Petros Grivas: Hello, I'm Dr. Petros Grivas. I'm a medical oncologist at the University of Washington and Fred Hanson Cancer Center. I'm so excited to be here in person at ASCO 2022, with safety measures of course, and I'm so excited here today to host Dr. Andrea Apolo, who's a [inaudible 00:00:16] senior investigator at the National Cancer Institute, and really, really a legend in the field of GU oncology. Andrea, thank you for being here.

Andrea Apolo: Petros, thank you for having me. I'm thrilled to be here.

Petros Grivas: It's great to be in person again, right, after three years?

Andrea Apolo: It is very nice.

Petros Grivas: I was so happy to see this amazing work that you did, and you had this session on Friday at 1:00 PM, the inaugural session of ASCO, discussing clinical trial endpoints in the adjuvant setting for RCC kidney cancer and urothelial cancer. Can you tell us more about the session and how it went?

Andrea Apolo: Yeah, so it was a great session because the audience got to ask questions to the FDA. We got to review clinical trials that have been done, phase three clinical trials in the adjuvant setting for both high-risk kidney and bladder cancer. And we got to review FDA approvals. We looked at what are the trials that are ongoing right now in this setting. And it was a great discussion. I think it's such an important area right now, it's a very timely area because a lot of the agents that are approved in the metastatic setting, that are effective, of course, are being brought into earlier states of disease, the muscle invasive setting. And rightly so, because we want to be able to cure more patients and not wait till they develop metastatic disease.

But one of the points that we made there is not only looking at the trial design and the efficacy, but also looking at treatment burden, toxicity, cost, and really is there a survival benefit? I think that that's really important because if you can give the therapy later on, and it's the same survival as if you gave it early, we are overtreating a lot of patients early. So, although a lot of these trials that we discussed really have disease-free survival as their endpoint, we need to look at the overall survival outcomes that are presented as the data matures, because that's really going to impact the patients. And that I feel like this session was very patient-centric, so we can really... We had a patient advocate on, we had the FDA so we can really talk about what are the important issues for patients within these clinical trials.

Petros Grivas: Absolutely. And you had all these voices, the FDA regulatory, it was yourself, Dr. Pavlos Msaouel, who's discussed about the statistical considerations, and Deb Maskins, the patient advocate there, and very into the seminar discussion and the audience was very engaged.

Andrea Apolo: And we have all these approvals now, so we have an approval — Pembrolizumab was approved for the adjuvant setting in kidney cancer, high-risk kidney cancer. I would say it's a little bit different than the other approval because in this study they actually also used M1 patients. So, it's not just high-risk, renal cell carcinoma patients, they were also M1 patients. So that's a little bit different in terms of who are the patients that are enrolling are. And now for bladder, we have nivolumab approved. And I guess sunitinib is also approved but really not utilized in the adjuvant setting.

Petros Grivas: And one of the questions was about the endpoints that you mentioned. This is for overall survival, and of course, we see pros and cons, ideally and philosophically overall survival would be ideal, but it sounds like the FDA is accepting DFS, disease-free survival, as an approvable endpoint for now. It's an endpoint you can achieve sooner, or earlier, and the patients appear to agree with that. Any thoughts on that part?

Andrea Apolo: Well, I think it makes the drug accessible quicker because the trials are smaller, the follow-up is shorter. So if there's an effective therapy, and generally these are agents that are effective in the metastatic setting, and if they're effective in the metastatic, they're being tested in the adjuvant setting. So, we're hoping that if there's a disease-free survival benefit, that there will eventually be an overall survival benefit. It depends how much the disease-free survival benefit that we see. But it doesn't always come out. I mean this is exactly what happened with the sunitinib, and there was a disease-free survival benefit with adjuvant sunitinib in renal cell carcinoma, but there wasn't an overall survival benefit. So that was disappointing. So, it doesn't always correlate, but I think if there is no overall survival benefit, should we really be using that agent? And I think that's an important discussion to have with patients.

Petros Grivas: Very insightful points, Andrea. This is a very important point and at the end of the day, we want to help the patient, and overall survival ideally will be positive in those trials. At the middle of this great conversation, we have Dr. Pavlos Msaouel, assistant professor at MD Anderson Comprehensive Cancer Center who just join us. Pavlos, welcome.

Pavlos Msaouel: Nice to meet you guys here.

Petros Grivas: We're discussing here with Dr. Apolo about ... This is a very important session on endpoints in adjuvant clinical trials in kidney cancer, RCC, and urothelial carcinoma, and we're discussing pros and cons of overall survival and ideal endpoint disease-free survival. What's your take on that discussion?

Pavlos Msaouel: Yeah, so the discussion was fantastic and very engaging, and each one of these endpoints, OS and DFS, have their pros and cons. So when you ask a patient what are they interested in with regards to survival, they want to know, "Okay, if I take that adjuvant therapy as opposed to surveillance, am I going to live longer?" That's what they care about with regards to overall survival. Now, it used to be that we were much more capable of answering that question using the hazard ratio estimate for overall survival from randomized trials of adjuvant therapies. But one of the things that is gradually changing now, and it's a good problem to have, is that because we have subsequent therapies confounding that estimate, it is becoming more and more difficult to answer reliably that question for patients. So we need to actually develop new methodology to more properly estimate OS outcomes such RCTs.

Petros Grivas: And that's a difficult part. How do you make these assessments in a quantitative way, and how do you break it down for the patient, for them to make an informed decision? How do you incorporate patient values in this discussion? And we had a great patient advocate in that session talking about the patient voice, and any comments, Andrea, about how to incorporate patient voice in trial designs, since you have led many workshops in that regard?

Andrea Apolo: Yeah, I think it's important to include patients in these workshops, to include them in cooperative groups, study designs. And it's very nice that we're at the alliance and we always have a patient advocate there as we discuss a trial. We really want to hear from the patient advocate, "Would this be something you would be interested in? What do you think about the design?" Patients, there's a lot of things that are very important to the patients. They want decreased visits, they don't want placebos. I mean coming in for saline, especially during the COVID times right now, but even before that, it's a big burden for them. And there's a lot of patient specific issues that come up in trial design that it's really important to capture. And speaking of patient voices, I think assessing quality of life is something that's really important within these trials that we are still learning and developing the right tools, especially in the adjuvant setting. We just don't have great tools to assess quality of life in the adjuvant setting.

Petros Grivas: That's a very important point, and the instrument we have, as you mentioned, are mostly used in the metastatic setting when disease is present symptomatically on scans. We don't have that validated tool in the adjuvant setting. We have to work on that. And Pavlos, do you think that we can potentially incorporate those instruments upon validation as co-primary endpoints in the adjuvant setting? Do you think that's physical strategy?

Pavlos Msaouel: It is. Not only that, but there is methodology whereby we can, let's say, estimate the joint probability of both having an efficacy, hard efficacy outcome, like let's say recurrence versus not, and at toxicity, grade three or more toxicity, and improvement in quality of life. So, we actually do have already the methodology that can estimate the joint probabilities of all of this, and then depending on the trade-offs that the patient gives us on risk and benefit, how much toxicity are they willing to accept? And what is the efficacy endpoint that they're interested in? And what is the quality of life instrument and aspect that they're most interested in not impacting? And then we can come up with decisions for each particular patient even.

Petros Grivas: That's so exciting. Do you foresee the future we can use maybe artificial intelligence tools, machine learning, helping us in that discussion with a patient? Is that something realistic?

Pavlos Msaouel: Absolutely. That can help, first of all, with software that we can use to both elicit this utility functions that allows us to estimate those trade-offs for patients. In addition, we can use that software to make the estimates for the clinicians and the patient. And at the same time we can go even further to other fields like psychology where they have done amazing work on decision making. Graphic design as well, decision theory, overall operations research, all of these fields have a lot of things to teach us. And economics as well, to teach us on how to make patient specific decisions in a more coherent and reliable way.

Petros Grivas: Sounds fantastic for the future to be able to incorporate all these different aspects in a decision-making point of care. That will be the ideal scenario. Switching gears a little bit to the neoadjuvant setting, and Andrea, you have done a lot of work in the field and have led workshops with the FDA. Any comments about the ideal endpoints in the neoadjuvant setting?

Andrea Apolo: Yeah, I think that's a hard question. And we don't have good endpoints right now. We use pathologic complete response as an endpoint. It's a quick endpoint, but we need better outcomes measures to really assess how the patient is going to do. One thing I mentioned during the session is that now a lot of trials, a lot of phase three trials are being designed with the sandwich approach, with the neoadjuvant, followed by adjuvant therapy. And I think it's great to bring these therapies forward, but there should be a step wise approach. If the patient is not responding to this therapy in the neoadjuvant setting, should we really continue it in the adjuvant setting? I would say no. And we may be treating over patients. What have they responded great and they have pathologic complete response. Do those patients need additional adjuvant therapy? And then for how long?

Andrea Apolo: So, there's a lot of unanswered questions, but there's a lot of trials ongoing right now with this sandwich approach, with the neoadjuvant and the adjuvant. And another thing too that we talked about during the panel is back in 2017 the FDA held a workshop in collaboration with the NCI and SUO, and one of the goals was to harmonize the eligibility criteria and the radiologic assessment because we compare across trials all the time. And are we really comparing the same patients? Were there any variant histologies? Did they receive prior neoadjuvant chemotherapy? It was a primary versus upper track for urothelial carcinoma. What kind of surgery did they have? Did they have a partial cystectomy? Did they have a radical cystectomy? What about the lymph node dissection? There's all these different issues in terms of eligibility that we don't really think about when we're comparing across trials.

Andrea Apolo: And the second thing was the radiologic assessment. So what is an NED scan to enroll into a trial if somebody has a little blip in their lung? Are they NED, eligible to enroll, because you're supposed to have an NED scan with no evidence of disease. And what is progressive disease, radiologic progressive disease? So subjective and I didn't realize how subjective it was until we had this workshop. And the workshop was great because it was urologists, medical oncologists, radiologists, patient advocates, and the FDA. So we really got to discuss these and come up with a conclusion, and these are published and available in the FDA workshop and the FDA website.

Petros Grivas: And I highly encourage the audience to read that report because it's so much collective wisdom, I call it an experience, how to develop principles and guidelines, how to define eligibility criteria for trials, how to define progression.

Andrea Apolo: And we had one for neoadjuvant therapy also in terms of endpoints.

Petros Grivas: Which is fantastic. Both adjuvant and neoadjuvant. Any comments, Pavlos, about the neoadjuvant trial approach?

Pavlos Msaouel: I would say I completely agree on the importance on having a tightly controlled trial design. So such workshops are key because that is what will give internal validity to the experiment. It's kind of like the lab experiment. We want to make sure that we do tightly controlled experiments that have high internal validity that will give us knowledge that we can then transport to the clinic. So that's the big starting point. And that includes what are the endpoints that we're measuring and how can we make them exactly as objective as possible. There will always be some subjectivity, but such workshops and their voice actually gives us a little bit of more understanding how much it is and how to mitigate it as much as possible.

Petros Grivas: Absolutely. Totally agree with you. And one of the last questions I have for you is the AMBASSADOR trial. We have one negative trial, Andrea, the adjuvant atezolizumab versus observation, IMvigor010, with some interesting CD-DNA data that were published in nature. We have a positive trial based on DFS endpoint. We don't have OS data with adjuvant nivo versus placebo, CheckMate 274. So maybe the tiebreaker here, maybe the adjuvant AMBASSADOR which is close to accrual, 702 patients—

Andrea Apolo: We completed accrual, so we actually stopped accrual once nivolumab got FDA-approved. So thank you for asking about the AMBASSADOR trial. The AMBASSADOR study is a randomized phase three trial of adjuvant pembrolizumab versus observation in high-risk bladder cancer patients, post-radical surgery, and it has a co-primary endpoint of disease-free survival and overall survival. And we stopped accrual, 95% accrued and because nivolumab became a standard of care and we didn't think it was appropriate to randomize patients to the observation arm at this point. So anyway, the trial is maturing and we're awaiting the data. So now we have a trial where we're awaiting data, a negative trial, and a positive trial. So, it's a little complicated in bladder cancer, so we'll see what the outcomes show.

Petros Grivas: Absolutely. We look forward for the results of AMBASSADOR, and, as you mentioned before, the plethora of phase three trials in the sandwich approach, the neoadjuvant/adjuvant. So the future is very interesting and the field is moving forward. I would like to thank both of you for your time here today for your overall contributions in the field, GU oncology, and I'm hopeful that in the future we're going discuss more about the advancement in the field. Thank you so much, Andrea and Pavlos.

Andrea Apolo: Thank you for having us.

Pavlos Msaouel: Thank you for having us.