Read the Full Video Transcript

Zachary Klaassen: Hi, my name is Zachary Klaassen, urologic oncologist at the Georgia Cancer Center in Augusta, Georgia. I'm delighted to be joined on UroToday for an ESMO 2024 discussion with Dr. Deepak Kilari, who is a medical oncologist at the Medical College of Wisconsin. Today, we're going to be discussing the ORACLE study, which is outcomes with novel combinations in non-clear cell renal cell carcinoma. Deepak, thanks, as always, for joining us on UroToday.

Deepak Kilari: Zach, thank you so much for having us. Zach, thanks so much for having me. And I'm really excited to share the results of the ORACLE study that were presented at ESMO 2024. So the ORACLE study stands for Outcomes with Novel Combinations in Non-clear Cell RCC.

And as you can see from the authors list here, this was a multisite, retrospective study involving academic centers all the way from the West Coast to the East Coast. We had centers like UCSD on the West Coast, and you can even see on the East Coast, as well as multiple centers in the Midwest. So again, it is capturing data from major academic sites across the country. And we’re excited—I’m excited to present the results of the study today.

So as we all know, non-clear cell RCC is a very heterogeneous disorder in the sense that it comprises 25% of all renal cell cancer diagnoses, but it’s not just one histology. There are multiple histologies, and each histology behaves differently. While we have seen multiple advances in the treatment of clear cell, in terms of treatments for patients with clear cell RCC, these have really not translated directly for the management of patients with non-clear cell RCC, mainly because there are fewer patients within each histology in the non-clear cell cohort.

So clearly, there's a paucity of data to guide management of non-clear cell RCC due to the rarity and the heterogeneity of these tumors. And we know that non-clear cell RCC patients have limited treatment options and also worse outcomes, for the most part, compared to patients with clear cell RCC.

So what we did was—this project first started off when I had a patient with chromophobe RCC. And my go-to typically had been everolimus for a long time. And then when I saw the data for len/everolimus, I was really impressed at how len/everolimus did for patients with chromophobe RCC. But when you look at the numbers in the Phase II study that was presented a couple of years ago, there were very few patients.

So initially, I wanted to start off by looking at real-world outcomes with lenvatinib and everolimus for chromophobe. But then as we started this project, a lot of co-investigators on the study said, why don't we look at different combinations? And why should we limit ourselves just to chromophobe? Let’s look at other non-clear cell histologies. So what initially was supposed to be one combination and one subset of non-clear cell RCC became a pretty big cohort looking at different combinations in different non-clear cell histologies, and not just the first-line setting, but also second-line, third-line settings.

So the eligibility for patients to be in the ORACLE study was that they had to be above the age of 18, they had to have a non-clear cell histology diagnosis of locally advanced or stage IV RCC, and they had to have received a combination treatment which included IO+IO (that’s ipi-nivo), IO+VEGF therapy, or VEGF+mTOR therapy, or any other combination as well. And they had to have at least one dose of combination to be included in the study. And the combination could have been received in a front-line or a later-line setting.

So this was, as I said previously, a multicenter retrospective analysis evaluating real-world outcomes. And we do have pretty good data evaluating these combinations in a first-line setting, but not really in a second-line or a third-line setting. And also, we don’t have good real-world outcomes. So this study was capturing real-world outcomes as well. We were capturing demographic data, clinical data, treatment outcomes with the combinations.

And the primary endpoint of the study was objective response rate, and this was by investigator review using the RECIST 1.1 principles. The secondary endpoints included time to treatment failure, duration of disease control, clinical benefit rate, and overall survival. And outcomes are reported based on when they received the first combination therapy. Kaplan-Meier survival estimates were used to assess time-to-event endpoints.

So these are the baseline characteristics. We have 350 patients in the database, but for the purpose of ESMO, we had some missing variables, so we included only 253 patients. The median age in this cohort was 59 years old. Thirty percent of patients were female. What’s noteworthy was that up to 25% of patients actually were African American, and this was self-identified.

And the majority of the patients had good performance status. Fifty percent of patients had de novo stage IV non-clear cell RCC. And if you look at nephrectomy status, approximately 70% of patients either had a partial or a total nephrectomy before going on study or before getting combination treatment.

And then if you look at histologies, we have different histologies reported here. And again, it’s noteworthy to mention that these are all academic sites that have GU pathologists on board, so I would trust these pathologists to make the accurate diagnosis. So 15% of patients had chromophobe, 3% had collecting duct. The majority of patients in this study had papillary—that was 38%. And the second most common group was unclassified at 35%. And we had a few patients with translocation RCC as well.

And if you look at the number of metastatic lesions, four-plus—50% of patients did have more than four-plus lesions. And in this cohort, again, the majority of patients had intermediate- or poor-risk disease. Almost 80% of patients had intermediate- or poor-risk disease.

And if you look at whether any combination was received in a first-line or later-line setting, up to 70% of patients had the combination therapy in a front-line or a first-line setting, but we had 20% of patients that got treatment in a second-line setting, and then we had 12% that received combination therapy in a second-line or a later-line setting.

And again, there were patients that got different treatments before getting on to the combination. So we had approximately 8% that got IO and VEGF therapy separately before getting the combination, and we had, as I said previously, a majority not getting any treatment before getting on to the combination treatment.

And in terms of the combination treatment, as you can see here at the bottom, 50% of patients had IO+IO—that’s PD-1 plus CTLA-4 inhibitor—as their combination treatment. But then we had—the next most common combination was PD-1 or PD-L1 treatment in combination with VEGF inhibitors, at 40% approximately. And then we had a few patients with VEGF+EGFR, and we had VEGF+mTOR at 12% as well. And we also captured if they had sarcomatoid or rhabdoid component, as you can see on the right-hand side of the table.

So these are the primary outcomes for the study. And as you can see here, what we did was, to summarize things and to put things into context, we looked at different combinations, right? We looked at if they got IO+IO, if they got IO+VEGF, or if they got VEGF+mTOR. So we divided them up based on what treatment they got.

And then we also divided patients up by what subtype of non-clear cell RCC they had. And if you look at the response types, we have overall response, and then we have the clinical benefit rate, which is basically CR plus PR plus stable disease.

So what captures my eye when I first look at this data is there’s a differential response to treatment, right? So for instance, if you look at the first column here, the IO+IO combination, we had 124 patients in the IO+IO combination. And if you look at which patients had the best outcome, translocation RCC—that’s the first one here—had a 50% clinical benefit rate. And then chromophobe, interestingly, also had a 50% clinical benefit rate. And then if you look at unclassified, we had 41% that actually had a clinical benefit rate.

So if you move on to the next one, that’s the IO+VEGF combination. If you look at translocation, we actually saw that 75% had response or clinical benefit to IO and VEGF combination. And the next thing that stands out to me here is that papillary patients had a 69% clinical benefit rate. And then if you move on to VEGF plus mTOR inhibitors, the maximum benefit was seen in chromophobe at 77%.

So to me, this suggests that, yes, these combinations work for most patients, and there’s some type of response. But if I see a chromophobe patient, it’s hypothesis-generating that a VEGF+mTOR combination might actually benefit the most. If I see a patient with papillary RCC, I would probably think about an IO+VEGF combination because the clinical benefit seems to be high at 70%. And if I see a patient with translocation or unclassified, I would probably think about either IO+VEGF or IO+IO.

And again, these are just hypothesis-generating. This is not prospective randomized data. But again, when we have a small number of patients, we want to make sure that we can generate data that helps patients and providers in clinic.

So the next thing that I want to highlight is that irrespective of whether you have a sarcomatoid or rhabdoid component, we really did see a benefit with all of these combinations. And the other very interesting point that we noted was that if you get these combinations in a first line, the response seems to be the best. But if you do get these combinations as a second-line therapy—right? First therapy, yes or no—so if you get these combinations in a second-line setting, there still seems to be a benefit. Not as robust as what we saw with the first-line setting, but again, we did see benefit.

And then if you look at it here, if your risk category is high—I mean, if you have good risk, you have the best benefit, but if you have intermediate and poor, you still see a benefit, but the benefit is not as good as you typically see with a good risk patient. And then irrespective of whether you have bone metastases or not, you still see responses with these combinations.

So then what we tried to do is—well, let’s see that people have an early response. But what about time to treatment failure? Again, these are hypothesis-generating. I mean, they don’t have p-values to justify saying that this is what we should be doing.

But again, if you look at the first curve here, chromophobe—and if you look at chromophobe, I think that VEGF+mTOR—actually, if you look at the bottom here, VEGF+mTOR is this one here. A PD-(L)1 checkpoint inhibitor plus a VEGF inhibitor is this color here. And then IO+IO is this straight line here.

And if you look at the chromophobe, time to treatment failure seems to be better with VEGF+mTOR. And if you look at papillary, VEGF+IO, or PD-1 or PD-L1 plus VEGF, actually seems to be doing better. And then unclassified—I think the numbers here seem to be too small for the green one to say anything, but if you look at these two here, I think unclassified seems to be doing slightly better with the IO+IO combination.

What about survival? I don’t think we can really make any conclusions based on these curves because they seem to be overlapping. But again, these are small numbers, and dividing these small numbers up is always difficult to find true answers here.

So to summarize, what did we learn from the ORACLE study? We did see that there was differential antitumor activity observed with combination therapies in non-clear cell RCC. We did learn that activity was seen with different subtypes with these combinations, not only in the first line, but also in the later line of treatment. And also, the response rates and survival with combination therapies in this multicenter data set were inferior compared to those seen with clear cell RCC.

And we clearly need prospective studies in this setting. But again, with all the challenges that we’ve seen with the other prospective studies in terms of how long it takes to accrue, I think we need to capture more real-world outcomes to help us better manage our patients. So with that, Zach, I would like to pass it back to you. And again, thanks for giving us the opportunity to discuss our study.

Zachary Klaassen: Deepak, wonderful presentation, really going through some of the highlights. And that one figure—a lot to unpack there in terms of the differential effects. When I was looking at your study, several things popped out. I think the one that I want to just dig into a little bit was you mentioned about one quarter of patients were Black or African American.

And we know historically these have been underrepresented populations in trials. Not only that, but with a difficult non-clear cell RCC population, you take that plus the disparity in clinical trial enrollment. Is there thought about looking deeper into this 24% of patients and seeing what their outcomes are? I know we’re getting into small numbers here. But again, this is an important number that you guys have in your study, about one quarter Black or African American.

Deepak Kilari: Zach, that’s a great question. That is our next planned subset analysis. And as you rightly pointed out, we are worried that we’re probably slicing and dicing the data, and it might not be that meaningful. But again, you’re right. Our African American representation in clinical trials is less than 10%. And on top of that, if you have non-clear cell, it’s probably much smaller than that from the trials that we have already reported. So yes, we do plan to look at it. And hopefully, we are planning to submit an abstract to ASCO. So be tuned for that.

Zachary Klaassen: Excellent. We’ll look for that for sure. I think when you look at real-world outcome data, like you guys have, from some very high-powered centers, a good number of 253 patients, specifically for, let’s say, papillary, maybe even chromophobe—more of the common non-clear cell histologies—how may this data that you guys have inform future clinical trial design in this space?

Deepak Kilari: That’s a great question, Zach, in the sense that, I mean, these are all hypothesis-generating. And we ask ourselves, based on these data, do we plan further prospective studies? So I do think that’s what we should be looking at. I think we need to expand upon our database before we can draw definitive conclusions and say, hey, we’ve seen a better—I mean, yes, we’ve seen a better response, but do we plan a clinical trial based on that? I think that’s yet to be seen.

We are planning multiple other substudies with ORACLE. So Dr. Brugarolas and Dr. Payal Kapur from UT Southwestern are interested in looking at histologies and actually correlating that with what has been reported. So that is what we’re planning as one substudy. We’re planning on looking at genomic predictors of response. We’re planning on looking at African Americans and outcomes.

But we’re also planning on looking at what happens after you get these combinations, right? I mean, everybody is so focused on first-line setting, but what happens in a second-line setting? What happens in a third-line setting? It’s great to have a combination that works for first line, but then is there a better second-line option? Is there a better third-line option for these subgroups? That’s what we’re also trying to capture and hopefully discuss at the next couple of meetings.

Zachary Klaassen: I think that’s important because as difficult as it is to recruit to first line, it’s going to be more difficult for second or third line. So this is where the real-world outcomes really become important for those patients.

Deepak Kilari: I totally agree, Zach.

Zachary Klaassen: Deepak, always great having you on UroToday. Maybe just a couple of quick take-home points for our listeners.

Deepak Kilari: Thanks, Zach, again. So I would say that non-clear cell RCC is a very heterogeneous disease with a paucity of data. So I think we should be looking at developing more prospective trials to understand this disease better.

We did observe differential activity of these combinations in each subset of non-clear cell RCC. Based on our data, I would suggest that probably a VEGF+mTOR inhibitor would work better for chromophobe RCC. For patients with papillary RCC, I would probably recommend an IO+VEGF combination. And for patients with unclassified, I would probably recommend IO+IO or IO+VEGF combination.

Regarding collecting duct and translocation, the numbers are small. We need more numbers to say that this might be better than that regimen. But I think we need to expand on this data, and we need to learn more. And we have many other subset studies planned from this data set.

Zachary Klaassen: Very exciting. Great summary. And again, Deepak, thank you for your time on UroToday.

Deepak Kilari: Thanks, Zach, for having me again.