Panelists Share How CLEAR and KEYNOTE-426 Data Presented at ASCO 2023 Impact Treatment Decisions

A roundtable discussion, moderated by Brian Rini, MD, of the Vanderbilt-Ingram Cancer Center, focused on treatment considerations for frontline and refractory renal cell carcinoma, including a discussion of new data presented at ASCO 2023. Dr. Rini was joined by a panel that included Tian Zhang, MD; Pedro Barata, MD; and Michael Atkins, MD.

In the next segment of the roundtable series, the panelists debate the use of IO/IO versus IO/TKI regimens for frontline RCC therapy, reacting to the updated data presented on the lenvatinib/pembrolizumab and axitinib/pembrolizumab combinations.

Watch the next segment in this roundtable series.

—

Dr. Rini: Tian, I’ll start with you again. Given what you heard at ASCO, these immunotherapy (IO)/tyrosine kinase inhibitor (TKI) updates, what do you think of them and is it going to impact your practice and how?

Dr. Zhang: It’s great to see that these overall survivals (OS) are improving for patients who receive immunotherapy-based combinations. I think you’re right in the hazard ratio is changing early on; it’s a model of the bad effects of sunitinib alone for the poor-risk inflammatory disease. But then as we get actual events occurring, then we have a more accurate model of the comparison. What my takeaway is back in, you all know better than I, 15 years ago where International Metastatic RCC Database Consortium (IMDC) criteria initially came out, we were looking at median survivals for these intermediate/poor-risk groups sy two years, nine months.

These have now been pushed out to 4 or 5 years. That’s really, really encouraging for me to see. When a patient with kidney cancer walks through the door, we’re not just looking at IL-2 or hospice, we’re talking through all of these treatment decisions and possibilities and that’s the opportunity to select the first-line therapy. To your point, every patient only gets 1 first line and one try at first line. How do we determine that optimal approach for first-line treatment I think is the challenge now. I love that we have these opportunities available. We have the treatment options available. My challenge in clinic is now how do I select the right patient for the right treatments?

Dr. Rini: Unfortunately, the data presented don’t really help us. But it sounds like if you were ipilimumab/nivolumab leaning and IO/TKI in select patients, these data probably don’t change your practice that much. I don’t want to put words in your mouth.

Dr. Zhang: I don’t think so. I mean use all of the combinations with maybe the exception of axitinib/avelumab is very low on my list. But the others I pretty much use if you take a smattering of my patients, they’re on all of the combinations as frontline approaches.

Dr. Rini: Pedro, what do you think of the data and does it impact your practice?

Dr. Barata: Great summary by the way. I agree. I’m thinking perhaps I’ll highlight 2 or 3 other points that we’ve been seeing as well. I think this data in general is kind of reassuring for IO/TKI users. In a sense that you started getting numbers, the median absolute number, the patients were all different in all these trials, but to your point Tian and about median OS, we’re in the 50-month range and higher for CLEAR, I believe it was 53, 54, we’re seeing the same with cabozantinib/nivolumab close to 50 [months]. I think the median on CHECKMATE-214 with ipilimumab/nivolumab is around 47, 48 months. So definitely we’ve come a long way.

The other thing now we are seeing is the association between depth of response and outcomes. We have that breakdown depends on how we define that. Do we want to group or lump together CRs [complete responses] and near CRs with the PRs [partial responses], 70% versus 50%. But no matter how we do it, that gives perhaps some arguments to folks who are IO/TKI users that, “Boy I’m looking after more response rates because that can predict longer, better outcomes in the future.”

Another way to look at that is to your point is what happens to patients who are able to get to the 3 year mark completing IO without progression? Obviously, we’re selecting up to probably a third between a fourth to a third of the patients. We saw that in CLEAR. We saw that in CheckMate-9ER. I mean we are seeing that now and those are destined to do very, very well in the long term. Finally, when we look at this, this is a heterogeneous group of patients. We have the poor, intermediate, and good risk, which by the way we’re not quite sure how good that classification is today, but we still use it because it’s the best we have. But this is still not enough follow up for the good-risk population for instance. Whatever we do, they’re going to do well. That’s perhaps one of the explanations where when we see the curves, no delta there, no big difference.

Part of that is we’re looking for follow ups of 40, 45, 5-year follow up. We need longer to actually assess a proper difference. At the end, I think coming out of ASCO, looking at all these data, I think people who think that the only way to offer a cure, whatever that might mean to us, is with an IO/IO approach. I think we’ll continue to do that, but I really think this data will support the use IO/TKI in all those cases. That might be 10% for you. For me it might be 30% or 40% or 50% of the times. It might be even more for the IO/TKI users out there in the community. I really think the data confirms that what we’re doing is the right thing.

Dr. Rini: It doesn’t sound like the data will change your practice that much because you used the word confirm kind of confirming those benefits. I think I also heard you say it’s not necessarily going to change a community oncologist who sees 3 or 4 patients a year or 5 patients a year. You don’t think these data will turn them into IO/TKI users or turn them into ipilimumab/nivolumab users?

Dr. Barata: That’s a great question because I can answer that in many different ways. What I mean confirm is when we look at how are IO users approaches or what is the landscape right now and what are people doing out there? We know there’s about 1 of 4 patients getting ipilimumab/nivolumab today, at least in the US. That means the rest of them, about 20% to 30%, are getting TKI monotherapy and the rest are getting an IO/TKI. What I mean by that is IO/TKIs are being more utilized today than the alternatives. I do believe these data will get some folks out there to start using IO/TKI in the good-risk population, for example, using a little bit less so of TKI monotherapy, for instance. Others would say, well I feel comfortable because the median OS for TKI monotherapy in the control groups of sunitinib, for example, actually doesn’t look that bad and I’m going to wait for longer data on that.

That’s what I mean by reassuring. I think the data is aligned with what people are doing and I don’t predict major changes in their practice. Now last answer is what do I think people are doing out there when they’re seeing 5 patients a year? I don’t think they’re trying one IO-based approach, a different one every single time.

I think what they’re doing is they get comfortable with 1 of them, maybe 2, and they’re going to use it more and more to get comfortable managing toxicity, which we’ll talk perhaps today and understand what to expect from that combo, whatever that preferred combo would be.

Dr. Rini: Okay, fair. Mike, I’m guessing you have some strong opinions?

Dr. Atkins: Yeah, so I tend to disagree with you. The IO/TKI regimens were compared to sunitinib and by the overall survival curves, as you went further out, converged. The overall survival hazard ratios, the 95% confidence intervals around each of those, were 0.99 at the top. Just barely significant. Keep in mind we’re comparing to sunitinib, not a pure immune therapy. If you look at the tails of those curves for the IO/TKI arms, they’re in the 18% to 20% range for PFS [progression-free survival], which is below the 30% tail of the curves at 5 years for ipilimumab/nivolumab. They’re not producing durable benefit. If you actually look at the response durations, they’re under 50% in terms of durable responses. While ipilimumab/nivolumab has still 60% of patients responding at 5 years; 85% of patients who are complete responders at 5 years. Even with pembrolizumab/lenvatinib for complete responders or near complete responders there’s a median duration of response. Many of those patients are progressing.

Dr. Rini: Can I interrupt you for one second?

Dr. Atkins: Sure.

Dr. Rini: Do you think that that lack or not lack of durable response, but the more limited durable response, and I totally agree with you, is it stopping the IO at 2 years? Is it just lack of ipilimumab upfront or both?

Dr. Atkins: I think it’s a mistake to think that stopping the anti-PD-1 is the reason. I think there’s a sort of misunderstanding about the way immunotherapy works. It works by producing a potent immune response early. That potent immune response can eliminate resistant cells if it’s potent enough. By the time you get to 6 months or a year, I don’t think it makes much difference anymore. I think as we get to the second-line discussion, I think we’ll see data to support that. I don’t think that continuing the anti-PD-1 would’ve changed things. I view what happened after 2 years in the KEYNOTE-426 study as just what happens if you’re getting a TKI. Over time, those patients progress, and the shape of that curve looks like a TKI curve. The shape of the IO/TKI curves look like TKI curves just pushed over to the right.

In summary, I think immune therapy needs to be given first because it works better in the frontline than it works in the second line. It produces durable responses and treatment-free survival responses that persist after treatment stops. As we’ll get to when we talk about in the second-line therapies, it doesn’t negatively impact the ability to respond to a TKI in the second line. Except for patients where you think you’re only going to get 1 shot at treating them, why not give them a shot of what looks like a functional cure at least?