KEYNOTE-B61 Study Results and a Focus on Non-Clear Cell Kidney Cancer

A roundtable discussion, moderated by Brian Rini, MD, discussed the latest data from the American Society of Clinical Oncology Genitourinary Cancers Symposium 2024 in clear and non-clear cell RCC. Dr. Rini was joined by David McDermott, MD; Elizabeth Plimack, MD; and Martin Voss, MD.

In the next segment of the roundtable series, the panel shifts to focus on the current state of non-clear cell kidney cancer, with a focus on the KEYNOTE-B61 study results presented at ASCO GU.

Watch the next segment in this series.

—

Dr. Rini: Let’s move on to non-clear cell. Martin, we’ll start back with you. KEYNOTE-B61 was a trial of lenvatinib/pembrolizumab across non-clear cell histologies, I think the largest non-clear cell trial ever done or certainly it’s up there. Do you want to summarize those data briefly and then we can talk about our individual approaches?

Dr. Voss: Yeah, of course. This is a study that’s previously been analyzed and actually reported in print already, so it’s fully accrued and analyzed. The first planned analysis was presented previously and published in Lancet Oncology in 2023. Laurence Albiges was the first author on the publication, and this is 158 patients-strong study, single-arm, multiple institutions conducted across 14 countries and enrolled patients to be treated with standard lenvatinib/pembrolizumab at the dosing schedule that we’re all used to until disease progression or intolerance with a max of 2 years of pembrolizumab, as we’re all used to doing in clear cell RCC. Primary endpoint for the study was objective response, and this was enrolling patients with untreated non-clear cell variants. For the most part, those were papillary kidney cancer patients, about 90 patients with papillary, 30 chromophobe patients, which is the largest chromophobe group reported prospectively, unclassified patients, and a small group of translocation patients, I think it was 6.

In the original publication, response was assessed/reported, as well as secondary end points and tolerance, and it showed a robust signal across disease groups, maybe more so in some than in others. Again, the chromophobes stood out as the ones that seem to benefit least from an IO-containing combination. But the papillaries and the unclassifieds had a high response, and we confirmed this in this analysis. The original analysis had 15 months median follow-up and this one has 6 months extra, so about 23 months median follow-up, which was helpful in some ways and just reaffirmed the rest. For instance, there was no median duration of response reported in the original publication. Now with more follow-up, we can see that median duration of response is just short of 20 months, which I think is clinically meaningful. The level of efficacy is robust. Specifically in papillary, it was about a 50% response rate including CRs.

Similarly, there are response and there was also CR in translocation patients. No CR in the chromophobe group. Response rates have gone up in all the groups a little bit with more follow-up. As you would expect in the chromophobes, somewhere in the low 30s, which we can then argue is that the lenvatinib alone or is it lenvatinib for most patients and maybe less IO effect, but this additive effect that we see across all kidney cancer variants as we had seen in clear cell, maybe a little bit less clear there. But I do think it is the biggest dataset that’s out there now, and I do think it sends a very clear message, and it did show across all groups a very low PD rate also. Similar to what we’ve discussed before. I think this is a regimen that I think should be considered for most patients with non-clear cell variants.

Dr. Rini: Betsy, I see shaking your head and you said you’re a lenvatinib/pembrolizumab user for clear cell, so I assume, is that your approach in non-clear cell based on these data? Or what are the nuances to that?

Dr. Plimack: That did become my approach when I first saw these presented back at ESMO. Just because it’s such a large series, I really hone in on papillary because I think the numbers are just very small in the other cancers that I see regularly, mostly chromophobe, but to get that CR rate, it’s about a 10% CR in papillary. I have to hypothesize that the IO is doing something or is somewhat responsible for that. David had data on pembrolizumab alone in papillary and there was activity, so I think it just makes logical sense to link the 2. I think the data support that, the large numbers support that, and just anecdotally in my experience, it’s gone really well in clinic. It’s maybe some soft reasons, a top hard data that this is really the standard of care for me in papillary.

Dr. Rini: David, as Betsy alluded to, KEYNOTE-427, I think it was cohort B, was pembrolizumab monotherapy in a similar heterogeneous non-clear cell cohort. I think the good news is in this field, we’re starting to get actual large studies in these diseases and not just case series and such. They’re not yet large randomized trials, but they are narrower confidence intervals so we can actually assess the activity. David, what do you think of these data, and what’s your approach to, let’s just use papillary as the most common example. What’s your approach to the metastatic papillary patient?

Dr. McDermott: I agree with my colleagues. The lenvatinib/pembrolizumab data is impressive. I’ve tended to use for papillary more nivolumab/cabozantinib, but that’s more just experience in what we have done, and maybe it’s the cabozantinib targeting metastasis story, which hasn’t been fully fleshed out, but maybe there’s something there, but I think lenvatinib/pembrolizumab is just as good, totally reasonable in that setting. I think to your point, though, we can accrue these patients to trials now, so we should be doing more of them and potentially exploring new mechanisms of action because we could actually execute those trials. I think those patients certainly need the trials, and also comparing datasets. Comparing the data that Martin collected with the KEYNOTE-427 data would be good not just from a clinical standpoint, as Betsy mentioned, but from a molecular standpoint. What are the signatures of the responders in those patients? That’s information that’s knowable and we should try to explore it with our industry colleagues I think.

Dr. Rini: You tend to give nivolumab/cabozantinib; lenvatinib/pembrolizumab is reasonable. Are you giving anybody ipilimumab/nivolumab? Any non-clear cell?

Dr. McDermott: Not generally. The data there is okay, but there’s no evidence that obviously across trial, that ipilimumab is adding much in those patients, so I generally stay away from it.

Dr. Rini: Okay, but non-clear cell with sarcomatoid, you might give ipilimumab/nivolumab?

Dr. McDermott: Yes. Absolutely.

Dr. Rini: Everybody agree with that?

Dr. Voss: I would say that there’s just no data for it, but the signal is so strong in clear cell that I don’t think it’s unreasonable, and the CheckMate 920 study did show efficacy, but regardless if it’s sarcomatoid or not. I think it’s the type of scenario where the patient may weigh in also. Sometimes I have a young patient who really wants a CTLA-4 inhibition because they’ve done their research and they feel very strongly about it, and I think then it’s our job to help them understand the limitations of the data that is there. They have non-clear cell. There just isn’t that strong data.

They need to know what the pros and cons are, what the risks are, and I would not keep that patient from the therapy. We’ve had a conversation. I get the sense that they understood what I tried to tell them. I think one other practical consideration that I think is important, especially for those of us who are at centers that have non-clear cell trials, sometimes it is important to also think through sequencing long-term, and with the median duration of response under 2 years, I think it is important to keep in mind what is going to happen afterward and what’s available.

Sometimes I might favor 1 regimen over another because in the second line, I already have sort something in mind or I have a strategy that’s emerging as a clinical trial opportunity for this patient that they otherwise might not have. If I had a cabozantinib-based second-line combination trial that I think is promising for that particular entity, I might be reassured to give them lenvatinib/pembrolizumab in the first line because then if it did fail in the first year or so, I might have a good option for them to sequence onto. Because I do not think that based on the data we have, lenvatinib/pembrolizumab is definitely stronger than nivolumab/cabozantinib. It’s a more robust dataset, it’s more patients, but I think the 2 trials tell the same story. High response rate in papillary, low PD rate, and so forth.

Dr. Rini: Agreed. For the group, is there any patient, barring absolute contraindications, that you’re giving upfront single-agent TKI, any non-clear cell patient you’re giving just cabozantinib monotherapy to any patient or I guess immune monotherapy? Barring again, the obvious solid organ transplant or barring something where you just can’t give the immune drug. Is there any patient phenotypes where you’re just giving 1 drug?

Dr. Plimack: I would say chromophobe. I’m really not convinced the IO is adding much. It tends to be a very cold tumor. I do sequence it and if there’s any PD-L1 signal, then I would give an IO-containing regimen. But those tumors tend to grow really slowly and shrink really slowly and you’re in it for the long game, and so I think it’s reasonable to do TKI alone there. I’ve done it.

Dr. Rini: You do TKI monotherapy and you choose cabozantinib?

Dr. Plimack: I choose cabozantinib.

Dr. Rini: Then if they progress and are PD-L1-positive, then you might consider immune therapy?

Dr. Plimack: If they’re PD-L1-positive, I would do it upfront. I would include it upfront. Yeah, if they’re PD-L1 negative, I would start with cabozantinib typically. Then lenvatinib/everolimus actually had some very small series data in chromophobe that I thought was interesting, so I’ve sequenced that in as well.

Dr. Rini: Lenvatinib/everolimus, just maybe for the audience, there was a grand total of 9 patients with chromophobe treated with lenvatinib/everolimus, 4 of whom responded, so it’s the most quoted 9-patient subset in the history of oncology. It certainly has activity, right? We’ve all used it, and it has activity, although I very much believe the 35% response rate from lenvatinib/pembrolizumab, which I believe wasn’t chromophobe versus that 44% where the confidence intervals are probably zero to a 100. But where do people use lenvatinib/everolimus in chromophobe? Martin, we’ll start with you.

Dr. Voss: I do use it. I do think that through the years it’s clearly come up again and again that M2 inhibition has merit in this particular entity. I think it’s reasonable to use lenvatinib/everolimus upfront in someone who’s symptomatic from their disease or who needs a response. It’s just notably more toxic than giving TKI monotherapy or TKI/IO. That’s based on personal experience, which I think is okay given how little data is out there. I sometimes favor everolimus/bevacizumab over lenvatinib/everolimus. We did a phase 2 study at our institution that included some chromophobe patients, and I had 2 patients that were on that regimen with me for years. They came off the bevacizumab at some point for the proteinuria, but they stayed then on the everolimus. It’s much easier in my experience to give them a TKI-based combo. But the strategy is the same to me. We know it’s effective. With all the limitations of the data, I think it’s reasonable to give knowing the toxicity, keeping that in mind.

Dr. Rini: I think as you can tell, when you get down to these super rare subsets, we all probably do something different and none of them are wrong or right, frankly, because they’re almost anecdotal-based. But I think we all agree we need better insight into the biology of non-clear cell in general and certainly chromophobe, but sounds like a relative consensus for IO/TKI in this patient population with the caveats and nuances that we mentioned.