraP study, the two randomized controlled trials that have been reported for Pluvicto. The TheraP study done in Australia was a randomized Phase 2 comparing lutetium PSMA to cabazitaxel in post-docetaxel patients, and the primary endpoint was PSA response rate. What you can see is that the PSA response rate, PSA 50, a 50% decline in PSA, occurs in about 2/3 of patients who received 617 and about 1/3 of patients who got cabazitaxel. There was no overall survival benefit that was demonstrated in this study, but it wasn't empowered to detect that. Here's the VISION study, which was the study that led to the approval of Pluvicto. It improved both rPFS as well as OS as you can see. The comparison here, importantly, was not chemotherapy in the post-taxane space. It's known that cabazitaxel has more activity and improved survival as compared to a second androgen receptor signaling inhibitor, so just keep that in mind. Nonetheless, there is an overall survival benefit of about four months in patients who received the 617 as compared to a second AR signaling inhibitor.

So the other biomarker that we have to consider are genetic sequencing biomarkers. Just so everyone's on the same page, we'll talk about what a germline mutation is versus a somatic mutation. A germline mutation is a mutation that actually occurs in a germ cell. Seen here in red is a spermatocyte that has a mutation, and post-fertilization, all of the daughter cells will harbor that mutation. So in the fully formed adult organism, every cell in the body has that mutation. The gametes from that adult organism, half of those gametes will have the same mutation, and that results in an autosomal dominant pattern of transmission. Somatic mutations only occur in the affected tissue, and these are acquired mutations that develop post-fertilization. So, who gets somatic mutations? All patients with metastatic CRPC, about 25% of the patients, will have a mutation that is actionable, so we can use a PARP inhibitor or a checkpoint inhibitor.

You can also consider doing sequencing earlier in the disease course, even in metastatic castration-sensitive disease because if you have tissue that's available and don't use it, it may sit for many years while the patient has castration-sensitive disease. As that tissue ages, it becomes less useful for next-generation sequencing. Importantly, there are not too many differences in the primary tumor compared to a metastatic tumor for the mutations that we're currently interested in for the purposes of pairing with a specific therapy. So it is okay to use primary tissue, of course, the preference is fresh metastatic biopsy. For germline testing, these are the NCCN recommendations for all patients who have high-risk disease or greater, so regional disease, metastatic disease. Patients who have intermediate risk disease and have a cribriform or an intraductal pattern, they should also have germline sequencing because those patterns have been associated with homologous recombination deficiency such as BRCA2.

Then, earlier stages of the disease, patients can get germline testing based on family history, personal history of another cancer, or ancestry, for example, Ashkenazi Jewish population have an increased risk for BRCA mutations. So this is the frequency of the various germline mutations with BRCA2 being the most frequent followed by ATM. Overall, about 12% of patients with metastatic disease have a germline mutation irrespective of family history. So family history may not be captured either because there's incomplete penetrance related to a germline mutation or the patient actually doesn't know his family history. About 5% of patients with localized disease will have a germline mutation. As you can see, almost all of these genes that are involved in hereditary prostate cancer syndromes encode for proteins that repair DNA. There are some exceptions such as HOXB13, which is an androgen receptor transcriptional repressor. So its absence will lead to upregulation of the transcriptional activity of the androgen receptor, which in turn, can promote the onset of prostate cancer.

An important point when considering germline mutations, they can be detected in about 10% of patients who've had somatic sequencing and the somatic sequencing does not detect one of the potential germline mutations. There are technical reasons for that. So the absence of a relevant mutation in somatic sequencing shouldn't preclude one from offering germline testing. So there were three controlled trials that were reported last year that looked at the question of combining a PARP inhibitor with an androgen receptor signaling inhibitor. I'll go through these quickly. This is the first one known as TALAPRO-2. They all had a similar design of an androgen receptor signaling inhibitor plus or minus a PARP inhibitor. In this study, the AR signal inhibitor was enzalutamide and the PARP inhibitor was talazoparib. Some of these studies had patients who had homologous recombination deficiency as well as those patients who did not.

But the benefit and the approvals for these drugs and these combinations are limited to patients who do have homologous recombination deficiency. What you can see in the top curve is the rPFS in favor of the combination of talazoparib and enzalutamide compared to enzalutamide alone, and that's about a 12-month improvement in rPFS. There's no OS benefit that has been established. Here's a similar study for olaparib; rPFS is improved. Here we get about eight or nine months of rPFS improvement. Here's a similar study with niraparib and abiraterone, again, about a six to seven-month rPFS survival advantage, no OS advantage established. I should point out there was another Phase 3 study that looked at the question of enhancing AR-targeting therapy even more with two AR signaling inhibitors and comparing it to one. So this was ADT + abiraterone and apalutamide versus abiraterone alone; there was no OS benefit, but there is about an eight-month rPFS advantage.

Interestingly, the FDA did not approve this despite the fact that the other combinations that I just showed you of a PARP inhibitor and an AR signaling inhibitor did receive regulatory approval for patients in a very similar disease space. I want to mention sipuleucel-T; it's really underutilized in the United States. This is a cellular vaccine and what it involves is leukapheresis, taking the patient's peripheral blood leukocytes and then stimulating them ex vivo with a vaccine, which is a fusion protein between prostate acid phosphatase and GM-CSF. That is done three times over four weeks, and the resulting stimulated cells, probably dendritic cells, are reinfused after each leukapheresis, so they get three infusions over four weeks. This is extremely well tolerated. In the original study that led to the regulatory approval of this, the median overall survival benefit was about four months. But there was a signal that African Americans had a particularly large median OS benefit.

So they did a registry study called PROCEED as some showing here. Compared to Caucasian patients, it was about 2000 patients; African Americans have about a 9 1/2 month additional median OS benefit. That's massive when you think that many of the drugs that we are looking at only improve median OS by about four months or so. If you actually take the patients whose PSA was less than the median in that trial, which is about 29, the median OS benefit overall compared to whites was 21 months greater. So this is something you should really consider for an appropriate patient. This is used pre-chemo in relatively asymptomatic patients without visceral organ disease. It's difficult to use in the sense that it's very expensive, and coverage can be an issue. So just I'll end here with this neuroendocrine aggressive variant concept, so neuroendocrine carcinomas, aggressive variant cancers as the name implies, are very aggressive clinically. They have a different underlying biology than their adenocarcinoma counterparts.

For the most part, these will develop after patients receive multiple AR-targeting agents. As an adaptation to the AR-targeting agents, the tumor becomes AR null or AR independent. It may express the AR, but the AR is not transcriptionally active. The manifestations of this entity have been characterized to some extent, and what you can see is that there's a neuroendocrine type, a stem cell-like type, and double negative where there's no AR expression and there's actually no neuroendocrine marker expression. Clinically, these can be difficult to distinguish. These are differences that we're largely making by molecular profiling. It's not quite yet ready for clinical implementation. So in terms of managing this, I should also point out the neuroendocrine carcinoma is extremely rare at the time of initial diagnosis. It's probably less than one or 2% of all cases.

So we're really talking about the treatment-emergent neuroendocrine carcinoma or adenocarcinoma. So as I said, it arises as an adaptation to AR-targeting therapy, typically, manifesting clinically by visceral organ involvement, lytic as opposed to blastic bone involvement, low PSAs, short duration of response to hormone therapy. Some of the genetic lesions that we see in these variants are shown there. P53 and PTEN mutations in an adenocarcinoma can be predictive of the potential for that particular case to transform from an adeno into one of these neuroendocrine aggressive variants. These can be PSA negative; however, an elevated PSA does not rule out the presence of this entity because there's spatial heterogeneity. So a patient may have adenocarcinoma in one site and neuroendocrine carcinoma in another, and you biopsy the neuroendocrine that doesn't produce PSA.