Read the Full Video Transcript

Alicia Morgans: Hi, my name is Alicia Morgans, and I'm a GU Medical Oncologist at Dana-Farber Cancer Institute in Boston, in the US. I'm so excited to have here with me today, Dr. Simon Crabb, who is an Associate Professor in Medical Oncology at the University of Southampton and the Associate Clinical Director of the Southampton Clinical Trials Unit in Southampton, UK. Thank you so much for being here with me today, Dr. Crabb.

Simon Crabb: Very much. Thanks for having me.

Alicia Morgans: Well, thank you so much. And I'm really excited to talk with you about the analysis that you presented regarding the rucaparib arm in the ATLANTIS trial.

Simon Crabb: Sure. So, ATLANTIS is actually a platform study running in the UK. It's an investigator-led study, and it has multiple parallel, placebo-controlled, phase two studies that run in the maintenance setting after first-line chemotherapy for advanced urothelial carcinoma, and the selection for which of these studies you go into is determined by biomarker analyses that are done during the course of the chemotherapy on archival tumor samples. And the study we presented at this meeting was a rucaparib comparison, so rucaparib versus placebo in patients that had a biomarker for DNA repair efficiency. We did this study, of note, before the data for avelumab maintenance treatment in this disease. So, we felt that placebo was a reasonable control arm for this study. So we took patients that had this biomarker, and we randomized them, if they were eligible, to treatment within 10 weeks of completion of chemotherapy, assuming that they did not have progressive disease at the end of the chemotherapy.

And the biomarker was based on three possible reasons for being positive. So, firstly, high genome-wide loss of heterozygosity, and we used that because there is data for rucaparib in ovarian cancer that suggests that that will act as a predicted biomarker for a benefit or the alteration in any of a list of 15 DNA repair genes, or finally if patients were already known to have a germline mutation of BRCA1 or BRCA2. So, any of those being positive, made you biomarker positive ... and it turned out, in the patients that we prescreened, that was getting over 30% of the patients that had one of those present. Just over half, it was due to high LOH in terms of why they were biomarker positive.

Alicia Morgans: So that's actually such important information on its own, knowing the prevalence of these alterations and, of course, understanding the difference between which are included because of the high LOH versus the germline mutations and otherwise. So really, really interesting data. What did you find in terms of your primary analysis and those outcomes of interest?

Simon Crabb: Sure. So, the primary endpoint was progression-free survival. We found a difference in the primary outcomes. So, progression-free survival in patients receiving rucaparib was 35.3 weeks, and, in patients receiving placebo, 15.1 weeks. So the hazard ratio was 0.53. The P-value was 0.07. This was an efficacy signal-seeking study, so we set our target for 20%, so slightly higher than perhaps what people are used to.

So, by the criteria that we set, this was a positive trial, and what we also found when we started to break the biomarker down, we think actually probably most of the benefit in this trial comes from people that actually had a gene alteration in their cancer. So, there are data that are emerging and being put into the manuscript and suggests maybe that is the key indicator for benefit from this approach as opposed to high LOH on its own.

Alicia Morgans: So that's so important, of course, because if we're really including those high LOH patients as well, that can dilute the signal and then lead to that borderline significance. So what do you do with that information? How do you think about that finding in terms of future clinical trial design, for example?

Simon Crabb: So, I think it's critically important.  When we started this trial, we felt that it was probably necessary to have biomarker selection. It certainly has been the case in most of the solid tumors. Although, in this meeting, we now have data that suggests, actually, perhaps that's not true in prostate cancer when you're combining with a hormonal agent. So I don't think it's an absolute rule, but we felt that the biomarker selection was probably the right way to go. Of interest, at this meeting, there was also the data from the trial, BAYOU, which was looking in a different setting for urothelial carcinoma, but for patients who were platinum ineligible.

They tested combining olaparib with durvalumab immunotherapy, and their trial was negative overall in all-comers without biomarker selection. But in about 20% of patients in that study, there was an alteration in a list of DNA repair genes. And, in that subset, they saw a benefit in progression-free survival, also. So I think these two studies together both now indicate that it may be possible to find a subset of patients with this disease where, in both studies, there appears to be a benefit in progression-free survival. To come back to your question, yes, I think future studies are going to have to look and consider very carefully how they are doing biomarker selection.

You have to make a choice at the beginning of a study like this, what you are going to call "biomarker positive," but I think we are going to need further work. It's probably not going to be as simple, certainly, as BRCA1 and BRCA2 genes. So we are going to need a wider panel, I think, but more work is going to be needed in this area, definitely.

Alicia Morgans: Well, that makes a lot of sense. And, you know, the other thing that makes sense, as compared to the other studies in this particular arena, looking at PARP inhibition ... As you mentioned, the BAYOU study did seem to have a benefit in that subset that was selected. And, in both of these trials, these are not patients who have had platinum failure. So these are patients who may still be sensitive to platinum or who have not been exposed yet to platinum. How do you think that actually comes into play when we're thinking about PARP inhibition, which, at least in the prostate cancer world, we think about mechanistically being similar in terms of the way that these drugs attack the cellular machinery?

Simon Crabb: I think so. I mean, certainly in our trial, of course, there were actually two points of selection. The first, chemotherapy. They then had a biomarker applied to that. So we are talking about a fairly rarefied group of patients that have these dual reasons to think that they might potentially benefit from PARP inhibition. And, of the people that we randomized in our study, 90% had had an objective response to their prior chemotherapy, which is ... excepting cross drug comparisons, that's higher than you would expect normally for this group of patients.

So I think we were enriching for a group that you would hypothesize would benefit from PARP inhibition. And I think, on the data that we have so far, that seems like the right way forward for future studies. I think I wouldn't personally want to take forward a study that looked at an unselected population. I think we need to think about this as a subset of this disease.

Alicia Morgans: That makes complete sense. And, as we sort of wind down, I just want to make sure, were there any new safety signals or any areas of concern that were identified with this patient population, with the use of rucaparib?

Simon Crabb: No, there weren't, so the safety profile was basically consistent with our experience of this agent in other cancers. We saw slightly high rates of fatigue, rash, nausea, and raised ALT levels in people who got rucaparib. But this was all generally of grade one or two severity. We didn't see any grade four and certainly no grade five toxicities. So we felt, overall, this was a tolerable treatment, and this is, of course, patients who have just finished chemotherapy. So I think it's an important point that, if you're going to continue with ongoing treatment for potentially a prolonged period of time, it does need to be tolerable. And so we were comfortable that was the case.

Alicia Morgans: Absolutely. Well, great. Well, as you think about the final message to the listeners on this particular arm of the study, what would that message be?

Simon Crabb: So, the study was positive. It showed a progression-free survival advantage for patients given rucaparib in a biomarker-selective group of patients. It was a small study, it's not practice-changing. What we do believe is that it now justifies further development of this approach. So we think a phase three study ... again, in a biomarker-selective population ... would now be appropriate.

Alicia Morgans: Well, I sincerely look forward to hearing more and look forward, of course, to talking with you about other data from the study, as those other arms readout. Thank you so much for your time and your expertise, and best of luck to you and the investigators as you carry this idea forward.

Simon Crabb: Thank you very much.