Johann de Bono, PhD, MB, ChB: Currently, we’re waiting on renewed guidelines, newer guidelines from the NCCN and for organizations like ESMO to really put forward recommendations. Regarding my personal recommendations, I think a man with advanced prostate cancer should get somatic (ie, tumor tissue) testing of their DNA. If they have a mutation, in their tumor biopsy, of these DNA repair genes with good quality control DNA—because, obviously, you can get quality-control—failed tissue, poor tissue—if you’ve got good sequencing results and you have a somatic tumor mutation with a high allele frequency, 50% or so, that would suggest a germline mutation, and you should do germline testing. If there’s nothing in their tissue and their tumor tissue, you may not need to do germline testing. The reason I’m saying this is that I do think that all these men should be getting sequencing now because this is likely to impact their therapy with drugs like carboplatin and PARP inhibitors. And while we don’t yet have randomized trial data that really prove the benefits survival-wise, the data are very compelling.

Chris Parker, MD, FRCR, MRCP: I can see testing is worthwhile for clinical trial eligibility, but I’m wondering whether or not it impacts on practice in the clinic. If you’re thinking about DNA repair defects, why not just give a trial of platinum? We can’t use PARP inhibitors in the clinic right now, but we can use platinum. So, why test for BRCA? Why not just use platinum?

Johann de Bono, PhD, MB, ChB: Well, if you gave platinum to everybody, I think you would be causing significant toxicity with minimum benefit to two-thirds of patients. I take your point, but if I was the patient and you were treating me, I would mandate you sequence my tumor for these genes because there’s a 1-in-3 chance that I would actually have these repair defects that can result in complete responses. Now, let me actually push back on what you’ve said. The reason I’m pushing back is that we have data from other diseases with the same genomic aberrations—ovarian cancer, breast cancer, now even in pancreatic cancer—that, actually, these cancers with these same aberrations as prostate cancer get survival benefits from platinum, from PARP, and get complete responses. These are very bad prognostic cancers. So, while I do take your point, you have to not only use prostate cancer data. I think you have to look overall at all the data for other cancers. And I know, Bertrand, you want to comment.

Bertrand Tombal, MD, PhD: I’m thinking of the concept that you brought up, which are the nonresponding patients. I think that especially with the first-line treatment of advanced patient who are on ADT, nonresponding patients deserve more attention than we give now. The problem is that we have only defined responses in the clinic based on PSA. That’s a problem. We should be more careful in being more stringent of what a responding patient is. Nonresponders deserve more attention.

Johann de Bono, PhD, MB, ChB: It’s interesting that, actually, these patients with the BRCA gene often have very low PSAs, which makes these patients very hard to manage. I think that’s a salient point, because it really is that more physicians are treating patients just following PSA and they don’t do scans.

Joe O’Sullivan, MD, FRCPI, FFRRCSI, FRCR: For sure.

Johann de Bono, PhD, MB, ChB: I think if there’s 1 message that I would like to give through this talk today, for any treatment, it’s to make sure that you’re using scans as well as PSA to monitor therapy outcome.

Chris Parker, MD, FRCR, MRCP: The practical point I wanted to make is that people need to think of using platinum. So, if you’re my patient and you’ve had all the standard therapies, you could have palliative care, but if you can’t get access to BRCA testing, it’s at least worth trying or thinking about using platinum.

Johann de Bono, PhD, MB, ChB: Yes, you could do that. You can give carboplatin AUC 6 mg/mL 3 times weekly based on the Cockroft-Gault formula. You could give it for 1 course, 3 weeks, or for 6 weeks, 2 courses. It’s well tolerated in this population. I’ve done it in many patients. If the patient clearly responds, continue, and if after 1 or 2 courses they don’t respond, you can stop. I have to be honest, that will result in overtreating a lot of patients. The cost of a gene test today is, in the United States, $100. That’s all it costs, although many companies are charging a lot more than that. In our lab, it costs £300 or £250. This is not a difficult test. So, I think it would be very sad if patients don’t have access to these tests.

Bertrand Tombal, MD, PhD: Across Europe, we can base a business case demonstrating that in patients who are diagnosed with very aggressive disease—they’re 65, they show up with advanced metastatic disease—these tests, sequencing the genome of these patients today, are less expensive than 1 month of the new abiraterone.

Joe O’Sullivan, MD, FRCPI, FFRRCSI, FRCR: Of course.

Bertrand Tombal, MD, PhD: And so, that’s why we have to do it, because I agree with Johann. It may not change the initial ADT plus abiraterone, but if this patient does progress or doesn’t respond well in 6 months, you need to know what he’s bearing. Because you’re going a different way than the traditional one. We need to learn to do that.

Johann de Bono, PhD, MB, ChB: I would like to raise an important point from the presentation here at ESMO, and we’ve presented similar data at AACR, that these patients with these DNA repair defects, as Bertrand alluded to, do respond very well still to hormonal therapy and docetaxel. Although they may respond for shorter durations. Some respond for a long time, but I think it’s important that these patients get the standard therapies so that they can benefit from them. I do think that particularly patients with BRCA2 and, in my experience, patients with ATM have a much worse prognosis. And, actually, data from David Olmos’s group here in Madrid, that were presented at this meeting, have confirmed that if you’ve got a BRCA2 mutation, your outcome is much worse. I think that probably is due to not just the BRCA2, but also that when these cancers have a germline BRCA2 mutation, they get a second hit and then lose the other allele and lose BRCA2 and RB together. It’s a monoallelic loss of RB, but we know that RB is haploinsufficient. So, if you just lose 1 RB allele, you really made that cancer a lot worse.

Transcript Edited for Clarity