 The last question I want to get to is about the future. What excites you the most for the next big thing in renal cell carcinoma? It could be a frontline treatment, it could be second line, it could be a new drug, or it could be a study. Anybody? I want to hear from all 3 of you, but what you are most excited about for 2021 ASCO, or beyond, in renal cell carcinoma? Monty, do you want to start? You’re smiling. You’ve got something.

Sure. We covered it at the time: Rana had mentioned glutaminase inhibitors.There are a lot of exciting target compounds emerging for the disease. I wanted to give a shout-out to a study that 1 of my former fellows from City of Hope wrote. He’s now in Brazil, Paulo Vergueiro. He wrote this beautiful trial looking at nivolumab-ipilimumab in combination with a probiotic. We’re most of the way through enrollment for this single-institution study, but I’m hoping we’ll see more of that: Trials that come out that try to hinge on other elements of RCC biology, acknowledging that hitting 1 target isn’t necessarily going to work for every patient’s disease. This is a study that randomizes patients to either nivolumab alone or nivolumab-ipilimumab combination with this unique oral probiotic that increases beneficial species with the gut—things like 

—and they may promote nivolumab-ipilimumab’s activity.

 That’s fascinating. You’ve published some great work in this area, and again, this is recognizing that there are unappreciated interactions between the host and these drugs as well as ways to manipulate it. That’s cool.

Rana, do you have a favorite out there? Maybe it’s going to be to your trial with radium, I don’t know.

That’s a good point. We’ve got radium-cabozantinib, with or without radium, for people with bone metastasis and RCC, looking at introducing a radio-pharmaceutical in the solid tumor space beyond prostate cancer. That trial is currently enrolling through the Alliance for Clinical Trials in Oncology. You brought up an interesting point about patterns of progression and the use of focal therapy, as well as working through the NRG on a focal therapy trial for progression post–I/O. There is an opportunity for the integration of multiple modalities and different mechanisms of action.

We know the new kids on the block with MK-6482. I’m excited to see how that’s going to show in clinical trials. It is probably 1 of the most tolerable agents out there, barring some anemia and hypoxia, which are easily reversible. It’s exciting to see the novel mechanisms of action of the growing list of new agents that are currently in testing.

 Fantastic. Neeraj, do you have a favorite up there?

I’d like to add that, in the immediate future, I’m looking at the approval of cabozantinib with nivolumab. I’m sure the data will be impressive, so I cannot wait to see those PFS, OS, and hazard ratio data, knowing that cabozantinib is the best TKI among all the TKIs for metastatic RCC.

Another trial I feel excited about, having used high-dose interleukin-2 for many years, is for the NKTR-214 drug, which is the PEGylated interleukin-2. That is being tested, as we know, in different trials with NKTR-214 plus nivolumab versus control arm, oral cabozantinib versus oral sunitinib. The control arm also includes cabozantinib.

There is another trial with NKTR-214 plus ipilimumab, plus nivolumab versus ipilimumab-nivolumab. The trial is already moving forward. These trials will hopefully give us higher complete responses that will be more durable.

Looking at the history of high-dose interleukin-2, I’m looking forward to higher CRs and more durable complete responses with this combination.

 Fantastic. Thanks to all of you for this rich and informative discussion. Before we conclude, I’d like to get final thoughts from each of you. First Neeraj, any final thoughts?

We can talk about all the data we can, but what it really comes down to is treatment decision-making. As Rana said earlier, patients’ wishes, patients’ preferences, and patients’ view all affect how the disease should be treated. Keeping that in mind, immunotherapy combinations and VEGF-TKI plus immunotherapy combinations are playing a huge role in fulfilling those wishes for our patients.

 Fantastic. Rana, any thoughts on your end?

The treatment landscape for advanced RCC has been rapidly changing. The paradigm has been shifting over the last decade and a half. Patients are living longer, and they’re living better, but we still have a lot of work to do. There are still gaps in where we are with treatment, so it’s been exciting to see these changes take place and be a part of that process. But until every person with RCC is cured, we still have work to do.

 Absolutely. Monty, you get the final word.

It’s a little wonky, and it’s something we didn’t talk much about today, but I would say to get genomic profiling done if you feel comfortable. At this point in time, it might seem to have more limited applicability of renal cell carcinoma versus lung cancer, but in terms of making the second-, third-, and fourth-line decisions based on biology, we can look for the usual suspects like poor pathway alteration.

I just saw a paper pop up today that suggested that 20% of renal cell patients have DNA damage-repair alteration. Down the line, it’s going to be useful to have those genetic bits of data in your back pocket, as Rana mentioned, especially for non–clear cell.

 Fantastic. Thank you all again, and thanks to our viewing audience. We hope you found this 

Videos

Future of Treating mRCC

 What about that patient now? We’ve alluded to this several times: The patient who’s been through several lines of therapy. Part of that, for sure, is disease biology. Nobody gets to fourth- or fifth-line therapy without having disease biology that is indolent. Some of this is the treatment effect, but whatever you don’t kill is presumably coming back stronger. However, these are patients who are able to tolerate that higher tumor burden, tolerate those multiple lines of progression, and tolerate those multiple lines of therapy. The patients have to be stronger as well. 

You get to these patients in this third-, fourth-, or fifth-line setting. Any philosophy on how you manage those patients in terms of the treatments? For our community guys, there are no clinical trials out there. There are guidelines, but there are no trials. There are no guidelines for fourth or fifth line. It’s all lumped together. What do you tell your colleagues? How do you manage it yourself in terms of these patients? Anybody? I’m opening it up. We’re getting loose; we’re near the end. Anybody, jump in.

I’ll jump in and say that, with the I/O–VEGF and the I/O combos being in the frontline space, whether somebody starts with nivolumab-ipilimumab or starts pembrolizumab-axitinib, they’re going to get 1 of those options. There are data for re-treating post I/O, certainly from the FRACTION-RCC trial, from retrospective series, I/O, post I/O has a response rate around 15%. Just because you’re out of that frontline space, it doesn’t mean you can never give somebody nivolumab-ipilimumab. Yes, there are insurance issues, but if you’re in the fourth or fifth line and have never seen nivolumab-ipilimumab, then you could consider that. If they have never seen pembrolizumab-axitinib or axitinib but have seen nivolumab-ipilimumab, then maybe you would consider giving them pembrolizumab-axitinib. That’s certainly totally off-label, but when you’re getting into that fourth- or fifth-line space and somebody has already received an I/O combo, like cabozantinib or lenvatinib-everolimus, and they’ve already seen a single-agent TKI,then you think, “What am I going to do?” It’s not out of the realm of reason to consider those combinations.

If somebody is still doing well, and you’re considering fourth- and fifth-line therapy, it’s important to think about a tertiary care center close by where you can refer for a clinical trial. There are a lot of interesting agents out there that are in-testing that have shown promise in early studies: The glutaminase inhibitor and the F2 alpha inhibitor. There are a lot of exciting things out there that are also going to change the second-line and post–I/O space, so thinking about those options for patients is important.

 We’re going to get to that in just 1 second, but I do want to just bring up something. For these patients who get to these fourth- and fifth-line settings, there’s 1 common theme. It’s not universal, but it’s a common theme: They have almost all had, at some point or another, some focal therapy. It may be for the brain metastasis that Monty mentioned, or it may be for bone metastasis. It may be radiation or CyberKnife. It may be a debulking metastasectomy or other symptomatic complication. For my chronic renal patients who get there, part of that is weaving in when to use those focal treatments. I don’t have any trial on this or any data. But from my own experience and looking at my own clinic [the Duke Cancer Center], I look at those long-term survivors, and it’s frequently been a mix. It’s been a multidisciplinary care effort to get them those 6, 7, 8, 10 years of metastatic renal cell carcinoma.

Treating Patients With mRCC Beyond Second-Line

 That’s right. Neeraj, I always give you the toughest questions, so I’ve got to come back to you here. I want you to think about this 1 carefully: Off-study, what are your criteria for switching somebody from frontline to second line? What constitutes real-world progression to you? Does that type of progression affect what you do in the second-line space? I’m going to follow-up on that second point, but right now, how do you call a patient a progressor in that frontline setting?

Obviously, this is not like what we do in clinical trials.

That’s a straightforward, 1-line answer to begin with. I never take a patient off treatment because 1 target lesion has grown out by 30% or 20%, while other nontarget lesions are stable or mildly increase in size. The most important criteria for me in my clinic at the Huntsman Cancer Institute to take a patient off-study is the clinical progression in addition to emerging progression. It’s never emerging progression alone unless there’s a high-volume disease progression or rapid disease progression, which I’m going to come to later. Half of this answers your second question.

Yes, the patients are doing well. In fact, today I saw a patient doing well with a 30% increase in the size of the main lung lesion, but other lesions are doing pretty well. He’s tolerating the treatment well. He’s planning on going on a vacation next month. I don’t want to rock the boat. We decided to repeat the scans 3 months from now.

In this case, I decided to continue the patient on the same treatment despite objective disease progression but obvious ongoing clinical benefit. The patient is going to come back to me after 3 months, and at that point, if the patient continues to have slowly progressive disease, I may continue the patient on the same regimen if clinical benefit is still happening.

There may be other situations: The patient may show a rapid disease progression with multiple new lesions. This patient, by the way, is on ipilimumab-nivolumab combination. I will use cabozantinib or cabozantinib as second-line treatment, or maybe lenvatinib-everolimus if, hypothetically speaking, the patient was on cabozantinib with nivolumab combination. If the patient was being treated on that combination trial and a patient is having disease progression or rapid disease progression, I’ll start lenvatinib-everolimus. If there is a situation where a patient who is on a VEGF–TKI [tyrosine kinase inhibitor], such as sunitinib—I still have patients on sunitinib—and they have mild disease progression, and I see that I have already noticed the disease to have progressed over the last 2 or 3 scans, and the patient wants to continue to be on an oral pill, then VEGF-TKI still has a part to play. I will start the patient on cabozantinib in the second-line setting. It’s so nuanced in real-life compared with clinical trials.

 It is, and those are good, illustrative examples of how you choose 1 strategy or another. Monty, any thoughts on those? Any other examples where a certain type of disease progression leads to a certain type of management, call it second line or just a modification of frontline?

No, Neeraj summarized it beautifully. I have very little to add to that. I struggle if, for instance, there is CNS progression. Those are certainly challenging episodes. Oligo-progression is another. It was alluded to earlier. I try to incorporate radiation therapy or surgery where relevant. I think he gave a great all-encompassing answer.

 What about this other scenario, where it’s not disease progression, but it may be chronic intolerance. We talked about the patients who have these severe, immune-related AEs. But what about good old-fashioned chronic fatigue or weight loss, whether it’s the drug or the diarrhea or what have you: patients dwindling down. Maybe we lower the dose of drug down as they get a bit weaker. Then there may be some slight progression, as Neeraj said. They’re just getting a bit worse, but it’s not anything dramatic. How do you manage that? In those scenarios, do you consider that it’s then time for a treatment break? Is it a second-line treatment? Do you switch class of therapy? How do you manage those patients, Rana? What do you think about in terms of your toxicity management around time for a treatment change? There is not necessarily clear progression, but it’s more toxicity driven.

I try to maintain a strategy where I can keep the patient on the therapy as long as possible. If that means a lower dose or maybe a lower dose with a modified dosing schedule within the week, if it helps them stay on therapy longer, then that’s the strategy to maintain as opposed to that persistent grade 2 toxicity. You’re right: It can be pretty life altering for patients. Especially with the TKI regimens, it may be that every 2 to 3 weeks, you get a weekend off. You get 2 days off, and that’s enough to just reset the clock. Patients will know. especially those who have been on the TKI for a long time, or if somebody may be having progression but was dose-reduced already. It’s not a lot of progression, but they dose-reduced for toxicity, and you’re not going to challenge them back to the next dose level; maybe you’ll do alternating days. I know that’s not necessarily in the guidelines, but that’s the clinical application. For example, with cabozantinib, you may have had somebody you dose reduced down to 20 mg because of toxicity. They’re doing great on the 20 mg, but maybe they’ve got a little progression. It’s not enough to say that you want to switch gears altogether. I’m not going to take you back up to the 40 mg. But 3 days a week, we do 40 mg, and the remaining days we do 20 mg. We’ll get creative because what I want to do is this: I want to milk every ounce of goodness out of that therapy that you’re on because it’s not as if we have an infinite number of options at our disposal. I try to maximize while also being cognizant to not lose a window of opportunity where you may not be able to get them on the line.

Sites of disease also matter. I’ve had patients with pancreatic metastasis who chug along, and they’re changing over time, but there is no other disease elsewhere. We know that pancreatic metastasis behaves different from other metastatic sites. Maybe they’re having some lung progression, and those patients are biologically different, so being creative to limit toxicity and maintain efficacy is key.

 Those are great points. I take the same philosophy. Similar to what Neeraj said, I try to keep that frontline therapy going as long as possible even through those slow-progressing signs and things. As you said, Rana, to me the reason is this: Every time we make a treatment change, there’s a risk. There’s a risk when we’re taking a break from therapy to make that treatment change. There could be a flare, there could be trouble getting on that next regimen, and there could be insurance issues or other issues. There could be a tolerance issue with the second regimen that’s worse than the first, yet it’s tempting. 

I know you said we don’t have that many treatment options, but last time I looked there were 11 drugs for renal cell carcinoma. I’m going to talk about this next: I’ve seen people who have gone through multiple lines of therapy: 5, 6, 7 lines of therapy. To me, the goal is not to use every drug in the book. That’s different from prostate cancer, where I’d like to use every drug in the book. In kidney cancer, I like to maximize each treatment line as long as I can without being pigheaded or stubborn about it. I use these creative techniques because when we either modify the dose or interrupt the dose—I like interruption more than modifications, by the way. But when we interrupt doses like that to make the drug more tolerable for patients, we’re not necessarily compromising the biologic effect. When we start at a lower dose and plan to start a drug at half its dose, we may be undermining the effectiveness of that drug. When we start at the full or near the full dose and then we modified it based on tolerance, then I feel like, because that’s a drug-driven adverse effect, then we’re still getting an adequate dose in patients. No, I don’t check pharmacokinetics on every patient, but that’s my poor-man’s pharmacokinetic: Is it a drug-related adverse effect, and if it is, I don’t mind making modifications around it. I like both of those strategies. Monty gets a pass because he agrees as well. We’re all saying the same thing on this.

Strategies for Switching From Frontline Therapy to Second-Line Therapy in mRCC

 I want to go back Monty. He did such a good job summarizing our frontline phase 3 studies that got us into this frontline space. Monty, in the second-line space, it’s a bit older. These studies aren’t quite as fresh, but they’re still the best studies we have in the second-line space, and they’ve led to a lot of the approvals of these drugs and drug combination that we have in these patients with previously treated metastatic renal cell carcinoma. Can you walk us through what data we have and what’s shaped the NCCN Guidelines and other guidelines based on those studies?

I can try. I’m still scarred because you gave me an A-minus, and I’ve never gotten and A-minus in my life.

 I believe that actually. I believe that.

I’ll dial it way back. I’ll go back to RECORD-1.

Yeah, it was way back, right; everolimus versus placebo, a very modest benefit in terms of progression-free survival. We then go to AXIS, which was a comparison of axitinib against sorafenib, not a placebo control. It was a modest gain: about 2 months in terms of progression-free survival with no overall survival OS benefit. Then came the slew of studies that reset the bar for second-line treatment where you had to have an OS advantage to demonstrate effectiveness. CheckMate 025 compared nivolumab versus everolimus; it showed benefit in terms of response rate and overall survival. You then have the trifecta: The METEOR trial came with an improvement in progression-free survival, response rate, and overall survival; cabozantinib over everolimus. Another study that becomes important in the context of the current first-line discussions that we’re having is lenvatinib with everolimus. This was examined in the context of a randomized phase 2 trial looking at lenvatinib-everolimus versus lenvatinib monotherapy versus everolimus. With that comparison of lenvatinib-everolimus against everolimus monotherapy, you see a significant benefit in terms of progression-free survival with a modest benefit in OS. However, it was certainly not the statistic that demonstrated that, with only about 50 patients in the treatment arm. That’s a quick rundown of the available second-line data sets that we have. Dan, am I missing anything there?

 We won’t ding you on the AXIS study. That’s OK. That’s great, but all those studies were done in the era when we weren’t using frontline I/O. If there were some frontline I/Os, then it was a small proportion of the patients. We’ve now got both frontline I/O and frontline TKI. How do you interpret these now? Do you say, “Well, there are those same robust benefits and everything up front?” Are you now rethinking it? Do we need a whole new set of phase 3 studies in this refractory setting to guide us?

I’m leaning a lot on these retrospective data sets. They come with all the caveats of encompassing a highly selective patient population, and a lot of these are limited to robust academic medical centers. The signal that seems to be emerging is that there is good, if not better, benefit in the second-line setting with agents like cabozantinib and similar data for axitinib. I’m leaning heavily on those to establish what I use in the second-line setting.

To echo something that Neeraj said earlier, the cabozantinib has become my preferred second-line treatment if I haven’t already used it front line for a patient.

 That’s exactly right. It’s interesting. The same logic holds true in the second-line as it does in the first-line setting. We can’t necessarily count on patients getting to third- or fourth-line treatment. If you look at the number of patients who get to second-line therapy but don’t necessarily get beyond that, at least when you look at some of these larger outcome databases, it’s about half. It suggests that we need to choose wisely in the second-line space.

Rana, what are your thoughts on the guidelines in the second-line space? Where do you use the guidelines to help you? Are they helpful at all? Do you have any thoughts on how you tease out what to pick next in these patients?

Unfortunately, right now, I don’t think the guidelines for second line reflect how we’re practicing for first line, because all these trials, as you’ve stated, were conducted in the VEGF era, with VEGF/TKI alone in the frontline space, which is not how we currently practice. There is a critical unmet need there for deciding the best treatment option post–I/O. How do we select therapy in somebody who has received an I/O combination, whether it’s ipilimumab-nivolumab or whether it’s pembrolizumab-axitinib, golimumab-axitinib, nivolumab-cabozantinib, or what have you? Our frontline treatment is changing. Therefore, not to say it has to change, but we need to evolve the second-line treatment to match the frontline setting.

This segues into the CONTACT-3 study, which is ongoing. It is a critical trial that is definitely meeting an unmet need in several aspects. One, it is a post–I/O trial, and it’s designed that way: post checkpoint inhibition. It’s also a trial that includes non–clear cell. Largely everything we’ve been talking about thus far has been clear cell. All the trials were designed for patients with some clear components. This trial includes patients with unclassified and papillary histologies, which is a great unmet need. The trial is randomizing patients to cabozantinib with atezolizumab versus cabozantinib alone. It’s going to answer an important question about what the role is of I/O, post I/O, and what the best strategy is in the second-line space post–I/O–I/O or I/O–VEGF strategy up front.

I agree that cabozantinib is the best TKI we have to treat kidney cancer. It combats that mechanism of resistance to VEGF inhibition with enzymic activity, and it’s the best drug out there from a VEGF standpoint. When cabozantinib first came around, we didn’t know how to use it well: the dosing was a bit off, but I find that it’s probably 1 of the easiest TKIs to use. Whether somebody starts at 60 mg, 40 mg, or 20 mg, it’s an easy TKI to use, and the adverse effect profile is predictable. It is my go-to if somebody has not received it in the frontline space. That may all change with the nivolumab-cabozantinib data, where we’re going to have to wait and see what that data look like when you use your best drugs up front.

Piggybacking on that, COSMIC-313 is also looking at the triplicate of nivolumab-ipilimumab-cabozantinib versus nivolumab-ipilimumab. It’s an important trial where the control arm is an active I/O–cabozantinib regimen. The field is going to continue to evolve over the next couple of years, but we need strategic trials looking at sequencing. What is the best sequence, and what do we do for the non–clear cell cases? Do we mirror what we’re doing for the clear cells? Those are unmet needs, but I don’t think the guidelines reflect right now what we’re doing in frontline treatment. They’re a little outdated, so we’ve got to wait for the trials to read out.

Clinical Trials Shaping Treatment in Relapsed/Refractory mRCC

 With that, I want to move now to the second-line setting. We’ve talked about all this change in landscape in the frontline setting. The rapid uptake of these TKI [tyrosine kinase inhibitor]/IO combinations. Right now, it’s axitinib and either pembrolizumab or avelumab, but there may be more coming. We mentioned the Checkmate-9ER study, and there’ll be other studies with pembrolizumab, lenvatinib, and others.

I want to think now as we start to use more of these combinations in the frontline setting: what does that do to our second-line setting? How are we thinking about second-line therapy now in renal cell carcinoma? Does it mean that we have to start all over again, or can we apply some of the older data? What are some of our goals in these second-line settings, and how do they differ from the frontline settings? What is the impact now on the tolerance and quality of life of patients in second-line and beyond? Neeraj, you want to kick us off in this area?

Yes. It’s an interesting question. Nivolumab was one of the standard second-line agents until just last year, and now with ipilimumab/nivolumab being used in first line or axitinib/pembrolizumab and soon cabozantinib/nivolumab combination, I’m sure it will show fantastic results based on the known efficacy of cabozantinib as a single-agent TKI, which you have established, Dan, in the first-line setting. What is the role of nivolumab in the second-line setting ?

We are seeing rechallenge data: there are data coming out after using these PD-L1 inhibitors again in second line after prior use of PD-L1 inhibitors in the past, but most of them are being used in combination with VEGF/TKIs.

Right now, if you asked me about second-line nivolumab, what is the potential, what is the market, or what is the use of second-line nivolumab in my practice? It is pretty much nil. It’s nothing because most of my patients are getting one of the PD-L1 access inhibitors first line, barring a small number of patients who are getting single agent VEGF/TKI. In those patients, I sometimes tend to use VEGF/TKI again because they’re responding so well to the first VEGF/TKI. That’s how it is panning out. With pembrolizumab being used in the upfront setting, or ipilimumab/nivolumab being used up front, the default choice for me is cabozantinib in second-line therapy. This is the most efficacious monotherapy in metastatic RCC.

Once patients progress on single agent monotherapy with cabozantinib, the most commonly used regimen in my clinic is lenvatinib with everolimus. Now, if the cabozantinib/nivolumab combination gets approved next month, I hope, then I’ll start using the cabozantinib/nivolumab combination in the first-line setting. I’m going to struggle because I’m sure this is going to be an effective combination based on whatever experience we have with these 2 drugs. The second-line agent may end up being lenvatinib with everolimus, or lenvatinib with pembrolizumab if it gets approved based on second-line data presented at ASCO. It’s going to be fascinating to see how things are going to evolved in the next 1, 2, or 3 years.

 Fantastic. Let me ask you 1 quick follow-up question because we talked about IMDC in the frontline setting; is there an IMDC in the second-line setting? Do you use it? How do you use risk stratification in the second-line setting?

No. I only use IMDC in the first-line setting.

 In the second-line setting, you treat them all pretty similarly?

Yes…but I don’t treat them similarly; I look at many other nuances. For example, what kinds of drugs have been used in the past, their performance status, the patient’s wishes, their choices, and so on. The IMDC criteria do not play as much of a role in the second-line setting as they play the dominant role in first-line setting in my clinic. The IMDC doesn’t have much of a role in the second-line or third-line settings.

 Monty, any thoughts on this? Do you think about risk stratification?

I tend to agree with Neeraj. Although you certainly can opt to use the IMDC second-line criteria, it’s that clinical nuance that helps me decide what treatment to allocate to patients. At this point in time, I don’t think the second-line IMDC criteria are refined enough to help us allocate to 1 treatment or another.

 It’s interesting, isn’t it, because it seemed like it had some relevance early on when we’re using TKI monotherapy, but now that we’re using these combinations in the frontline setting, it’s almost like everybody who progresses is at least an intermediate risk. There’s at least some biology there that’s a little bit more aggressive. It’s also interesting to me, now that there’s been data out about rechallenge with agents. For a long time, we’ve known you can rechallenge with TKIs, and you can use different TKIs 1 after the other. In fact, some of these drugs got their approval in the setting of frontline TKI failure and then a subsequent TKI. We’re now beginning to see some data with IO studies that you can challenge with 1 IO after another, maybe with a treatment break of something different in between, but there can be continued response. Therein lies the hope, Neeraj. Even though we may be using some of our best agents and combinations up front, we’re not emptying the tank. There’s still opportunity to revisit these strategies. It may be in newer combinations over time and get continued, maybe not complete responses, but at least some disease stabilization and clinical benefit.

Choosing Second-Line Therapy in Treatment of mRCC

I’ve got to ask Rana to comment because she’s already talked about some of these new tailored therapy approaches, and the one we haven’t touched on is the Phase 3 PDIGREE study, and for my colleague, Dr Tian Zhang, I’d be remiss if I didn’t give a shout-out. This is another innovation in the field, it’s an adaptive, maybe not tailored, but an adaptive design.

Rana, any thoughts on that study and other studies that give us an idea of how we can begin to navigate not just making 1 treatment decision up front, but how we can begin to merge some of these strategies together?

Absolutely. The PDIGREE trial answers a credible question, and, even now, that’s more important with the Checkmate-9ER data being positive. It’s a critical question in the field of how we can escalate therapy for people who need therapy escalation, and can we do so for people who aren’t in the CR category and aren’t in the PD-L1 category? For the bulk of our patients who are in the middle, how can we get those responses to be deeper? How can we convert somebody with stable disease into a responder? That’s the messaging behind the PDIGREE trial. Patients start out with nivolumab/ipilimumab as the backbone, and then, depending on their response, if they fall into that middle category of non-CR, non-PD-L1, then they’re randomized 1-to-1 to continuing the nivolumab or adding cabozantinib to the nivolumab. It’s a critical trial answering a critical question in the field. 

There are other trials that have been reported out. The TITAN-RCC trial reported at ESMO last year. The HCRN study led by Dr. Michael Atkins and our OMNIVORE trial at the University of California San Diego Medical Center also looked at answering the question of whether everybody needs nivolumab/ipilimumab up front. A third of the patients have toxicity; does everybody need it up front, or can we take those patients who don’t have a CR and don’t have a PR, and then layer in the ipilimumab? Whether we do 2 or 4 doses at a time, can we layer it in later to see if we can convert patients into responders or not?

Now, these studies were small. They were phase 2, but they weren’t randomized trials to compare it with the standard of care, which is nivolumab/ipilimumab up front if you were going to do the IO/IO combination, but they were illustrative of the fact that the upfront ipilimumab is important. For all of these 3 studies in aggregate, the CR rate was quite low: less than 10%, ranging from 2% to 0% in some of the trials. For the conversion rate, the rate of converting somebody who’s got stable disease into a responder was about 10%. It speaks more to the idea that, if we had a biomarker to identify them, then, yes, but we don’t. It recommended against that approach without more data, but these trials are very illustrative because a clinician may be faced with the idea, “Let me try the nivolumab, and then if they don’t do well, I’ll add in the ipilimumab. I want to spare the patient the toxicity.” It’s about shying away from that. These trials have shown us that. If there is somebody who you are worried about with the ipilimumab, and you’re worried about toxicity, there is this 10% conversion rate. The early phase 1 studies showed that you can salvage patients with ipilimumab/nivolumab, but it can’t be a received practice.

Yeah. It’s a great study even though you might consider that a negative result. It’s an informative result because there’s such a temptation to do just what you said, and that is to say “Well, let me start with 1 and add another.” We do it all the time in oncology, and we think, “Well, that won’t matter,” yet here’s a study showing that, in fact, it could matter a lot. You’re absolutely right. It’s an important, informative study in terms of whether you’re considering that combination. To get the best results, you need to consider that up front.

Tailored Therapy Approaches for Upfront mRCC Treatment

 Monty, any thoughts on your steroids use in these patients? Do you use steroids the same way when you’re dealing with an immune-related AEs for say ipilimumab/nivolumab versus an immune-related AE for say pembrolizumab/axitinib? Is there any difference in how you use the steroids in those different scenarios and discontinuing and restarting drugs in those scenarios?

I wish I could give you something nuanced for the process of management. I find that I am quite similar in my strategy for both. I start with Solu-Medrol or prednisone, maybe in the context of some GI or liver-based toxicity from the drugs. Perhaps in the context of nivolumab/ipilimumab, I’m a bit more liberal in jumping to the second-line therapy just because we’ve all seen situations where it’s significant.

 That’s fair. I feel the same way: for the combination of IO/IO, you can’t reverse that. With the IO/TKI [tyrosine kinase inhibitor], we would hold the TKI, so it’s going to wash out. It may take a few days, or it may take a week, but it’s going to wash out. With the IOs, they’re not necessarily going to wash out. For a lot of these toxicities, they sometimes happen with the very first dose, but more often, it’s with that second or third cycle, so you’ve loaded them up, and that’s going to take awhile. I think about treating with steroids a bit longer in those courses, and I think about what you said: moving on to the second-line agent a little bit sooner if we’re not seeing the results. Those drugs work much better if we do them earlier rather than later. That’s helpful advice.

Finally, are there patients who you would look at up front and say that they are not an IO candidate? Are there criteria? You mentioned some of the things in the past that excluded them from trials, such as autoimmune conditions, but are there other factors or other things you’re concerned about with IO therapy that you would hesitate and consider using more straight TKIs?

Rana and Neeraj both alluded to this, but it’s when you have that sense or that gestalt that the patient is not going to be very compliant. Say, for instance, that the patient had their nephrectomy at your institution 10 years ago and then never went for a single follow-up scan, and they show up a decade later with extensive metastasis. That’s the individual for whom you may balk a bit. You have to have a certain level of trust between physician and patient to apply IO-based therapy.

 That’s exactly right. It’s challenging because you’re just getting to know these patients; they are new patients to us, and we are making a lot of important decisions up front and trying to get a sense of this. This is an important aspect to this.

Treating Side Effects of Therapy for mRCC with Steroids

 Before we move on from this frontline setting, though in the interest of being fair and balanced, it’s important for us to recognize the limitations of these regimens in terms of their tolerance. They’re different. The reality is that we have difference in not just the adverse effects but in the mechanism of those adverse effects as well, from things like diarrhea and liver function test abnormalities to the chronicity of these adverse effects with TKIs versus with IO therapies, as well as how we manage it with steroids versus other supportive cares and now beyond steroids. The world of immune modulatory symptom management has gotten incredibly complex and requires multidisciplinary teams, as Rana mentioned early on.

Rana, do you want to talk to us a bit about adverse effects?

 Take on the IO/IO combination, the IO/TKI combinations, and TKIs alone. How do you size that up, and what are you looking for? Then, when do you pull the trigger on these multidisciplinary teams?

It’s important to understand the nature of the toxicity profiles that we see with the VEGF targeting agent or IO/VEGF therapy versus IO/IO. With VEGF-targeted therapy, the toxicity tends to be dose-dependent: it’s predictable, it’s chronic, and it’s sometimes on-target and sometimes off-target effects of therapy. When we look at the IO regimens, specifically the PD-L1 targeted therapies, they’re not dose-dependent, and they’re not necessarily predicable. With ipilimumab, we know that the toxicity can be dose-dependent, but it is also not predicable. It’s hard to know who’s going to get toxicity with IO/IO. We don’t necessarily have a biomarker to say we should be barring somebody who has an existing autoimmune disease; we don’t have a biomarker to say you’re going to get it or that you’re not going to get it. I have no idea who is going to be in the one-third category to whom I’m going to end up needing to give steroids or the small population of patients who end up getting infliximab or TNF-alpha condition. One of the biggest things is communication with the patient around the toxicity, and especially the IO/IO toxicity because of the somewhat unpredictable onset.

With that being said, when we look at the toxicity profile curves from Checkmate-214, most of the severe grade 3/4 adverse events did seem to come on within the first 6 months or so of therapy. Once the ipilimumab was removed, discontinued, or stopped, and they remained on PD-L1 inhibition alone, then the degree of severe toxicity did come down; whereas, when we look at the IO/VEGF, there’s that chronicity of toxicity that we continue to see from the chronic ongoing TKI. The other confounder with IO/VEGF is that there’s a lot of overlapping toxicity, particularly with the diarrhea and the hepatitis, that you’re asking, “Which one is it?” You don’t want to call everything immune-mediated right away because that has significant implications for rechallenging, steroids, and tapering.

The beauty of axitinib is that it’s got a short half-life; you can quickly stop it. You can try to tease out, “Did the diarrhea get a bit better? Did their LFTs get a bit better? Okay, maybe it’s the axitinib. Let me ride this out and maybe redose them at a lower dose or something like that.” It requires a lot more nuance and a lot more checking to see what it is, whereas, with the IO/IO, unless somebody’s got some concurrent ongoing comorbidity or something, you can point to the IO/IO agent, the nivolumab/ipilimumab.

The biggest thing is communication with the patient and, honestly, the communication to the nursing staff, as well, because they are the people who are at the first lines. When a patient calls and says, “I’m having a little bit of diarrhea,” well, how much is a little bit? What’s your baseline? We have huge education sessions with our nursing team, our LVN [licensed vocational nursing] team; the people who are the front lines, who triage calls, and who triage symptoms. We instruct the patient, for anything that you’re not sure on, you can call us, and we’ll walk you through it, or we’ll triage for you. Don’t do the triage yourself. It’s about educating the entire care team.

When there are significant grade 3 or 4 adverse events, there may be delaying the initiation of steroids or dragging your feet around it. Where I’ve gotten burned has been where I’ve just thought, “Should I, or should I not?” If there’s a high index of suspicion, and the patient’s not doing well, start them on steroids. Historically, we used to do a full-fledged diagnostic work-up to get a biopsy to prove that it’s immune-mediated, but we don’t necessarily have to go through that if the patient is sick and there’s a high index of suspicion. Just start, and if there’s no improvement, then add the infliximab.

 That’s great advice. I’m curious: we’re all at academic centers, and we at academic centers have specialists and subspecialists in these areas of GI, or dermatology, or cardiology or what have you, who help us when we get into these complicated cases and whatnot.

How do you advise folks in the community, Neeraj? You’re up there in Utah, and you have a lot of community doctors in relatively rural areas. How do you advise them on how to manage that and when to send the patient to a tertiary center for help with this or when to bring in their local specialist who can help them recognize that this is something more serious with this rash, or something more serious with this pulmonary situation? How do you advise folks there?

First of all, my experience is that my community colleagues are smart clinicians who are able to manage patients with a hundred different kind of cancers. That’s the first impression I have of my community oncology colleagues. To start off, I wouldn’t say advise; I would say for patients with metastatic kidney cancer, I may have used ipilimumab/nivolumab just recently, but many of my colleagues are using this for melanoma for many more years. By far, VEGF/TKI plus PD-1 access inhibitor combos are more popular with the community practices as opposed to ipilimumab/nivolumab, which is the preferred regimen, at least in our practice, in academic institutions. This has to do with the unpredictability of the colitis part with ipilimumab.

The single most important factor that drives the uptake of VEGF/TKI plus PD-1 combos versus ipilimumab/nivolumab in academic cancer centers is that we have more support: supportive oncology clinics, triage lines, nurse practitioners, and fellows. These are usually not available in community practices, especially, and as you said, Dan, in the Intermountain West area. Many practitioners are all by themselves and 300 miles out, so if somebody is getting ipilimumab, I tell them to have a low threshold for starting steroids versus somebody who is receiving axitinib plus pembrolizumab, for example.

 That’s great advice, and Rana said this, as well: we’d rather be safe than sorry.

Evaluating Safety Profiles of Upfront Regimens for mRCC

 When you have a patient with renal cell carcinoma [RCC] and use our IMDC [International Metastatic Renal Cell Carcinoma Database Consortium] criteria, do you have a sense that, “I have got to pick my best regimen up front because they may not make it to second-line?” What drives that sense of immediacy for treatment?

Always. You, Neeraj, Rana, and I have all contributed to several of these data sets, both in the context of IMDC and outside of that context, that reflect the high attrition from first-line to second-line treatment that we saw in the TKI [tyrosine kinase inhibitor] era. For instance, some studies, like the IMDC data sets, reflect how just 50% of patients make it to second-line treatment. Those numbers may be a bit better in the context of VEGF [vascular endothelial growth factor]/IO [immune-oncology], but we just don’t know yet at this point, by and large. I would suggest that data, which implies that a substantial proportion are not going to get the second line, means you can’t save the best for last here; you’ve got to put your best foot forward right up front.

 I agree with that. You mentioned bone metastasis; to me, symptomatic bone metastasis is one of the things that I worry about in this disease. These patients can run into trouble early on. Another thing is tumor burden. Patients who have symptomatic, high tumor burden when they have progression and they’re already carrying a critical mass of tumor burden…to me, the third is the type of symptoms the patients have. I mentioned pain, and the other to me is weight loss. When I have a patient, and we see this with kidney cancer, the patient has a lot of weight loss. We’d like to think that’s like liposuction and body fat, but it’s mostly muscle. When you lose muscle, it makes everything that much tougher to tolerate and be able to get to your second or third line of therapy. There are a few key factors to me when I’m looking at these patients. Some weigh a bit more than others, and that really makes me think, “Gosh, I have got to take that first best shot.” If they don’t tolerate something, then I can dial it back, but we may not get a chance to do some fantastic salvage regimen if they don’t get some disease reduction in this first-line setting.

I want to be conscious of time, but I want to bring up one more thing that Rana brought up, and then I want to ask about adverse effects because these are the other key things. Rana brought up biomarkers, and biomarkers are key. I know all of us have done studies and invested a lot of time and energy in understanding biomarkers. There is 1 biomarker that’s out there: it’s prevalent, and it may show up on every tumor profile you’ve been getting it in your path reports regularly. It’s the PD-L1 [programmed death-ligand 1] status. I’m curious. When you see a tumor that’s PD-L1 positive or you see a tumor that’s PD-L1 negative, how does that factor in your decision? Does it factor in your decision at all in how you’re going to manage that patient? It’s not part of our IMDC. Neeraj, do you use that information?

I get that information, but I don’t use the information. That’s how I report that.

Most of the sequencing platforms report this information.

 Yes. You’re interested but not enough to act. It’s not going to weigh in your decision on how a patient is treated.

I think more about whether I’m going to use a PD-L1 access inhibitor and whether I’m going to use a combination regimen versus a VEGF/TKI-based regimen. That’s a bigger decision-making factor in my clinic at the Huntsman Cancer Institute in Salt Lake City, Utah.

 Do the others have the same feeling? Rana, you’re smiling. Do you use it a little?

First, we know that PD-L1 status in RCC is prognostic. People who are PD-L1 positive seem to do worse compared with people who are not. When we look at, for example, the 3 first-line trials with approved therapies in the frontline space, each of those trials used a different assay to determine PD-L1 positivity. It may be that the responses are going to be higher in patients who are PD-L1 positive, but patients who are PD-L1 negative can respond. It’s not a pure biomarker. You’re not going to restrict therapy around PD-L1 status. Right now, it hasn’t been integrated in clinical practice. Yes, we get it, we look at it, and we haven’t even validated the right assay yet. There is still a lot that needs to be done.

In each of these studies, there was some elegant work that came out from JAVELIN Renal 101 (NCT02684006) and Checkmate-214 (NCT02231749) evolving, but each trial did its own set of biomarker studies, and there was not a lot of integration across all trials. It urged the field for us to try to synthesize that and integrate the same biomarker panel across these trials, so we can learn something and try to better select our patients beyond just IMDC criteria. I don’t think we’re there yet, so I still think we have a lot of ways to go.

 It seems like something that could be induced. We’re using archival tissues to describe this, and you’re then treating patients with, say, a VEGF/TKI and an IO, how much of that PD-L1 status is changing in vivo in these patients and we don’t even know it. There are a lot of limitations in that archival tissue, but something that probably doesn’t change is something like a sarcomatoid element, and that’s an aggressive element to a tumor. It’s probably a bit more of an inflammatory element. Certainly, it has a different histology associated with the disease, and we don’t see it that often, but maybe 10%, 15% of the time we’ll see some significant component of a sarcomatoid element….

One of the most impressive outputs of all the immunotherapy studies has been the benefit that we see with the application of immunotherapy with sarcomatoid features. If I see a patient with a substantial proportion of sarcomatoid elements, then I’m definitely steering them toward nivolumab/ipilimumab.

 It’s interesting because you see it not only with the response, but even with a complete response, which, to Neeraj’s point, is remarkable because the sarcomatoid elements are rarely 100%. You usually get 10%, 30%; it’s some proportion of the tumor, yet you’re having much more dramatic effects, and it suggests that there’s some sort of bystander effect once you’ve got that immune activation associated. It starts killing all the tumor, and you can get these remarkable complete responses. We’re not necessarily talking about taking a couple of small lymph nodes and shrinking them away. These are patients with liver and bone metastasis responding. I agree with you. These are the cases for which, even if it seems a little bit of a stretch, I’ll consider that because it has such a dramatic result associated with it.

Choosing Upfront Therapy for mRCC

 It almost seems like every quarter there's a new regimen that’s coming out that’s showing excitement.

Monty, in April of this year, amid all the COVID-19 [coronavirus disease 2019] stuff and everything else we had going on, CheckMate 9ER had a press release showing a positive progression-free survival [PFS] and overall survival benefit to that. Of course, we haven't seen the data, but with first impressions from that press release and what you know about that regimen, what's your take on that? Is that going to be a viable option in this frontline space?

It’s impressive. My sense was that, after bevacizumab/atezolizumab came out, all the VEGF/IO [immune-oncology] regimens to emerge thereafter would face an uphill battle. If you’re competing against sunitinib, you’re going to have more and more patients who get upfront an VEGF inhibitor cross over and get second-line immunotherapy, then that would certainly make an overall survival bar even higher. Seeing CheckMate 9ER emerge with this impressive signal for overall survival and PFS advantage is incredible. I can't wait to see the numbers.

To a question that you asked earlier, what sort of follow-up are you going to need to see with some of these studies? I’d have to tell you that that may be contentious in the context of some of these emerging studies. With a regimen like the one they assessed in CheckMate 9ER with cabozantinib and nivolumab, I can definitely think of places where I'm going to want to use it up front, irrespective of the fact that it doesn't have substantial follow-up or won't have substantial follow-up at the time of the initial report. For those patients with bone metastasis and those patients with aggressive disease that were cast under the CABOSUN trial umbrella, those are folks that I’d probably want to preferentially treat with cabozantinib/nivolumab up front.

 When I look at those Kaplan-Meier overall survival curves, it doesn't matter if it's CheckMate 9ER, we haven't seen that one yet, but CheckMate 214 or KEYNOTE-426. What's impressive is the separation of these curves early on and within the first 12 months. That’s impressive because, remember, these are clinical trial patients. These are patients who met eligibility for clinical trials who had reasonably good performance status. They were intermediate- or high-risk, yes, but they were able to go on study, yet within 12 months, we see 25% of these patients in the sunitinib arms dying of disease. It tells you, as a quarter of this population is dead at 12 months, the immediacy of picking our best regimens. The separation of those curves early on in those first 6 to 12 months tells you that, for the sickest patients, these regimens are making a huge impact. You could talk about whether there's crossover and everything, but for these patients, there is no time for crossover. For these patients, making the first best decision is critical because they may not have a second-line therapy. We're going to talk about second-line therapy in a minute.

CheckMate 9ER:  Phase 3 Trial of Frontline Nivolumab + Cabozantinib

 Neeraj, these trials have different levels of follow-up. Some of these studies have just come out. You mentioned the CheckMate 214 study just presented 42-month follow-up data. How much of that factors into your comfort level in a recommendation for this frontline treatment, not necessarily having that 3- or 4-year follow-up data? Is that an issue for you? Is that something you feel comfortable with, still recommending some of these newer regimens?

Yes, a longer follow-up is always compelling for me, especially when I'm looking at the durability of complete response [CR], which is the first thing that comes to my mind when I think about ipilimumab/nivolumab. Yes, a 42-month follow-up is a big deal, and it continues to reinforce that this regimen induces those durable complete responses, at least until now. But not necessarily in the context of VEGF/TKI [tyrosine kinase inhibitor]-based regimens. When I'm not counting on complete responses, as Rana and Monty said, there are many patients who are going to be candidates for VEGF/TKI plus IO [immune-oncology] combinations. For those regimens, the durability of complete response may not matter as much as the overall response data, PFS [progression-free survival] data, or the absolute OS [overall survival] data. It’s relative. I don't expect, for example, cabozantinib, to induce durable complete responses for 5 years. However, I have patients, both in the METEOR trial and the original sunitinib trial, which randomized patients to 50 mg every day versus 37.5 mg, and I have a patient still responding to sunitinib from that trial. At the same time, I'm not expecting cabozantinib to show 5-year complete durability of response in most of my patients, so it's all relative. If I see data from a 42-month follow-up, it is definitely compelling.

 Neeraj, when you talk about a complete response, how complete is a complete response? Does it, in your mind, have to be pristine with not a millimeter of things left, or if there’s a 90%-plus reduction, then that’s good enough for completeness from your perspective?

I agree with you if it’s a near-complete resection. How will we know what that small spot in the lung is? We are not going to do a biopsy of that spot in the lung to see if there is still metastasis left. Many times, it’s scar tissue. Many times, it’s a benign lesion sitting in the liver, but we’d have to do a biopsy to figure out if it is really a benign lesion versus metastatic tissue. As long as I see the dominant lesions decreasing in size of more than 90%, I call it complete response.

 I do too. Rana, you brought up some of these other end points that patients are interested in and whatnot, and Neeraj has talked about these durable, near-complete responses. How do you view time off treatment as a goal for patients? That’s not something patients necessarily think about up front, but do you introduce that concept to patients when you’re starting on a frontline therapy? When do you consider a discontinuation of therapy in somebody who is having an ongoing response?

This is a continuing, evolving question in the field. With data from CheckMate 214, the way the trial was designed was that, if you had a significant toxicity while you were getting your nivolumab/ipilimumab, you had to discontinue treatment. This end point has evolved to treatment-free survival, which is basically the time that treatment discontinued to the time of initiation of the subsequent therapy. There are many different ways that a patient can spend their time off therapy. They can spend their time off therapy without any toxicity. They may have some minor toxicity, but that interval matters, and it’s becoming a new end point in the context of the IO/IO combinations, particularly people having to discontinue therapy for toxicity.

One of the biggest questions we get in the clinic is this: “How long am I going to be on this therapy, doctor? Am I going to do this forever?” The answer’s always been, “Well, indefinitely until it stops working or you have toxicity.” That’s always been our answer, but now the benchmark may be around 2 years. Patients are allowed to discontinue treatment at the 2-year mark if they were responding. We don’t yet have data on how those patients do. Time is going to tell what happens, but it’s a big question: can we safely discontinue therapy in responding patients?

To speak to that a bit, we looked at that in the context of a multicenter, phase 2 trial called OMNIVORE; it asked the question of can we stop therapy in responding patients? The trial was small; it enrolled 83 patients. For people who had a confirmed PR [partial response] or CR within 4 to 6 months of starting treatment with nivolumab, they could be discontinued from therapy. From the patients who were allocated to the arm to be discontinued, 42% at 1 year remained off treatment. The flip to that is that 60% weren’t, so how do we appropriately identify those individuals for whom we can safely stop therapy? That’s an area where we critically need a biomarker, a biomarker of extreme responders because some people can do quite well.

Speaking to the durability of response with the data coming out from the updated data from KEYNOTE-426 at ASCO [the American Society of Clinical Oncology annual meeting] and updated data from CheckMate 214 at GU [Genitourinary Cancers Symposium] ASCO, as we look at these progression-free survival curves over time, the nivolumab/ipilimumab combo seems to maintain a steady hazard ratio over time. There is not a big shift in that 0.71, 0.72 hazard ratio, and when we look at the KEYNOTE-426 data, we see a progressively worsening hazard ratio, with the most recent being 0.68 from the updated data with an intend-to-treat population. It doesn't look as good over time with the IO/VEGF, and again speaking to that, are we curing patients? For the durability of the IO/VEGF, time is going to tell what happens to those patients, but treatment-free survival matters.

 Absolutely. This is one of the nice things about the longer follow-up in these studies that we’ll learn. These data are all still relatively new, even 42 months is not that long in the natural history now of some of these patients. There's a longer story to these regimens.

Evaluating Long-Term Follow-Up Data for Treatment of mRCC

 Rana, I wonder if you could comment a bit on what the patient brings to the table. How do you incorporate patient preference into this decision process in terms of goals of care? Neeraj mentioned complete responses [CRs] as an important end point and as a differentiating point from the single-agent TKIs [tyrosine kinase inhibitors]. But is that important in every patient, in every setting? How do you factor in comorbidities and other factors with the patient? What are your thoughts on rounding out this whole discussion with patients?

That’s a key point. With the advent of IO [immune-oncology]/IO and IO/TKI combinations, our end points are changing, and the things that matter are changing. Before, it was largely PFS [progression-free survival] that we looked at. You may get a bit of bang for your buck with OS [overall survival], but it was largely PFS that led to approvals of our VEGF-targeted agents. But we now look at a lot of things. In addition to complete response and durability of response, we're looking at what’s the primary PD [progressive disease] rate? We're looking at what the toxicity profile of a regimen is, what the risks of immune-related AEs [adverse events] are, what the risks of use of high-dose steroids are. Do you have the multidisciplinary team to be able to manage that IO/IO toxicity? What's the toxicity with IO/TKI? What’s the chronicity of TKI therapy toxicity?

All that gets factored in when deciding what the most appropriate regimen is. A patient’s age, fitness level, and goals for therapy matter, clearly. You may have a favorable-risk patient who is young and healthy, and although they're a favorable risk, you still may be swayed to use an IO/IO combination for that individual. You may have a poor-risk patient who may be older and doesn't want to come in and get IV [intravenous] treatment; they live far away from the center, they have other comorbidities, and you’re going to say, “You know what, we're just going to do cabozantinib for you alone from the CABOSUN trial data.”

It's important to understand what the data show us, but there's a lot of gray and art in how we practice in those day-to-day engagements with the patient. You need to understand what you do know, know where the gaps are, and try to come to a therapeutic decision with the patient by interpreting the data and what the patient brings to the table.

Our end points are changing. With the patient populations in all of these trials, all we want to do is sit there and pin them head-to-head against each other and nitpick at every little thing. I'm guilty of that. They're all different. They were accrued at different times, in different parts of the world, with different patient populations, so understanding that is important as you apply it in the clinical setting.

 That’s fantastic. I agree with you, Rana. This is the art of what we do, right? There is what we might think is the best choice, and then there's the realistic choice that the patient can tolerate and still benefit from it. It's great that we have this spectrum of trials that span across here.

Monty, what are your thoughts on the applicability of these clinical trial data? For these clinical trial patients, if you look at them, a lot of our patients in clinic wouldn't necessarily qualify for this trial or that trial based on some eligibility criteria and other factors. How well do the CheckMate 214 data translate into your off-study patient population?

It's tough. The one thing that we’ve become accustomed to doing now in the clinic is screening intensively for any autoimmune conditions. That wasn’t part of my initial dogma in clinical practice, but I struggle with those patients, as one example. It's challenging for me to know what to do with somebody with mild rheumatoid arthritis. Could I potentially challenge them with nivolumab/ipilimumab? I'm tending to favor either TKI monotherapy or TKI plus IO in those cohorts. It's hard to know what to do with patients with poor performance status who might not meet the mark for many of these studies. There I do find myself leaning on nivolumab/ipilimumab to a greater extent, perhaps.

I have to agree with what you said, Dan. It’s challenging to know how to extrapolate from the trial data for some of these unique patients.

 There have been some studies to look at this: some of our nonclear cell population, some of our brain metastases populations, and some of the poor performance status populations. There are small, single-arm analyses, but at least they're giving us a glimmer of hope that there are some patient populations not represented in the original CheckMate 214 study that could still tolerate and potentially benefit from these regimens. It's not going to be the same result, that 11% CR rate, that good duration of response, and other things. The number of patients who get through all 4 doses of ipilimumab may be lower, but do you still see benefit in these populations?

I’ll take a stab at that. There are occasions in which I've been pleasantly surprised by how the nonprotocol patient in my practice responds beautifully to a challenge with systemic treatment, whether it's nivolumab/ipilimumab or cabozantinib in frontline from the CABOSUN trial, but I can definitely cite multiple examples where I've been pleasantly surprised by responses.

 I agree with you. This is one of the key things, if you treat enough kidney cancer patients, then you’re going to get these exceptional responders, and they're not always in the patient populations that you would predict they would be in. That’s one of the things that’s important out there, not to give up on these regimens because of 1 or 2 negative experiences. There will be those, but if you treat enough patients, then you will see these positive results, and they're life-altering results when you see them, in this frontline setting especially.

Patient Factor Consideration for Treatment of mRCC

 Neeraj, any thoughts on how you look at these data now and start to think about applying this in the clinical setting?

Yes. I would, obviously, start with the IMDC [International Metastatic Renal Cell Carcinoma Database Consortium] risk categorization first. Does this patient belong to a favorable-risk category, intermediate-risk, or poor-risk category? That’s the first decision I have to make after evaluating the patient.

Once we have decided that, based on laboratory tests and based on history, the patient belongs to an intermediate-or poor-risk category, then I look for the regimen with which I will have the most curative potential in this given patient. I used high-dose interleukin-2 for a long time until 2015 for the potential of that 10% chance of long-term remissions. If I look at these regimens in the first-line setting, I look at complete responses first, axitinib/pembrolizumab versus sunitinib; I think if you look at the complete responses with the recent update, 9% versus 3%, so twice as much complete response. If I look at ipilimumab/nivolumab versus sunitinib for intermediate-, and poor-risk patient populations, there is a 10 times increase in the complete responses: 1% with sunitinib, 10% with ipilimumab/nivolumab. That’s the big factor in my mind as far as decision-making is concerned.

I then look at the durability of complete responses. If you look at the durability of complete responses, it’s hard to beat the ipilimumab/nivolumab combination at this point. A 42-month follow-up was presented a few weeks ago at ASCO [the American Society of Clinical Oncology annual meeting] 2020, and we are seeing that one-third of patients are maintaining the responses overall. Their complete responses are more likely to be retained; most of the complete responders are still in response. Those are the factors I take into account for poor- and intermediate-risk patients.

For favorable-risk patients, I would admit that it is more challenging because there are favorable risk, and there are favorable risk. There's a patient with 1 lung nodule that is growing for the last 5 years. We have been monitoring it, and it is not changing at all. I will not do anything; I will go for active surveillance because this patient may be harmed by what I offer in terms of medications. Then, there are some patients who are again in the favorable-risk category, 5 different sites of metastasis, but their labs are completely normal and disease burden is quite high. In these patients, I would think about starting systemic therapy.

Depending upon what the disease volume is, I’d base my choice on the disease volume in the favorable-risk category. For low disease volume, it’s active surveillance. For just 1 or 2 sites of metastasis that are slowly growing and the patient is otherwise doing well, it’s single-agent VEGF/TKI [tyrosine kinase inhibitor]. I’d pick one of the VEGF/TKI options. VEGF/TKI can include pazopanib and sunitinib, but I then look at the CABOSUN trial. Cabozantinib was superior to sunitinib, so I don't hesitate picking up cabozantinib as a VEGF/TKI of choice for these patients. For patients who have more rapidly progressing disease, still in the favorable-risk category, I tend to pick up axitinib/pembrolizumab, for example, in these patients. Thinking about how I practice in my clinic, this is what I have to say.

 Interesting. That’s a good summary, Neeraj. What I like about that is it's thoughtful; you use the whole armamentarium. There's a place for each of these approaches in frontline patients: it's not one-size-fits-all. A big part of that is the risk stratification that you mentioned.

Applying Clinical Trial Data for Treatment of mRCC

 Monty, you can summarize for us some of the trials that define the National Comprehensive Cancer Network [NCCN] guidelines and some of the other guidelines we have for frontline therapy that have changed in the last 2 to 3 years to define both the TKI [tyrosine-kinase inhibitor] space in frontline therapy as monotherapy, as well as now the combinations IO [immune-oncology]/IO and IO/TKI. Walk us through the top-level studies and data that led to those approvals in the frontline space.

I'm going to try to do this in 2 minutes or less. It all started with CheckMate 214, didn’t it, which was the study of nivolumab/ipilimumab versus sunitinib in the frontline setting. We were all amazed with the initial data suggesting a profound benefit in progression-free survival, response rate, and overall survival, particularly in those individuals with intermediate- or poor-risk disease. Still, there is something to be said for how to manage good-risk patients based on that study. We can certainly dive into the details of that later.

The next study to report out is one that we don't talk that much about these days, but it has increasing relevance in other diseases; this is the trial with bevacizumab with atezolizumab versus sunitinib.This study demonstrated an improvement in progression-free survival with bevacizumab/atezolizumab versus sunitinib, but it fell short of demonstrating a benefit in overall survival. It's a little different from the data that we're seeing emerge with bevacizumab/atezolizumab in the context of lung cancer, recently with Impower150 in hepatocellular carcinoma, for instance. The topic of discussion today is going to be how to juxtapose the nivolumab/ipilimumab regimen against 2 other regimens, one being axitinib and pembrolizumab, looked at in the KEYNOTE-426 study against sunitinib. There you see a benefit in overall survival in the overarching cohort and certainly a benefit in response rate and progression-free survival across a stratum. The axitinib/avelumab study, or JAVELIN Renal 101, was a trial that looked at that regimen against sunitinib once again. That study showed a benefit in progression-free survival across a stratum, but it failed to show an improvement, at least to date, in overall survival.

How was that, Dan? Does that give you the gist of the studies?

 That’s great. The only study you missed was CABOSUN, and I'm not going to take offense. It was our Alliance study; we were one of the first studies to look at the intermediate- and poor-risk populations and show a difference in TKIs, that maybe not all TKIs are the same and that cabozantinib showed some progression-free survival superiority over sunitinib in that frontline setting.I'll give you an A- overall; it was a great segment.

That was one of my favorite trials, I can't believe I forgot it. Please don't kick me off the program.

 I won't. You’re forgiven, you’re absolutely forgiven. It's difficult because these trials are all done fairly contemporaneously with each other, so we don't have comparisons head-to-head, right? We have everything compared to sunitinib or VEGF-targeted therapy. We're seeing now superiority with these combinations over monotherapy, but the question comes up: are they all showing the same benefit, or are there subpopulations for which one approach may be a bit more preferable to another approach?

 Clinical Trials Defining the NCCN Guidelines for Frontline Treatment of mRCC 

 “Advanced Renal Cell Carcinoma: Putting Advances Into Practice.”I'm Dr Dan George, from the Duke Cancer Institute in Durham, North Carolina. Joining me today in this discussion are my colleagues, Dr Neeraj Agarwal, from the Huntsman Cancer Institute at the University of Utah, in Salt Lake City; Dr. Rana McKay from UCSD [The University of California San Diego] in San Diego, California; and Dr Sumanta Pal, from the City of Hope Comprehensive Cancer Center in Duarte, California.

Today, we're going to discuss a number of topics pertaining to the use of systemic therapy for advanced metastatic clear cell renal cell carcinoma [RCC]. We’ll discuss the latest research in the field, and the impact of recent clinical trials on making decisions around treatment selection. Let's get started with our first subject.

Let me start first with Rana, and we must start thinking about this in terms of frontline clinical trial data in advanced renal cell carcinoma. Before we make a treatment decision on a patient in the frontline setting, we usually start with some kind of risk assessment. Now, Rana, there have been a number of risk stratification models out in metastatic renal cell carcinoma; how relevant are they in your everyday practice for patients, how do you use them, and what do you use?

The risk stratification models are significant in this day and age, particularly the IMDC [International Metastatic RCC Database Consortium] risk factors that integrate laboratory values, performance status, and time to systemic therapy from diagnosis, so they're highly relevant. The model has been integrated into our frontline trials: KEYNOTE-426, CheckMate 214, to help us risk stratify patients to understand outcomes for different subsets of patients. The IMDC were initially validated in the context of the VEGF targeted therapy era. Predating that were the Motzer criteria, the Memorial Sloan-Kettering Cancer Center [MSKCC] criteria, that were validated in the cytokine era. Now in the context of these risk stratification rules being integrated into our frontline trials, they are helpful in deciding how to select therapy for any given patient.

There's been some elegant work that’s been conducted by David McDermott, MD, and his team from the IMmotion studies, speaking to the biology of each in the different risk groupings. The favorable-risk patients seem to have a more angiogenesis-high signature, and that seemed to have panned out in the frontline IO [immune-oncology]/VEGF trials, where there seems to be a signal that those patients may benefit from VEGF inhibition. Whereas, those who have intermediate- and poor-risk disease seem to have a more upregulated immune signature. The application of these risk stratifications is important. They’re critically important in not just determining systemic therapy but also in thinking about the role of cytoreductive nephrectomy in the context of people who have advanced RCC. Given the number of risk factors that any one individual may have, deciding for up-front nephrectomy or delayed nephrectomy may be important, so these are critically important in this day and age.

 Good, I agree. Monty, expand on it a little bit for me. We use IMDC probably more than some other models, but do you use that in every patient? How do you use it? Are there certain variables within IMDC that you look for, particularly to help you think about if there is an underlying biology to this? What are your thoughts on these models in your practice?

That’s a great question, Dan. When we talk with our colleagues in the community, they sometimes suggest that they might be using a gestalt impression of whether a patient is good risk, intermediate risk, or poor risk. As time goes on, Rana spoke to this beautifully: now that we understand the biological underpinnings of what good, intermediate, and poor is, I am sticking to the criteria more directly. I have to say, it's been a long time since I've not staged a patient with IMDC risk characteristics. I tend to be a bit of a purist there nowadays.

 That’s great. Neeraj, can you comment on this as well? When I see a patient who has a good performance status, I tend to think of them as a good-risk patient. However, when you start to look at some of the laboratory test results, and you start to look at the burden of disease and some of the timing of that metastasis from diagnosis, you start to add up these criteria. You can quickly realize that this patient may be a bit higher risk than you originally thought. Getting past this visual assessment and being more quantitative in this is extremely helpful. I've been burned by being fooled. The patient looks pretty good in the exam room, but then when you get the labs back the next day, you’re saying, “Wow, there's more going on here than I thought.”

Neeraj, what's your take on IMDC, and how do you think of this in terms of individual factors? Is there any one factor in IMDC that weighs more heavily on you than another?

Like Rana, you, and Monty said, I use it for pretty much all my patients. It’s a standard as a part of my initial visit now that IMDC will be mentioned, and without that, my note is incomplete. We are looking at all these factors in all patients. I’ve been surprised to see how patients can be categorized into poor-risk versus favorable-risk once we start counting the factors.

Which factor sways me the most: time from original localized kidney cancer to time of nephrectomy to onset of metastatic disease. That’s one factor I take into account and keep in mind. If a patient is developing a metastatic disease within 4 months or 5 months from nephrectomy, that’s a bad sign. If I see someone who is coming to me 8 years after original nephrectomy, it's a different disease.

 That’s a great point. It’s the only factor that speaks to the natural history of the disease; everything else is pretty immediate or static, and it's the one that is undeniably cancer. For anemia, there are a lot of factors that affect anemia. Or even hypercalcemia, it could be due to dietary things or sedentary factors or whatnot. Performance status is definitely affected by a lot of factors in the patient. That natural history of disease is so specific and relevant in kidney cancer, particularly in our good-risk patients, almost to the very good-risk patients. 

I want to get back to something that Rana brought up. She brought up the angiogenic profile that’s been looked at and correlated to some of our better-risk patients. It’s fascinating because we’ve known from the beginning that the loss of VHL [Von Hippel-Lindau disease] is a common biologic, angiogenic alteration that accompanies clear cell carcinoma frequently early on in this disease, and it drives that angiogenic profile. Everything else is secondary to that. Those inflammatory factors and other things get piled on and built onto that as the tumor burden increases and the genetic complexity diversifies, but it starts as an angiogenic biology. The tumors that retain that are pretty dominant on that. They tend to be the good-risk patients, but there are probably some of our intermediate- and poor-risk patients who still have an angiogenic profile; they just have other factors with it as well. We’re going to think about this now as we start asking questions about combination therapy.

Daniel J. George, MD, Duke Cancer Center

Articles