Alicia Morgans: Hi, my name is Alicia Morgans, and I'm a GU medical oncologist at Dana-Farber Cancer Institute. I'm so excited to have here with me today, a friend and colleague, Dr. Jonathan Rosenberg, who is the chief of the GU medical oncology service, and a professor in medicine, at Memorial Sloan Kettering Cancer Center in New York City. Thank you so much for being here with me today, Jonathan.

Jonathan Rosenberg: My absolute pleasure.

Alicia Morgans: It is my pleasure too. And I also am so excited to talk with you about a presentation you gave at GU ASCO 2022. And I think there were definitely, lessons learned from this study that you reported out. And of course, congratulations for getting the work done. That is always step number one, and sometimes the hardest, but you presented data from the BAYOU trial. Can you tell us a little bit about it?

Jonathan Rosenberg: The BAYOU trial was a randomized Phase II placebo control trial of olaparib and durvalumab, or placebo and durvalumab, in patients with metastatic urothelial cancer who are ineligible for platinum-based chemotherapy. That's a very difficult patient population to deal with. The treatments that we have are immunotherapy only, essentially, and still, the outcomes are suboptimal in that patient population, as the treatments don't work for most patients.

And so we hypothesized, based on preclinical data, that there might be synergy between PARP inhibition and immunotherapy, in inpatients with advanced urothelial cancer. And so, this study was designed to look at the PARP inhibitor olaparib in combination with durvalumab in that patient population who were platinum-ineligible.

And so, the study randomized patients, one-to-one, to either durvalumab and olaparib, or durvalumab and a matched placebo, in the patient population where people who were ineligible for platinum, not just cisplatin, but also carboplatin, as well as having other more traditional cisplatin ineligibility criteria, in terms of heart failure, neuropathy, hearing loss, or performance status, et cetera.

The trial was powered to look at progression-free survival in the intent to treat population, with a pre-specified analysis in patients with HRR mutations, homologous or combination of repair mutations, using the panel that's been evaluated with olaparib in other diseases. And we know, that identifies patients that might have enhanced sensitivity to olaparib.

The trial was conducted, and the results, unfortunately, showed that in the intent to treat population, there was no difference in progression-free survival for patients treated with the combination, compared to just durvalumab. In the HRR mutant patient population, however, there was a fairly substantial difference in the progression-free survival, from about 1.8 months to 5.8 months, with a hazard ratio of 0.18, P value that was highly significant, but the downside is that it's a relatively small subset of patients in the trial. And I think really, is hypothesis generating, and allows us to think about future directions.

The side effect profile didn't seem to be different than what you might expect with either drug by itself. So additive toxicities, rather than synergistic toxicities in the combination, more asthenia GI toxicities and cytopenias in the olaparib arm, and no increase, that we could tell, in immune related adverse events. So well tolerated, or as well tolerated as either of those treatments are in patients.

What the data, I think, suggests that, as we think about future steps, we really need to think about the right patient populations for PARP inhibitors in urothelial cancer, and that unselected populations may not be an appropriate group.

Alicia Morgans: So I think that's such a big learning point, especially in urothelial carcinoma, where we, as a community, have not necessarily really thought for as long, I would say, as some of the other communities, ovarian cancer, prostate cancer, even. Certainly, it has come to light, that these are mutations that are, of course, happening in these patients, but the history's a little less, it's newer. And so, as I think about that, I guess my big questions for you are, in these patients in the study, were these patients who had germline DNA repair defect mutations? Did they have somatic DNA repair defect mutations? Obviously, small group, you couldn't necessarily look at differences between the two, but that's something that seems to be not necessarily, a driver of response in prostate, for example, but perhaps it will be a driver in bladder. And so I'm just curious what your thoughts are there.

Jonathan Rosenberg: It's a very, very good question. And the assay we used was somatic only, both foundation medicines. So we don't have true breakdown of germline versus somatic in that trial. Some people argue that it won't matter. Other people argue that you really need to know the contribution of somatic versus germline. I'm more of the opinion, for this biomarker, that it may not matter quite as much as it might for some other biomarkers. I think the percentage of patients who have germline alterations in bladder cancer that are pathogenic and leading to the tumors, predisposing to the tumors, might be quite small, based on some work for Barry Taylor and others here at MSK, but that it still might play a role in sensitivity to treatment with PARP inhibitors in this disease. Even if it's not something that's acting as a tumor suppressor in bladder cancer, the way it might in breast cancer, ovarian cancer, or prostate cancer.

Alicia Morgans: Very, very interesting though, and something for us to keep digging into. And the other piece that I think is really interesting that you mentioned earlier, is that there may be a difference in response between patients who have been exposed to and have now resistant disease to platinum agents, versus patients who have not seen platinum previously. And what are your thoughts there?

Jonathan Rosenberg: I think it's a reasonably likely hypothesis, based on the data we've seen to date with other trials. We've seen data at this meeting with niraparib in an unselected population, and the ATLAS trial with rucaparib in a refractory population, and in the BISCAY trial, which used the same durva/olaparib in a refractory population. And then also, Simon Crabb presented data, in a maintenance setting, of platinum sensitive patients, also showing a progression-free survival advantage. And so all the other studies were negative. And the only studies that seemed to have a signal, although, I wouldn't necessarily call them positive, but signals, were in those where patients were not platinum refractory. And so, all the platinum refractory trials, no signal of activity, even in the presence of HRR mutations, whereas, in the platinum naive or sensitive populations, we see a signal. I think we have to pay attention to that.

Obviously, we'd want to think about it more scientifically in the laboratory. But I think given the availability of the drugs, the ability to measure the biomarker, and our understanding of bladder cancer biology, we are in fact, poised to think about where we go with these agents in bladder cancer. And there are some opportunities, either in the maintenance setting, or in the first-line setting that we might think about.

Alicia Morgans: I think that's a great segway to the next question, which is, really, where are we going from here, and how can what we've learned from this study, and from those that you've mentioned in this part of the field, how can that help us take next steps?

Jonathan Rosenberg: I think we have to have the fortitude to recognize that we need a biomarker selected population. We've done it with FGFR3, and managed to have a drug approval over erdafitinib, in a similar sized patient population, of about 15 to 20%. And so, given the patient population, and the sample size, and the percentage of HRR mutations, it's really, no different.

The most common mutations of BRCA2 and ATM, in the BAYOU trial at least, we don't know the relative contribution of different mutations with sensitivity in bladder cancer. And I don't think we totally have a handle on it in all the other cancers as well. And so we're going to need to, I think, cast a focused net, not too wide, but just right. Large enough to make sure we have enough people that we're confident are going to benefit from PARP inhibition when we design the next generation of trials, but not such a broad net that we bring in people who are inherently going to be resistant. Because it's not going to help them, and it's not going to move the field forward and help patients.

Alicia Morgans: I could not have said it better myself. Those are wonderful comments. And I think, at the end of the day, that's what our goal is, really, to help patients by moving the needle step by step. I sincerely appreciate your ongoing efforts, and I congratulate you and your team, for providing us with this data. And for your excellent commentary today.

Jonathan Rosenberg: Thank you so much. Have a great day.