Impact of Clinical Data on the Treatment of mCRPC

Eleni Efstathiou, MD, PhD: Can I make a comment, Dan?

Daniel J. George, MD: Please do.

Eleni Efstathiou, MD, PhD: I happen to be a CARD investigator and the reviewer for ESMO on the discussion for the PROfound trial. I disagree a bit, Tanya, on the fact that I never considered shifting from one androgen signal inhibitor to the other as a valid option for our patients. I'm not saying that because we can’t disagree on the subject before this data, but when this trial started, I thought, “This is a no-brainer study. Why are we doing this study? Why are we doing it? This is a waste of our time.”

It turns out that you are absolutely correct. In practice, a lot of us, even today after 7 months, and within The University of Texas MD Anderson Cancer Center, people shift from one to the other agent. Now, we have clear data. If you look at both trials, for me, shifting from one to the other was the equivalent to placebo, based on the outcomes. Within 2 or 3 months, all patients progressed on both of those, if you look at them carefully. This is where I'm saying , we need to go out there, as you said, and educate more people to not first consider going from abiraterone to enzalutamide and from enzalutamide to abiraterone.

Think about it more. It’s more of a Ben question…because it's harder to do the referral. You have to go to a different doctor to do the chemotherapy. It's also the fact that the old TROPIC data with cabazitaxel was so scary for a lot of people. There were 5% toxic deaths, but still, for those of us who use it every day, many of us can say that cabazitaxel is so much better than docetaxel. We got the FIRSTANA data to speak to that point.

A lot of things that we sometimes have not processed enough until we actually get into the practice of using it came out of these 2 trials. The thing about PROfound that was a bit of an issue for a lot of physicians is that it didn't show an overall survival benefit, but 80% of the men who were on the placebo arm were immediately offered to go on to olaparib within a couple of months, so we can't judge anything from the overall survival data.

I was a bit surprised by the approval across the board, but that was to be expected. One of the things, as you pointed out, Tanya, is that they were largely BRCA2, BRCA1. And the ATM data, it was a bit of a mixed bag there, so we need to get better at it, but that was truly groundbreaking because it opens the door immediately for the approval of targeted therapy for prostate cancer. That's the first one. That ESMO session was for me, important, at least for 2019.

Charles Ryan, MD: I would like to point out that although I'm very enthusiastic about the PARP inhibitors, and just as excited as everyone else, it's really important to notice that despite the presence of somatic BRCA2 alterations, not everyone responded particularly well to these therapies. When I saw that, I wondered why. There were some really interesting data that are coming out. PROREPAIR-A was a study that was presented at ASCO this year, which looked at the presence of 1 or more of these alterations.

I think this is the next level we're going to get to. Right now, we're thinking, “Is there BRCA2 mutation, yes or no?” If there is, great, we've got a PARP inhibitor. That's true, but when we see RB [retinoblastoma] loss and other genomic alterations, we know we're dealing with a different prognostic category. This gets back to what we said earlier, which is that we're beginning to talk in that language about the genomic complexity of the disease. It’s not simply whether a mutation is present or absent, but about the totality of the biology, which is really encouraging and exciting.

Daniel J. George, MD: Doesn’t it also speak to your point about tumor volume and heterogeneity? All of those patients had to fail abiraterone or enzalutamide. They're not only metastatic castration-resistant, they're 1 step further. In the nonmetastatic, castrate-resistant setting, if you had someone with a BRCA2, you would imagine that PARP inhibitors would work even better.

Charles Ryan, MD: Right. If we can attack the BRCA2-altered prostate cancer before it loses TP53, or before it loses RB1, we're going to see better results. This is just the general theme that we're talking about, which is that earlier may be better. Simpler biology is better, etc. I think that the PROfound data, while really encouraging, and the rucaparib data also, opened the door to further questions and opportunities for us.

Tanya Dorff, MD: There’s also a lot of interest in combining some of these agents. There's this whole question about whether having those mutations also makes you more sensitive to immunotherapy with a checkpoint inhibitor, which we know has performed relatively poorly in prostate cancer in unselected populations. But the olaparib with durvalumab data are interesting, because you wouldn't have necessarily expected to see responses in an unselected population.

They had a 12-month PFS of 36% in patients without the mutations that you would use to predict sensitivity to the PARP inhibitors, and that's much higher than with checkpoint alone. It’s much higher than you would expect with PARP inhibitors alone, so is there some synergy there? I think that's one direction the field is moving in, in addition to the next generation of PARP inhibitors that may have broader effects or differential effects. Olaparib shows some responsiveness with ATM. Rucaparib did not. I think we're going to have to get into the weeds, to Chuck's point, about how complex it is to understand the genomics and the heterogeneity, and what that means when we're choosing treatments.

Charles Ryan, MD: Yes, it’s almost ironic, Tanya, because you're absolutely right. The irony is that we're also seeing data on the combination of abiraterone and PARP inhibitor in the PROpel data and in others. There are studies launching that are looking at PARP inhibitors plus AR [androgen receptor] inhibitors in the unselected patients. We may be going into the front door, saying we're going through all this genomic selection and have these targeted therapies, and then go out of the back door saying you don't need to do that in everyone.

Eleni Efstathiou, MD, PhD: You brought up the very nice experience we had at ASCO. There's one more thing that's coming out for the DDR patients, and again, we don't want to go into the weeds because the PROfound data went widely for anything you can pick out. Monoallelic was good to go since that specific assay, and we had some answers with regard to going from monoallelic coming out for some of the papers that had been published that actually supported that.

Some of the other approaches have been more purist in going for biallelic alone. I think we want to walk away a bit from being so selective moving forward. The approvals that we got from the FDA are speaking to a specific assay, but maybe we want to comment that moving into the future, we might be using other assays, as well. Cost is an issue and we don't want to be bound. For now, the approval for olaparib was for a specific Foundation Medicine assay companion diagnostic. The Myriad Genetics, Inc, I believe, is for germline assessment.

I wanted to bring up one point that came up from ASCO, as well. It is likely that those men who have tumors that harbored DDR alterations may not be at all responsive to taxanes, and that had come out in a retrospective analysis from Elena Castro, MD, et al. There was a poster about looking prospectively at such candidates where cabazitaxel really underperformed. It performed as badly as transitioning from abiraterone to enzalutamide and from enzalutamide to abiraterone. There’s more to look into in the future about selection of these types of patients who have more aggressive phenotypes based on molecular characteristics.

Transcript Edited for Clarity