Charles Ryan: Hello from the PCF 25th Anniversary Retreat in Carlsbad, California. I'm joined today by my colleague and friend, Alan Bryce. Alan is the Chair of the Division of Hematology and Oncology at the Mayo Clinic in Scottsdale, Arizona, and, today, we're going to talk about PARP inhibitors, most specifically rucaparib, and, Alan, you've been involved in the TRITON2 trial, and you are presenting that data here at the PCF. Tell us a little bit about the TRITON2 study.

Alan Bryce: Yeah, so the TRITON2 study is looking at the use of rucaparib in patients with alterations and homologous repair of recombination genes in prostate cancer. We know from ovarian and breast cancer that PARP inhibitors can have a role in this select population and, in fact, these same alterations that are effective for selecting therapy in breast and ovarian cancer exist in prostate cancer patients.

Charles Ryan: This is an ongoing study. It's a Phase II, and you said that there are a number of different molecular selection criteria, so what are they?

Alan Bryce: As we look at it, about 25% of prostate cancer patients more or less will have alterations in a select group of genes including BRCA2 most commonly, BRCA1, ATM and a list of other potential DNA repair-associated genes such as such a CDK12, FANCA, BRIP1, and some others. So this study took a group of 15 of these genes as a selection criteria for treating patients with rucaparib. So as with many of these molecularly directed studies, there are really two steps. There's a screening step for testing the tissue or the patient's germline to see if a mutation exists, and, if it does, patients then would be enrolled in the study to receive treatment with rucaparib, which is an oral PARP inhibitor.

Charles Ryan: Would one hypothesize, and we've talked about this quite a bit, that the benefit or the response to a PARP inhibitor would be the same across all these different mutations, and I think that question gets me to the next question, which is why are you analyzing the data now?

Alan Bryce: Yeah. Yeah, that's right, so the answer is we don't know. As far as before this study, there were certainly hypotheses from smaller case series at least raising the possibility that any of these alterations would lead to a response. However, those of us I think who look at this data a lot and have seen these patients, we all have the sense that there is a significant difference between these alterations and the response one would expect, so the design of the study is meant to cast a wide net to capture this large group of alterations. It splits them into different cohorts so that they can be analyzed separately.

Without question, the heart of the study is patients with BRCA mutations, so BRCA2 being the most commonly mutated homologous repair-related gene in prostate cancer. That's about five to one more common than BRCA1, and then ATM also being the third most common alteration. I think those of us in the field have an expectation that BRCA is the most important gene, and, of course, this study sets out to prove that.

The reason this analysis is being done is, with this design of looking at a large panel of genes, we feel it's important to constantly assess the response rate amongst the various genes in question whether any of the genes should either remain in the study or, perhaps, be removed if we have enough data to suggest an amendment to the design.

Charles Ryan: What are you reporting today?

Alan Bryce: So we see that amongst the BRCA1 or 2 mutant population, amongst patients with measurable disease the RECIST response rate is 44% in the first 25 patients we've analyzed. And then in terms of PSA 50 response rate, so that's a reduction of PSA by 50% or more, we're reporting a 51% response rate. So really very encouraging results for an early unplanned analysis to be sure, but very encouraging results in the early going. 

Charles Ryan: What's the signal on duration of response?

Alan Bryce: If we look at the duration of response, it's actually quite encouraging. The longest patient on the study so far has been on rucaparib for 52 weeks. We see most patients exhibiting a response within the first eight weeks, and most patients who exhibit that response remain on the study now with a median followup approaching six months.

Charles Ryan: Great, and what about safety?

Alan Bryce: Safety is encouraging. The safety is exactly what we saw in the previous studies from ovarian and breast cancer. There really are no surprises or any particularities in terms of the toxicities in men or the toxicities in prostate cancer patients. Keep in mind that most PARP inhibitors have never been used in men to any significant degree because they were initially developed in breast and ovarian cancers, so, in this study, the major toxicities are what we would expect, that is, myelosuppression anemia, certainly, PARP inhibitors do have a mild suppressive activity, and also some GI toxicity, a low rate of nausea, low rate of vomiting, some dysgeusia, anorexia or whatnot, but nothing too dramatic, certainly easier than docetaxel for the most part.

Charles Ryan: Great, so, certainly, as a first pass in prostate cancer, rucaparib looks to be reasonably safe. We're definitely getting on efficacy signal. Are these data being used to change the study? You've mentioned that before, but trying to identify signals, positive or negative, how's the study going to change based on these?

Alan Bryce: Yeah, so the important point there is certainly amongst the list of other non-BRCA mutations. So in the ATM population, we have 18 patients there so far, and it is an interesting dataset in the ATM populations in that we see patients who have prolonged stable disease, although we've yet to see patients develop a RECIST response. One patient has had a PSA response, and the question will be whether that is it... will be enough to translate into a durable disease control, and we don't know yet with only 18 patients.

Looking at the longer list of genes, we've seen responses in a gene called BRIP1, another in FANCA, and we've had 13 patients accrued with CDK12 alterations. In the CDK12 population, we haven't seen any responses amongst the first 13 and, given that it represents over half of the planned cohort C, the decision has been made to amend the study and no longer enroll CDK12 patients, so that was really the driving motivation behind doing this analysis now.

Charles Ryan: Which seems very reasonable to not offer patients a therapy that doesn't look like it's going to benefit them if they've got this particular mutation, and, of course, the other interesting thing is the CDK12 story is really going in an immunotherapy direction, I think, so that's a potentially different avenue for those patients, so great.

My understanding is that rucaparib now has been granted breakthrough designation status by the FDA, and that means that there is going to be continued and, perhaps, a more accelerated review of the data as they emerge, and there are... There's the TRITON3 study that's going on that is a randomized Phase III study in the pre-chemotherapy or the chemotherapy-naïve patient, and that's the only in accruing BRCA1, BRCA2 and ATM patients at this time, and so I think these data are really going to help bolster the case for TRITON3, and ultimately both datasets, TRITON2 and TRITON3, are going to be relevant to the potential use of rucaparib as a standard approach for prostate cancer, and it's an exciting time because we're moving into the phase of personalizing therapy for prostate cancer, and your work and your explanation I think helps us move further forward on understanding who we should sequence and what we can expect to do with those results at least for now in terms of clinical trial enrollment.

Alan Bryce: That's right.

Charles Ryan: Thank you for joining us, and congratulations.

Alan Bryce: Thanks, Chuck. Thank you.