Transcript:Raoul S. Concepcion, MD, FACS: Let’s circle back a little bit about the AR-V7 and the role of taxanes again. We know that there have been a couple of papers that suggest that if the patient expresses this splice variant, is positive for AR-V7, you shouldn’t give them an androgen axis modulator, but taxanes apparently have some efficacy in these patients. Give us the proposed mechanisms behind that?

Jorge A. Garcia, MD, FACP: I think one of the challenges of the splice variant topic is what Chuck mentioned. Not every patient has it, number 1. And number 2, the biggest utility of a marker like this, in my opinion, would be a cost containment situation. If you were to have a patient with a splice variant, then you know that you’re going to save that patient access to expensive therapy. But I remind you that it’s only around 25% to 30% of patients where you see that splice variant.

So the question is what happens when there are 70% of patients, and how those patients are actually becoming resistant to existing agents that we have. Now, I think we do not understand well the data with the splice variant and whether or not harboring that splice variant will impact your ability to respond to taxanes.

What is more intriguing to me is if you look at the existing data as to what happens to a patient that undergoes primary therapy with either enzalutamide or abiraterone and how you manage the patient when they become resistant to that therapy. If you have a splice variant, I think with the limited data that we have, I probably wouldn’t put that patient on an oral agent.

But the question also becomes if you don’t have that splice variant and you get on therapy, are you developing, in the future, a similar ligand binding mutation that can actually lead to resistance? And that’s something that we don’t know.

So for most of us, I don’t think I would use the splice variant to make my choice as to who needs chemotherapy. I think that it remains a clinical question. It remains who’s in front of you in the office and what are the needs of that particular patient. But I think it’s important to emphasize that, for instance, back-to- back oral therapy is just not the most intelligent strategy to undertake in the context of where we are.

We know that if you start with abiraterone and then you move the patient to enzalutamide, the likelihood of benefit, whether it’s PSA measure and/or progression-free survival is pretty low. And if you start with enzalutamide and move the patient, whatever agent you start, back-to-back oral therapy is not the smartest strategy with the data that we have.

So, the biggest question becomes if you have someone who actually progresses on an oral therapy, knowing that back-to-back oral therapy is not the best therapeutic strategy, is what did you do for that patient. I think most of us outside of a clinical trial would probably move that patient to systemic chemotherapy.

Charles J. Ryan, MD: I just want to make a quick point. I think that the AR-V7 story for all of its excitement is just one piece of the molecular profiling that we’re beginning to do. There was a paper in the New England Journal of Medicine about BRCA2 mutations and their importance in prostate cancer.

I think moving forward, it’s simply not going to be acceptable to not understand some of the molecular biology of the disease. Androgen receptor, glucocorticoid receptor, DNA repair, these are all emerging. These are all coming to prostate cancer from other cancers where it’s done more routinely. And I think that in five, 10 years if we were to repeat this conversation, we’ll have a whole panel of molecular profiles we’ll need to be looking at.

Michael Fabrizio, MD, FACS: But isn’t that the exciting thing right now with taxane therapy is that really it’s independent of the androgen receptor. They work, so that’s the good news. If you look at the data, we’re even looking now at being able to identify molecularly which taxane may be better, depending on markers such as the retinoblastoma gene, which was developed at Thomas Jefferson University, and showed that cabazitaxel may be a better drug in that setting. So I think we’re really getting down to some benefits of even the taxane therapies here.

Raoul S. Concepcion, MD, FACS: Judd, do you think we need to be more aggressive, understanding that these are mutations and that prostate cancer, especially in the CRPC is a polyclonal disease? I personally don’t think so and I’d be curious to see what you guys are doing at Duke. Are you being more aggressive in terms of doing biopsies of the met, especially bone?

Judd W. Moul, MD, FACS: Great question. So we have a protocol and they’re selectively doing that but it’s for research purposes. And whether that will play out in the trenches with routine clinical care, I don’t know at this point. My suspicion is it will because we’re just starting to scratch the surface, and the technology is improving. Let’s say a urologist tuning into this program has a patient with newly diagnosed early M1 CRPC and he or she is trying to decide on an oral agent, is there a phenotype of a patient who should go directly to chemotherapy. Aside from the molecular standpoint, when should the urologist not do an oral agent and say from a practical standpoint, I need to get this guy to a medical oncologist because he needs chemotherapy immediately?

Jorge A. Garcia, MD, FACP: I think that the data with the splice variants gets really validated, and it’s indeed the reason why someone should not move forward to an oral therapy. It doesn’t really mean then that patient should go straight forward to chemotherapy either. Because if you have someone who has no symptoms, who feels great, who’s interested in quality of life but who harbors that splice variant, it doesn’t really tell me that that patient needs chemotherapy immediately.

Because timing in that context also hasn’t been tested. We don’t know if starting someone on chemotherapy when you have one bone lesion is better than when you have two bone lesions. Granted that we now understand that maybe earlier for selected patients is better. But I think the biggest question when I think of this type of testing is I think that if I have prostate cancer and I happen to harbor that mutation, then I certainly don’t want to be on that oral therapy, right? But it doesn’t tell me what is the next best treatment for that particular patient and that’s the biggest question.

So I think most of us, at least in the GU medical oncology space, continue to utilize clinical features, prognostic factors that have been tested over time to define who’s the right candidate for chemotherapy. And I don’t think a molecular test at this point is prime time for us.

The data from de Bono that were just published in the New England Journal of Medicine discussed DNA methylation changes and whether or not PARP inhibitors are the best therapies for patients who may develop epigenetic changes in their tumor over time. If you look at their data, only 30% of patients have those epigenetic changes, and if you have a patient, for instance, that has a BRCA2 deletion specifically who becomes castration resistant, would that be enough for you genomically to say, I’m going to put you on an agent that actually targets that? I don’t think we’re there yet, and one of the challenges with genomics is that we continue utilizing archived material to define what happens in the metastatic deposits.

And the challenge also comes that biopsying prostate cancer patients in the CRPC setting is very challenging because most of them will have only bone disease. Not a lot of patients have soft tissue disease, and whether or not you have three different genomic aberrations which one will be the one that you want to target therapeutically.

Charles J. Ryan, MD: We’re doing that on a large scale with part of our Stand Up to Cancer dream team program, done about 200 of these biopsies now. And it’s possible, and I think that they’re going to be meaningful, but the pathologists and interventional radiologists need to be ready for this.

I would also point out that there’s the genomics, which is obviously very important. There’s the somatic mutation issue versus germline mutation issue, which is going to need to be sorted out. And then, actually, histology of these metastatic tumors can tell a lot. You can’t go to a prostate cancer meeting now where there’s not a number of presentations on neuroendocrine and small cell phenotypes, and many people have their own definition and outcome analysis of an individual with small cell or neuroendocrine prostate cancer. And that’s something where we need to get uniform language and uniform outcome analysis on. And the conversations are happening. It’s just that we don’t have the answers yet.

Raoul S. Concepcion, MD, FACS: Correct. I think it’s fair to say, like you had alluded to earlier, is that where we are in prostate in terms of molecular testing, we are 10, 15 years behind what everybody else has been doing in breast, colon, and lung cancers.

Charles J. Ryan, MD: Hopefully not 10 or 15 any more but it’s still behind.

Raoul S. Concepcion, MD, FACS: Great discussion and lots of great points. Again, I think this is something that I believe all of urologists need to have some basic understanding that we have these mutations and theses variations. Putting people on these drugs, whether it is LHRH therapy, androgen axis modulators, or chemotherapy, changes the tumor environment. We induce these mutations and so this begs the questions later on about sequencing, combinatorial, layering of therapy, whatever verbiage you want to put on there.

Transcriptn Edited for Clarity