Toni K. Choueiri, MD

 Toni, you’re the leader of the COSMIC-313 study and also the belzutifan studies. Do you want to speak to those trials in terms of where they stand and what people should be watching for and anticipating with those studies? 

Thank you, Bob. COSMIC-313 is a study that asked the question, “Is 3 superior to 2?” We went through this lovely talk from 1 to 2, to perhaps a 3, where cabozantinib is going to be added to nivolumab and ipilimumab vs controlled nivolumab and ipilimumab and modern control finally, not sunitinib. The study just finished. Accrual about to wrap up was primary end point PFS [progression-free survival], secondary end point OS [overall survival] with over 800 patients. In intermediate and poor risk, it’s important to see if we can push survival more and push the CRs [complete responses] more. 

The second study you alluded to is a study with a totally new agent. Here, we’re not adding drugs that are already approved as single agent or combo. Belzutifan is an HIF2 inhibitor, and Tom in a very eloquent way discussed the biology of clear-cell RCC [renal cell carcinoma] initially. HIF2 is a transcription factor thought for a long time to be completely undruggable. HIF2 doesn’t have those small pockets where the drugs could go until recently, with elegant crystallographic work showing it could. We could have molecules and the second-generation molecule MK 6482 belzutifan showing a response rate and a refractory population of clear cell RCC of more than 50 patients. It’s mostly at 25% with quite tolerable adverse effects. You don’t have diarrhea, hypertension, and anemia, which is on target because you target downstream HIF2. 

Down regulating actually could be a pharmacodynamic biomarker with these drugs. Hypoxia was another, which is likely on target. Toxicity was HIF2 physiology. We’re taking this to a phase 3 trial post-VEGF inhibitor plus PD-1 inhibitor, belzutifan against everolimus later lines. And belzutifan is being introduced first line and second line with other studies that are being planned. The 1 thing about this treatment, as Tom mentioned, is that the time to respond seems to be not 2 to 3 months with single-agent TKI or less than 2 months with the combos, but it takes time—around 6, 6½ months with it. It doesn’t work quickly if you want an average. But it’s a nice set addition, and I hope this would be there. Data also in VHL [von Hippel–Lindau] syndrome associated clear-cell RCC, where we saw some image of blastoma is shrinking. Just yesterday or the day before, there was an FDA note about. VHL syndrome is scheduled, I believe in September. This is all good news for the renal cell cancer field.

This transcript has been edited for clarity.

Videos

Toni Choueiri, MD, closes with promising new studies that will impact metastatic clear-cell renal cell carcinoma therapeutics in the years to come. 

Promising Studies for the Future of Frontline mRCC

 I use both in my practice. I try to keep nivolumab-ipilimumab in the approved indication: intermediate and poor. I don’t use it in good risk. I don’t think it’s a mistake, because of the CR [complete response] rate, but I do not. That’s 1. In patients with sarcomatoid histology, it is the combo that should be the best standard because of the complete responses of 19% from CheckMate 214. That’s the second 1. The third 1, in patient with tumors that are not progressing rapidly, it was significant disease volume. Although we’ve seen responses with nivolumab-ipilimumab, I lean toward VEGF–TKI [tyrosine kinase inhibitor] rather than nivolumab-ipilimumab.

That’s an area where you’re going to get different opinions from all our friends and colleagues. I still think there’s an advantage. Especially with CRh [CR with partial hematologic recovery] and I/O [immuno-oncology] TKIs, the degree of shrinkage and the clinical benefit rates are excellent. You can always carve out patients that could be treated with just a single-agent TKI. I do have some of these patients in my practice, but the standard should be the default, should be I/O–TKI vs the other way around. Go into a favorable-risk situation if you made the decision to treat and not just monitor. The majority of people will make the decision to go forward with treatment that an I/O–TKI should be your default. Then you can carve out from them a smaller group that may benefit with just the TKI.

For instance, patients I just saw 2 types of patients today in clinic. One was an organ transplant patient. Obviously with a favorable-risk patient, you’re going to avoid your checkpoint inhibitor. Giving a single-agent TKI is reasonable. Another group that I’m leaning on, that we have the most data in the setting, is patients whose disease is predominantly in the remaining kidney. They’ve had a renal cell. They’ve lost a kidney from that. They have 1 kidney remaining, and it has tumors. And you’re channeling them like a patient with VHL [von Hippel–Lindau], and you’re trying to be careful and preserve kidney function. But sometimes we end up having to put them on a systemic therapy with the hope of making tumors they have smaller to allow a percutaneous intervention or just to stabilize things.

I’ve had patients on a single-agent VEGF TKI, and most of us are aware of the data coming out of Houston, Texas, in the past where that was somewhat instrumental. They’ve done a lot of that work on looking at the degree of shrinkage in the tumor from the TKIs. I’m sure I/O–TKIs do that at some level too. I’m just not as familiar with that degree of shrinkage. I’ll have those patients on a single agent. But that’s the minority. The default for me, mostly is I/O–TKI. If a select group seems to be ideally suited for just a single-agent VEGF inhibitor, then I would consider using traditional sunitinib or pazopanib.

I see a role for that, especially for the follow-up, and we understand the CLEAR data better. Right now, the agent with the longest follow-up and the most attractive 1 is the durability of the response that exists with the I/O–I/O. What drives that, the ipilimumab or not, no one knows. There are triplet studies planned that add ipilimumab into an I/O–TKI regimen. It will be interesting to see the benefits. I completely agree with Toni. I usually agree with Toni on most things. But the way he unpacked that is a reasonable approach, where you are using both. You can look at the way that the patient is presenting. Do they have time? The difference is the I/O–I/O combination is going to work in less people.

The clinical benefit rate is less. Do you have a patient who can afford to have I/O–I/O not work? As the volume of disease sets, it would allow that. If you have someone on the other end of the spectrum, where you really need to have an insured a rapid response, we see the data, and we didn’t talk about it at the time to response for lenvatinib-pembrolizumab. The cabozantinib-nivolumab is probably pretty quick. You can get these really symptomatic patients under control quickly.

The other thing that we also didn’t mention is expanding understanding about stopping it. How do you go about stopping an I/O–TKI vs the I/O–I/O? Some people would say it’s a disadvantage to be on an I/O–TKI and try to stop. Other people would say it’s an advantage because you could go off the I/O or the TKI and do it that way. There’s a lot that we have to understand about these drugs. If I was predicting that CLEAR study on the CheckMate 9ER trials to continue to evolve, where there’s evidence of durability of response, the more durable those responses are, the less I’d be inclined to use an I/O–I/O in those setting. But we don’t know that yet.

Both nivolumab + ipilimumab and lenvatinib + pembrolizumab TKI/IO combinations are discussed, with patient selection and appropriate use being the focus. 

Using TKI/IO Combinations in Patients With mRCC 

 Tom, you’ve used lenvatinib for over a decade and are probably 1 of the heaviest users of lenvatinib in the United States. What has been your take on dosing of the lenvatinib and adverse effects within the class that you see as favorable or unfavorable for lenvatinib in combination, with everolimus and pembrolizumab?

 You all hit on things correctly. Diarrhea, hypertension, and fatigue are key ones. I’ve never run into a situation, thank goodness, where any type of renal proteinuria has resulted in an impairment and kidney function that impacted the ability of the patient to receive other therapy. Mind you, many of our patients have impaired kidney function with 1 kidney, and we insult them with IV [intravenous] contrast and other things that can precipitate worse things. 

As far as tolerability goes, some of the newer, more active agents have stepped up a little the degree of adverse effects that patients experience, whether that’s with the cabozantinib-type drug as a single agent or in combination. Same with lenvatinib-everolimus or lenvatinib-pembrolizumab. We go into those therapies starting with an idea that over 50% of patients require dose reductions. I’ve found that instructing the patient at the beginning and painting a picture for them that as a field, we try to start at the highest dose possible in most situations to get them to that first scan, to see maximum benefit. The expectation is that more than half of people will require fine-tuning of the dose or a dose reduction. 

To start with the lenvatinib, I understand that there may be a formulation difference in the tablets. But for the lenvatinib-everolimus combination, when you come with the blister pack, you have dose levels. It becomes a very easy way to lower from 18 to 14 and then 10. We’ll see how that gets packaged out there, but it’s ease of use from that perspective. It’s something I’ve noted in my patients. If you go into therapy with the right mindset, trying to optimize the efficacy by maintaining dose intensity but understanding and making sure the patient understands that if they are experiencing an adverse effect, to bring it to our attention. That is expected in some regard.

In the clinical trials, half of patients did require dose reduction. And as Toni mentioned, the discontinuation rate by doing so is low enough that it makes you feel that, “Hey, if someone runs into adverse effects, I can dose reduce. I can handle it with supportive care to the point that the majority of patients are able to stay on therapy.” The frequency of this happening in the clinical trials is high, and patients still get a benefit. You have a lot of evidence that when you talk to your patient, they shouldn’t be afraid. Sometimes patients are if lowering the dose means they’re not going to respond. You have evidence to say that that’s not the case. We need to make the therapy tolerable to you so you can stay on it for as long as possible. We shouldn’t worry as much that it’s somehow going to get a lower efficacy. It’s a doable regimen.

To be honest, 1 of the drivers is something we haven’t mentioned: the experience of the prescriber. The newer drugs aren’t necessarily out there as long. Many of the prescribers aren’t as familiar with them. As we learn with sunitinib, sorafenib, and pazopanib that some of the adverse effects that were most troubling at the beginning of the approvals when they were first used have become less and less. We see adverse-effect profiles reduce as we’ve done later trials with these agents as comparators. There’s a learning curve when it comes to these agents. But the lenvatinib, although there are toxicity concerns, isn’t so far out there that it couldn’t be used in our patients.

Dosing strategies to achieve strong efficacy and ameliorate adverse events with the combination of lenvatinib + pembrolizumab in metastatic clear-cell renal cell carcinoma.

Dosing Strategies for Maximum Efficacy With Len + Pembro

 In terms of toxicity overall and adverse events, there are TKI [tyrosine kinase inhibitor]–driven toxicities but also some immune-related adverse events. You have to use some of the art because some toxicities are overlapping between I/O [immuno-oncology] and TKI. Nothing really much different that diarrhea and liver function test elevation could be due to pembrolizumab or lenvatinib, while hypertension is due to lenvatinib. Treatment-related AEs [adverse events] were higher in all these trials of VEGF–I/O. It’s hard to compare them because the reporting is different.

I look at the treatment-related AEs leading to discontinuation of either drug or both drugs. Overall, these tend to be less than 10% for both drugs. The question that is important is the doses, which are different from the TKI. You start at 20 mg when you combine with pembrolizumab. When you combine with everolimus, it’s 18 mg. When we combine cabozantinib with nivolumab, it’s 40 mg. Which dose is right? Is it the pembrolizumab combination with lenvatinib—with lenvatinib at 20 mg—the right dose because a significant number of patients needed dose reduction?

In my book, I start at the FDA and what the trial did and I go down. It definitely needs folks who are well versed in managing the adverse effects of I/O but also the TKI, because with the TKI it’s always about the exposure. It’s not just the dose but the exposure. There’s a lot of trial and error. Today, 15 years after sunitinib, we don’t know what the right dose of sunitinib. We haven’t solved that. So we have to be very careful because the goal is to keep the patients on the regimen for the longest period of time—let’s say first line with a manageable adverse-effect profile and an acceptable quality of life. That’s the goal.

 In your own experience, for people on this program, what adverse events for the community oncologists did you find to be prominent? People have talked about the proteinuria as being a standout with lenvatinib.

 I have seen the typical TKI overall adverse effects: diarrhea, hypertension, stomatitis. I agree with you. I’ve seen also dysphonia, which I’ve also seen with axitinib. I have seen proteinuria to be higher than I would expect with sunitinib or cabozantinib, for example. I have followed the protocol because of our experience from the CLEAR trial on what to do. But I would argue—maybe not first line for proteinuria, which is what I would do, but for later line—that the proteinuria in use by VEGF-TKI is not the same proteinuria from chronic diabetes, etc. Although there are some reports with bevacizumab of proteinuria in a setting if you want of TTP [thrombotic thrombocytopenic purpura] and HUS [hemolytic uremic syndrome], which is a totally different ballgame. But I’ve seen a lot of patients on TKI in general at the later line who are responding and have tolerated 3 or 4 g without major effects on the kidney function or others.

So it’s manageable. We probably need to be a bit more careful if it’s a first line, because first-line patients’ median survival is quite prolonged, luckily. But I would not have a knee-jerk reflex to completely stop the drug and move on if there is a gram of protein. I would make sure first to do a 24-hour and work with the nephrosis. Sometimes, because their blood pressure is hanging in the 160-to-170-mmHg range, they’re not appropriately treated or diabetes is not well controlled. You have several contributing factors. That’s my approach.

 The hypertension occurs early with this combination. Physicians need to be particularly vigilant about hypertension early. But from my own experience, the trial results reflected this very low incidence of liver toxicity with lenvatinib and also with lenvatinib and pembrolizumab, as well as the lack of high-grade hand-foot syndrome. I rarely see that with this combination of myelosuppression. Some of the standouts in terms of the favorable parts of this combination were those particular aspects of the toxicity.

Hutson and Choueiri compare the safety profiles of lenvatinib + pembrolizumab against other TKI/IO combinations for patients with metastatic clear-cell renal cell carcinoma.

Safety Profile of Len + Pembro From CLEAR in mRCC 

This was following the lead of some of the earlier I/O [immuno-oncology]–TKI [tyrosine kinase inhibitor] trials. It was an international effort, combining lenvatinib and an agent was already approved when in combination with everolimus, for differentiated thyroid cancer, and for gynecologic malignancies—like uterine cancer—when combined with pembrolizumab. That combination of lenvatinib-pembrolizumab vs lenvatinib-everolimus has already been approved for refractory RCC [renal cell carcinoma] against sunitinib.

It was a 3-arm trial, but the statistics were designed to compare not lenvatinib-pembrolizumab vs lenvatinib but lenvatinib-pembrolizumab and lenvatinib vs sunitinib individually. There were roughly 350 patients per arm. We’re going to focus on just the lenvatinib-pembrolizumab, the I/O–TKI part of that compared with sunitinib, which was the gold standard. At the first analysis that had been presented, we saw that with 27 months’ follow-up the median overall survival between lenvatinib-pembrolizumab and sunitinib had not been reached but had a hazard ratio of 0.66. We did a landmark overall survival at 12 months and 24 months. They were 90% and 79%, respectively. Median progression-free survival was 23.9 months with the combination lenvatinib-pembrolizumab vs 9.2 months. The hazard ratio there was 0.39. Look at the overall response rate: 71% with lenvatinib-pembrolizumab vs 36% with sunitinib. I like to think about clinical benefit rate, which adds that stable disease factor or the generalizable ability of the drug so that if you take the drug and give it to someone, it’s going to at least produce stable disease. That was sitting above 90% and was very similar to what Toni just mentioned with cabozantinib-nivolumab, a 90% clinical benefit rate.

We can’t really do fair cross-trial comparisons, yet we always do that. It’s in clear cell carcinoma; there are no crossovers. We have a little difference in outcome, and the numbers I just said with lenvatinib-pembrolizumab make it—at least by number—1 of the most active if not the most active I/O–TKI. But we must understand that there’s a follow-up difference among the trials. The trials have different percentages of the prognostic groups in them. The lenvatinib-pembrolizumab trial was heavily weighted on the favorable-risk group at the expense of not having as many of the poor-risk patients in it. The pembrolizumab was traditional dosing, but the lenvatinib was at 20 mg orally daily, which was higher than what traditionally has been used with lenvatinib-everolimus, which is 18 mg.

There are some nuances there that make it unclear: Is this really that much better than the other agents, or can that be explained by other factors? Nonetheless, we would all agree that based on the data presented with lenvatinib-pembrolizumab, the efficacy numbers put it at the higher end of what we see with I/O–TKIs.

The question from there comes into toxicity. We have not reported the health-related quality-of-life outcomes that were reported with CheckMate 9ER. We see that there are adverse effects with the combination of it in that patients do require dose interruption and dose reduction. You had mentioned correctly that it’s generally the TKI that drives that. It’s hard to differentiate among the regimens on tolerability, but when you have similarly active agents, tolerability is going to come into play as the ease of administering the agents will come into play. At the end of the day, the CLEAR study showed us what we were hoping to see, which was that the combination was doable and could be administered. Adverse effects, although they’re present, could be managed. We see efficacy levels of the combination that put it at a very high level as 1 of the most active regimens.



 One of the most interesting aspects of this trial was that it was a 3-arm trial and did have the third arm as lenvatinib-everolimus. The other trials have compared I/O combinations with sunitinib, but they have used a later-generation tyrosine kinase inhibitor that may be better than sunitinib. One of the interesting aspects of this was the lenvatinib-everolimus. What I found interesting was that both lenvatinib-pembrolizumab and lenvatinib-everolimus met their primary objectives by showing longer progression-free survival [PFS]. They also showed higher response rates, but the margin of benefit was longer or better with lenvatinib-pembrolizumab compared to the lenvatinib-everolimus. We saw a survival benefit with lenvatinib-pembrolizumab but not with lenvatinib-everolimus. I interpret that there is an important role for I/O therapy in first-line therapy, that the I/O therapy helps drive the efficacy and result in a longer survival. I found the lenvatinib-pembrolizumab efficacy data particularly impressive with regard to the metrics you mentioned and also a high complete response [CR] rate of 16%. The objective response rate was over 70%. It looks to me like a very effective regimen.



 I am quite happy that there’s another VEGF–I/O combination to make it here. The compelling thing about lenvatinib-pembrolizumab, as Tom eloquently said, is that we shouldn’t compare across trial. Baseline characteristics are a bit different, with lenvatinib-pembrolizumab having a bit more favorable patients. But I don’t know if that really explains the longest PFS today, which was close to 2 years. And the CR—not just the overall response rate of over 70%, but the CR rate of 16%. These were really quite compelling and intriguing. I look forward very much to the next update from CLEAR to see what are the PFS and CR rates and if they’re going to go up. These are the strongest data with lenvatinib-pembrolizumab, knowing that if you compare it—and you shouldn’t—with cabozantinib-nivolumab, there are more patients with favorable risk and more patients with kidney removal, which we know could influence the PFS and the CRs rate.




Thomas Hutson, DO, PharmD, and Toni Choueiri, MD, talk about the phase 3 CLEAR trial at length, including each of the 3 arms of the trial and key data. 

CLEAR Trial: Len + Pembro or Everolimus vs Sunitinib 

 Yeah, absolutely. It sealed the role of checkpoint inhibitors as a primary therapy for the disease. We know that it’s coming after the approval…of combination immune agents, like the ipilimumab-nivolumab regimen that you mentioned before. But we had wondered what would be the role of a TKI [tyrosine kinase inhibitor]. Toni just mentioned all the preclinical and basic science research suggesting that certain TKIs have immunomodulating properties, plus the underlying basic biology that we understood of the cancer, that continue to inhibit VEGF are going to be important if the mutation that results in the majority of is VHL related. This was a major trial to show us that this combination of a TKI plus an I/O [immuno-oncology] is going to have additive benefits that are going to push up response rates. From the original report, we had 9% complete response rates. We had overall tumor response rates.… We saw a median progression-free survival increase to 15.4 months when it was compared with sunitinib as a control. We saw for the first time that combining an I/O and a TKI together produced a benefit for us. That has entered the frontline setting as a viable therapeutic option. Unlike some of the concerns, and we could debate whether immune approaches have a role in the favorable-risk person, the I/O-TKI seem to have shown benefit across all risk groups. That was also something that we gleaned from this trial.

In both studies, the pembrolizumab-axitinib and avelumab-axitinib started at the same time from phase 1 and went through phase 3. The differentiator was the avelumab axitinib. First, avelumab is a PD-L1, not a PD-1 inhibitor. That’s the first thing. Both studies went from phase 1 in untreated patients, who showed significant but more direct activity, to phase 3. With avelumab, the activity from phase 1 in clear-cell RCC [renal cell carcinoma] was shown in terms of response rate to be higher in PD-L1 positive. Unlike KEYNOTE-426, the end point of PFS [progression-free survival] and OS [overall survival]—2 co-primary end points—were in PD-L1–positive population, which ended up in phase 3 being over 60% of patients. 

You asked about PD-L1 as biomarker. Each drug has its own PD-L1 definition. Avelumab was thought to be a bit of a differentiator compared with others because it’s unique and can mediate ADCC [antibody-dependent cell-mediated cytotoxicity] because it retained this native FC region. Overall, the study met its primary end point of PFS. And it ended up that PD-L1—positive or not—didn’t make a difference. Response rate also was doubled. The differentiator, unfortunately, is that whether in the PD-L1 positive or in the ITT [intent to treat], the combo of avelumab-axitinib still does not have an overall survival advantage over sunitinib. The final analysis did not happen yet. Therefore, despite the FDA approval of this TKI-I/O combination, it does not have an OS advantage.

This brings memories from the phase 1 of cabozantinib a long time ago. It’s been a journey, and I’m happy that now it’s the first line, 1 of the standards with nivolumab. We learned that this could be from another phase 1—by Dr Andrea Apolo at the NCI [National Cancer Institute]—that they could be combined. That’s the first thing. We settled on a 40-mg dose of cabozantinib, which is not the FDA starting dose of 60 mg with monotherapy. The study had a primary end point of PFS, and it did meet the primary end point with a hazard ratio of 0.5. Also, we had an OS advantage with a 40% decrease in the risk of death. The responses were doubled. The CRs [complete responses] were higher, and there were quite interesting quality-of-life data. One of the interesting things we insisted on initially was the first disclosure of the result That was at ESMO [European Society for Medical Oncology Congress] 2020… We insisted on having some quality of life. Patient-reported outcomes are the voice of the patient. We introduced 2 metrics, the FKSI-19 [revised Kidney Symptom Index] and DRS [disease-related symptoms]. Those supported the combination of cabozantinib-nivolumab over sunitinib. More recently, there was way more quality of life presented at ASCO GU [American Society of Clinical Oncology Genitourinary Cancers Symposium], going in the same direction that cabozantinib-nivolumab is superior to sunitinib. All the end point that we had, to my knowledge, were met.

Hutson and Choueiri elaborate on the FDA-approved TKI + ICI combinations and their data from the following trials: KEYNOTE-146, JAVELIN Renal 101, and CheckMate-9ER.

Toni K. Choueiri, MD | Authors

Articles

Promising Studies for the Future of Frontline mRCC

Using TKI/IO Combinations in Patients With mRCC

Dosing Strategies for Maximum Efficacy With Len + Pembro

Safety Profile of Len + Pembro From CLEAR in mRCC

CLEAR Trial: Len + Pembro or Everolimus vs Sunitinib

FDA-Approved TKI/IO Combinations in mRCC