Risk stratification and selection of first-line therapy for metastatic renal cell carcinoma (mRCC)

PROF POWLES: The series of things that go through my head when I see these patients is, I normally get their case reports first. And the first thing I think about is what histology do they have, number 1. And then number 2 is, where are their sites of disease? And those two are the first 2 thoughts, and clearly we’re going to talk predominantly about clear cell, and I think I’m going to talk pretty much exclusively about that, unless asked otherwise. I assume you’re talking about clear cell.

And the second piece is, where is the disease? Two questions: Number 1 is, have they had an operation previously if they have advanced or metastatic disease, which I assume we’re talking about again today. Have they had an operation previously? Have they relapsed after a nephrectomy? Yes or no. And then the second key question is, where are their sites of disease? And so do they have metastasis in the visceral wings to the lungs? Is it just lymph node disease?

I worry about brain, liver and bone, and I tend to put those together. And then lung metastasis and lymph node metastasis tend to worry me less.

Histology first, then sites of disease and disease burden, whether or not they’ve had a nephrectomy, yes or no? And then the last piece is then their IMDC score. What risk group are they in? Good, intermediate and poor.

And I like to have that kind of information before we see the patient and go in the detail of what their past medical history is like and how fit they are and what their expectations are and what they would like.

Actually, it’s relatively straightforward. We don’t have that many options. Histology. Nephrectomy? Yes or no? Sites and burden of disease, and then their risk score.

Role of cytoreductive nephrectomy in RCC

DR LOVE: Toni, where do we stand today on the age-old question of cytoreductive nephrectomy?

DR CHOUEIRI: I mean, Neil, this is a good question. And I think at least at our center, we have not really changed much what we do, even based on the CARMENA study, that pivotal study that showed that cytoreductive nephrectomy, at least when you get sunitinib, may not yield an overall survival compared to not getting cytoreductive nephrectomy, because we really did not include patients with poor risk or a patient with poor performance status. We still do it in the appropriate patient. But certainly I do believe that it was done more than it should be done in some situations.

I think it’s still an important part to remember that cytoreductive nephrectomy, especially in these patients that have limited disease burden. And a lot of them could be followed sometimes without therapy or could be rendered disease free, especially if they can get into surgery.

The bigger question is, sunitinib is not a standard today, and should we repeat the nephrectomy, the cytoreductive nephrectomy question with a more potent and modern therapy, such as the combination of pembrolizumab and axitinib or the combination of nivolumab and ipilimumab? And I think people are split here, and you will see some that say the style will not happen. And even if it happens, although there is, I would say, a SWOG proposal from one of the cooperative groups, folks are saying even if it is to happen, you have drugs that are more potent than sunitinib. The premise that cytoreductive nephrectomy can help a drug that is less potent won’t exist here, so why do the trial?

I think selection, selection, selection is important here. And it brings me back to a multidisciplinary care, where if you have metastatic disease and your primary is still in place, it’s very important to have your clinic next to a urologist and discuss this in a multidisciplinary fashion, Neil.

Overall survival and complete response (CR) improvements with the FDA-approved nivolumab/ipilimumab combination versus sunitinib for intermediate- or poor-risk previously untreated advanced RCC on the Phase III CheckMate 214 trial

DR LOVE: Tom, can you talk a little bit about some of the new data sets that have come out in the first-line setting? Toni mentioned pembrolizumab/axitinib, also avelumab and axitinib. And, of course, we had ipi/nivo before that in terms of some of the newer data sets that now are affecting first-line decisions.

Tom, can you kind of go through those and what your perspective is about where you utilize these strategies or where you would like to utilize them?
 
PROF POWLES: Yes. Neil, that’s great, and it’s become complicated, because, actually, there are 4 combinations where we have information on randomized Phase III data. The first is ipi/nivo, which is a PD-1 inhibitor, nivolumab plus CTLA-4 inhibitor, which is ipilimumab, and that’s a 1,000-patient randomized Phase III study. That has a label in both Europe and the United States, with Level 1 evidence in intermediate-and poor-risk patients. There’s no label in good-risk patients, because in that randomized Phase III study, the ipi/nivo, the immune therapy was outperformed by sunitinib. And as I see it, there’s a debate going on in that place, but that debate, as it currently stands, without that label, I think we should only be selecting ipi/nivo for intermediate- and poor-risk patients.

This trial has an overall survival, a hazard ratio of 0.63 in this population initially, so a big overall survival advantage for intermediate- and poor-risk patients. We had long-term durable responders, and we had better quality of life than we had with sunitinib.

An excellent combination. Patients living longer and feeling better with durable remissions.

DR LOVE: Before you go on, I just want to get Toni’s take on this. Any thoughts about why there wasn’t a benefit seen with the good-risk patients? Biologically it’s kind of hard for me to figure that out. What’s your take on that, Toni? If you could, would you use the drug in good-risk patients?

DR CHOUEIRI: Yes. Tom, I agree with everything Tom said, in one of those situations where Europeans and Americans do agree. Whatever he said, I’m in complete agreement.

PROF POWLES: I agree with you.

DR CHOUEIRI: I want to supplement, especially in the Brexit era, where we do agree with you, Tom. I want to supplement that, because Tom said something extremely important, and I want to make sure this is captured. Looking at overall survival, which, of course, has been seen in the pembrolizumab/axitinib and nivolumab/ipilimumab, is a great thing — it’s the gold standard in oncology. But one thing about the nivolumab/ipilimumab, which the combination of axitinib and pembrolizumab or axitinib and avelumab doesn’t have yet, is the rate of complete response that is over 10%. And let’s remember that this is a combination where you’re done with ipi and you continue nivolumab maintenance.

This is a good endpoint to look at, could be equivalent — and I don’t like always to put the word “cure” here — to, let’s say, long-term remission. Having said so, I would also argue that the data is very mature, and the follow-up with nivolumab and ipilimumab, it is not yet as much mature with a median of follow-up of 1 year, I believe, in both axitinib-based combinations. There is a chance that CR rate, the one becomes around 10% or gets higher with time. That’s an important point to agree on.

And another point I want to say is that yes, all these combinations — and you heard about pembrolizumab/axitinib, axitinib/avelumab, nivo/ipi — have been approved at least in the US. There will be other combinations also coming potentially, and all of these have the same comparator: sunitinib.

The landscape is going to be complicated. And what folks are going to do, and I don’t know if the question will ever be answered, which one is better, but the next question is, how to build on this?

And going back to your question about the good risk — I didn’t forget that, Neil — I don’t think we know exactly what is the difference. However, there is some translational data that comes from our group but also from Tom, who’s been heavily involved in the combination of atezolizumab and bevacizumab. What is interesting here is, perhaps the patient that responds to sunitinib or have had long-term remission have an angiogenic profile. Are these patients also the same good-risk patients? I think this is very important to look at, specifically because there must be something biologic where the IMDC risk group, which is a simple clinical variable, is capturing in an indirect fashion.

Activity of nivolumab alone or in combination with ipilimumab for patients with mRCC and brain metastases

DR LOVE: Tom, just to continue on a little bit more about ipi/nivo, what do we know about the effect of this combination in the brain? In melanoma we’re seeing really great results with the combination. How about in renal cell?

PROF POWLES: Neil, so I love the question, because I think it’s important. I saw a patient where we had this dilemma, and she was a lady, 37 years old — actually from Spain, originally — and has got really a big brain metastasis, very small lung disease. Has intermediate-risk disease. She’s not had a nephrectomy.

We’ve done a biopsy. She’s got clear cell renal cancer. And we’ve given stereotactic radiation. She’s got 2 or 3 different areas. And the difficulty now we have is, what is the brain data for ipi/nivo? And we don’t know.

The assumption that melanoma and kidney cancer are the same, that’s not true. We know melanoma has a very high mutational burden, has a different T-cell infiltration, different PD-L status to renal cancer, which seems unique. Just assuming things are going to be the same, probably fooled.

We’ve got some monotherapy nivolumab data from France, Laurence Albiges, excellent data. And he shows that there is some activity in monotherapy, but it doesn’t look fantastic.

What we do know is that sunitinib as a monotherapy, or pazopanib, doesn’t seem to be magic in these patients. I’ve seen some remissions with these drugs. But it’s not like they’re getting fantastic results.

Let’s have a look at the ipi/nivo data. Let’s collect that information. But I don’t think it’s currently a contraindication. And remember, if you don’t give these drugs to start with, particularly in Europe, it’s hard to subsequently give them. You can’t mix and match towards the end and suddenly give ipi/nivo fourth line or fifth line in. You want to give it now, or you don’t.

I think that my personal perspective of this is, it’s a reasonable approach. I actually look at axitinib and pembrolizumab as attractive in these patients as well.

Tolerability of nivolumab/ipilimumab; risk of immune-related adverse events (AEs)

DR LOVE: Toni, what about the issue of tolerability/toxicity with ipi/nivo specifically in the renal cell population, particularly as patients get older and have more comorbidities?

DR CHOUEIRI: Yes. I mean, it’s legitimate. I don’t think there is a direct relationship, let’s say, with, like, cytotoxic chemotherapy sometimes with age. I look mostly at physiologic age. But comorbidities are important to take notice. And I agree totally with the opening question when Dr Powles, his way to look at kidney cancer patients when they see their metastatic kidney cancer.

I will add 2 things that are therapy dependent and in context. I ask about prior autoimmune disease, which is very prevalent and many times not well captured in the past medical history. I had a patient that said they had arthritis, and when I checked the list of medications, and you know how we are busy in our oncology practice, seeing many patients, the patient was on methotrexate.

DR LOVE: Wow.

DR CHOUEIRI: Upon further asking them again — I had to send them to a rheumatologist, and, actually, they had rheumatoid arthritis. I have to consider again. Same thing I ask about the cardiovascular disease if I want to — especially stroke and recent myocardial infarction, CHF, if I want to use a TKI. Those are very, very important. But I would say both TKI/IO and IO/IO are generally safe. With nivolumab and ipilimumab, you have a significant risk of immune-related adverse events from both drugs, probably more so ipilimumab. And that led, at least on the Phase III trial, to treatment discontinuation due to the immune-related adverse events in around 20% of patients. This is something to really be careful about and educate the community about.

Recent FDA approvals of avelumab with axitinib (JAVELIN Renal 101) and pembrolizumab with axitinib (KEYNOTE-426) as first-line treatment for advanced RCC

DR LOVE: I want to go through all these new combinations that have come out. Tom, there are 2 combinations now that are approved, at least in the United States, with axitinib, avelumab and pembrolizumab. Can you go through the data with both of those and, again, what your clinical take is on that?

PROF POWLES: Yes. Thanks, Neil. I think this is data that has been available now for about 9 months with avelumab and axitinib, which I’ll start with, if I may.

The avelumab/axitinib data — avelumab is a PD-L1 inhibitor rather than a PD-1 inhibitor. We don’t yet know if there are big differences between PD-1 and PD-L1. And this is a really important question that we will need to get to the bottom of over the next few years.

Comparison of cellular distribution, activity and tolerability of anti-PD-1 versus PD-L1 antibodies

DR LOVE: Before you go on, could I just pick up on that in terms of kind of when checkpoint inhibitors came out originally? There was this question — at this point, Tom, is there any evidence in any cancer or any way to differentiate anti-PD — we’re going to talk later about atezolizumab, that’s another PD-L1 agent. Anything in terms of efficacy or tolerability at all that differentiates them?

PROF POWLES: No. It’s very difficult. And actually what we do have is, which I think is relevant, is the mechanism of action is different. And so PD-L1 inhibition spares PD-L2 binding. And PD-L2 binding was originally thought to be important. It was overexpressed in the lungs, and the association with that original pneumonitis story described in lung cancer with PD-1 inhibitors.

The PD-L1 inhibitor is sparing PD-L2 binding. Was thought it may be better tolerable. That’s one of the issues around there.

If you look at the distribution of PD-1 and PD-L1, the reason we measure in PD-L1 expression — we never measure PD-1 expression — is, PD-1 is expressed on a whole series of T cells and immune cells. PD-L1 is expressed on tumor cells and mainly antigen presenting cells. Their distribution is also different. I don’t think it’s just a hypothetical difference between the two. There is some real difference, both in terms of the distribution of expression but also the binding and the sparing of binding of B71 and PD-L2.

I think the assumption that they’re the same is probably not true. When we look from the back of the room as it currently appears, there does appear to be a series of positive trials in lung cancer for PD-L1 inhibitors and PD-1 inhibitors. But one of the more direct or, dare I say it, indirect/direct comparators has actually been these avelumab/axitinib/pembrolizumab/axitinib trials, and the reason why is, they were done at a similar time. It was a global enrollment. The control arm was the same. The standard dosing of axitinib was the same. The methodology was very similar. There are many more similarities than there are differences between these trials.

And what the 2 trials showed was, the response rate for axitinib and pembrolizumab and axitinib and avelumab were actually much higher than with sunitinib, in the region of 60%. And we also showed progression-free survival. Hazard ratios were really impressive as well, in the 0.6s. And so they were both very similar to one another.

The toxicity profiles in the 2 combinations were similar. But the one difference between the two at this particular follow-up period was, the survival signal was 0.53 for a hazard ratio for axitinib and pembrolizumab and 0.78 for the combination of avelumab and axitinib.

Now, that’s an interim analysis, and we need to see more mature data. But all things being equal, the PFS and response rates and toxicity profiles look similar, but there do seem to be subtle differences in this survival signal that may turn out to be really important.

Two things: Number 1 is, pembrolizumab, in a number of different cancers, has appeared to perform extremely well where other drugs sometimes have struggled. And so I think there is a halo effect around pembrolizumab. We haven’t got to the bottom of the biology of that yet, because there’s also some lung cancer data where pembrolizumab has exceeded, and nivolumab, another PD-1 inhibitor, has struggled in the same environment.

It’s not quite as black and white as PD-1 versus PD-L1. And remember that avelumab is a bit unique as both a PD-L1 inhibitor and it has some ADCC activity — it generates antibodies — and that may be relevant as well.

It’s one of those things where you scratch the surface of trying to look at the difference between these drugs, and you can come up with broad statements. But when you actually peel back behind these layers, it gets more and more complex.

To summarize, at the moment we don’t really know that there’s a difference. These 2 trials, the JAVELIN trial and the 426 trial, actually do show this difference in survival signal between the 2 drugs.

The probably easiest explanation as that currently stands is, the drugs are not the same, but one could argue we need to see more mature data, and there were subtle differences between the trials that may have accounted for some of the differences.

Similarities and differences in clinical outcomes among the JAVELIN Renal 101 (avelumab/axitinib) and KEYNOTE-426 (pembrolizumab/axitinib) trials

DR LOVE: Toni, it is said that you shouldn’t be comparing trials indirectly, although everybody does it every single day, but what is your take on this issue that Tom just went through in terms of these 2 axitinib trials and these results?

DR CHOUEIRI: Yes. I mean, at the end of the day, the data is the data. There is the preclinical evidence, and I can tell you, it was all emerging that these are preclinically the same, at least in a way that will get you the same effect, despite what Tom mentioned. That not just PD-L2 is spared with PD-L1 inhibitor, but interaction between the immune cell and the tumor cells are different. Their distribution is different and based on each tumor type.

And the clinical result came — initially they weren’t much different, but now they’re different. They are trials and histology dependent, but what Tom mentioned, that the axitinib/pembro and axitinib/avelumab have many more similarities than differences. And you have 1 study, was an OS benefit. Another, perhaps not yet, was an OS benefit, although response rate and PFS benefit are there.

For everything where a PD-1 is maybe better than a PD-L1 inhibitor, there are the counterarguments. The example I’m going to give you, which is probably the purest example, again, comparing across trials, which you don’t like. Tom doesn’t like. I don’t like. But we’re doing it.

So is single-agent untreated patients with clear cell RCC, where pembrolizumab data from David McDermott resulted in a response rate around 36%, I believe. With avelumab, it was only 16%. However, when you look at the median progression-free survival, noting these are not randomized — these are single-arm trials — it was in the 8-month range. It was, depending how you look at it, and PFS is a primary endpoint in many trials.

I think a lot to be said. I think with axitinib and avelumab, hopefully we’ll have another analysis for overall survival. And at that time, for more events we’ll be able to know better what’s going on.

The biggest thing is the future, I think, and hopefully we’ll touch base on many other combinations, (1), and second, building on a new standard of care in the control arm, which is nivolumab/ipilimumab, axi/avelumab, axi/pembrolizumab, with new agents and new assets.

DR LOVE: Yes, I want to get into new. But just one other thing to get back to this PD-L1 versus PD-1, Toni, my primitive understanding was, PD-L1 binds the ligand and PD-1 binds the receptor. Is that too basic, or is that the way it is?

DR CHOUEIRI: Basic is okay. I think basic is okay. I think this is very reasonable. I think L2 is spared in the PD-L1 when you spare L2. And there is evidence that PD-L2 may not be as inert as we think in terms of prognosis and cancer progression.

What Tom said, at least the rumor there, that initially when we start using these drugs and came in clinic in Phase I is that you need to perhaps spare L2. L2 is expressed in the lung at thepneumocyte level, and that’s why patients are having pneumonitis. And I would say there were a lot of cases of pneumonitis initially, perhaps maybe not if you consider the denominator, but perhaps they were hyped or emphasized more. And these were scary. However, I would say it was more Phase II and Phase III data now coming with thousands of patients — I don’t know what Thom thinks — but I don’t think pneumonitis emerged as a big, huge immune-related adverse event. If something emerged, it’s probably more so the gastrointestinal and the colitis rather than the pneumonitis.

DR LOVE: That’s interesting. Tom, any thoughts about that?

PROF POWLES: Yes. I think that what Toni said is correct. And I think that the biology of PD, PD-L1 inhibitors — because they’re expressed on a broad spectrum of cells beyond the tumor — and I think that complicates simple messaging. But, essentially, what you’ve said is correct. And that piece of the biology is relatively important.

It’s important to remember that, actually, we also measure PD-L1 expression in the immune component as well. That shows it’s expressed beyond just the tumor cells. And that can be prognostic in a lot of different cancers. And so I think it’s an interaction of immune and immune cells, T cell and antigen presenting cells, and of course T cells and tumor cells together.

I would look at it’s a complex piece that’s ongoing in terms of our discovery of how these drugs are working. But, essentially, what you said is correct. And to emphasize Toni’s point, when you look at avelumab response rates and pembrolizumab response rates on the Phase I as single agent, you can do similar work in second-line bladder cancer where KEYNOTE-045 has got a response rate in bladder cancer second line at 21%, atezolizumab, a PD-L1 inhibitor, of 13%.

You can sometimes see those differences. But you also can go away, and you can find areas where actually PD-L1 inhibitors, for example, small cell lung cancer and triple-negative breast cancer, where atezolizumab has performed extremely well and is leading the field.

I think sometimes we try and oversimplify these areas. I’m really keen that we try and keep it tumor specific. I think kidney cancer is different from melanoma. I think it’s different from lung cancer. We know that IL-2 actually worked quite well in some kidney cancer patients. That certainly wasn’t the case in lung cancer, and yet we’re using these PD/PD-L1 inhibitors in similar groups.

We need to be very clear that kidney cancer is specific, and we need to be looking at these drugs. Because they may be the same or different in kidney cancer, and that can be completely different from what’s going on in lung cancer.

As it currently stands, the jury’s out. But what’s available to us suggests that PD-1 inhibitors, nivolumab and pembrolizumab, have both had positive randomized Phase III front-line trials with survival, whereas atezolizumab with bevacizumab and with avelumab and axitinib, the PD-L1 inhibitors as it currently stands have not yet had positive trials for survival.

Two PD-1s both performing with survival advantages, as yet that’s not the case with the PD-L1 inhibitors. Common sense at this point suggests that this 1 group of drugs seems to be outperforming the other, as it currently stands.

DR LOVE: I was just flashing on Bob Motzer’s presentation of the sunitinib data. I think it was at Orlando, before ASCO even went to Chicago, and now here we are talking about all these combinations and multiple Phase III trials coming out. Amazing.

Indirect comparison of the efficacy and tolerability of nivolumab/ipilimumab versus avelumab/axitinib and pembrolizumab/axitinib

DR LOVE: Getting back to indirect comparisons, trying to build on this platform, Toni, how do you indirectly compare the efficacy of ipi/nivo versus these to axi combinations?

DR CHOUEIRI: It is definitely hard to compare, and hopefully there will be a trial, at least to compare between them. But there are some important differences.

The good risk, for example, the IMDC good risk, 20% of all patients — if you need to be treated — because many IMDC good risk have, at the same time, small tumor burden and may not be treated immediately. But let’s say you have something that is a patient, IMDC good risk, and you feel that they need to be treated. It is very hard to justify nivolumab and ipilimumab at this population, because sunitinib, in terms of response rate, PFS, seem to do better. But why I’ve seen some folks even using nivolumab/ipilimumab in good risk? Two reasons. You could have patients that don’t have a small disease burden, (1), even though they’re, on paper, IMDC good risk. Remember, these risk criteria, the part we came up with are certainly not perfect.

And second, the response rate in terms of CR, and there is a lot of emphasis on CR, and potentially putting the patient in a long-term remission remains higher with nivolumab/ipilimumab in the good-risk group. If you’re targeting CR and you have this discussion with the patient, it’s not completely unreasonable to think about nivolumab and ipilimumab. But saying it’s totally okay in good-risk patients without having this discussion with the patient is probably not the best idea.

DR LOVE: Tom, can you talk a little bit about the toxicity issues with these 2 axi combinations, putting together a VEGF TKI and a checkpoint inhibitor? What was seen, and what do you see clinically? And globally, how do you compare it indirectly, again, to the patients you use ipi/nivo on?

PROF POWLES: I think the issue, the first part to that question is, which patients, and are there ways of comparing the 2 groups from an efficacy perspective and certainly, I think, the choice is between ipilimumab and nivolumab or axitinib and pembrolizumab.

I can see advantages and disadvantages from an efficacy perspective of both, I think, from a PFS and from an overall survival perspective. Axitinib and pembrolizumab seems to be performing really well in unselected patients. I think that ipilimumab and nivolumab is not performing well in the good-risk patients and from a PFS perspective didn’t hit its original endpoint.

From a complete response perspective, and Toni’s talked about this, there’s an ongoing debate about what a complete response really is, because not all patients can achieve a complete response. And we need to articulate that debate over the next 12 months, which I think we’ll do. But certainly, the ipilimumab and nivolumab data has longer-term follow-up data as it currently stands. And as Toni said previously, we know that we’ve got ongoing durable responses in that group. And I think it’s that piece that makes the ipilimumab and nivolumab attractive.

There are attractive aspects of axi/pembro, which we’ve described, and there’s some attractive aspects of ipi/nivo. And then the next piece, of course, is around tolerability and toxicity. We haven’t yet seen the quality-of-life data on axitinib and pembrolizumab. The ipi/nivo quality-of-life data looks superior to sunitinib.

Many people talk about ipilimumab and nivolumab being difficult to tolerate. And I think it requires interventions, particularly medical interventions, while you’re giving the combination. But from a patient’s perspective, maybe only 20% or 30% of patients are actually getting into trouble with the drug, requiring steroids or cessation.

Importance of educating patients and community practitioners on the distinction between immune checkpoint inhibitor- and chemotherapy-related side effects

PROF POWLES: Now that, from a doctor’s perspective, is important, because if you let transaminitis or colitis or pneumonitis drift, we can put patients in harm’s way really quickly. The medical community really needs to be keyed up to these toxicity issues.

We are attached to 3 emergency departments at our hospital. We did a survey of all emergency doctors, what they knew about toxicity associated with these drugs. And actually, 90% of our emergency doctors found it difficult to distinguish between an immune checkpoint inhibitor and a chemotherapy drug. And that’s really important, because when patients get ill in the middle of the night, that’s who they’re seeing first.

And number 2 is, they’re also saying that neutropenic sepsis is a big concern of theirs, and so they’re not giving steroids, and they’re talking about giving antibiotics and fluids. That’s not good.

DR LOVE: Wow.

PROF POWLES: And so this huge education piece to talk about before we come into the detail of comparing toxicity.

Now, one of my key messages to people right now is, don’t worry about indirect comparisons to the 2 drugs yet. Let’s talk about that in the fullness of time. But please, let’s make sure our patients are safe. And so that’s make sure the patients are informed well, they attach to a 24-hour community that they can speak to, get advice from and then let’s make absolutely sure they’re not turning up blindfolded to the emergency department saying, “I think I’m on a chemotherapy drug, and I’m not feeling well.”

Patient information, absolutely crucial, 24-hour hotline numbers. Nursing and doctor support. That currently is the most important message as we’re giving these drugs in first-line renal cancer to unselected broad groups of patients. And that’s really where I’d like to start this discussion before we move on to the subtle differences between ipi/nivo and axi/pembro.

DR LOVE: And I think that’s really a great point.

Choosing among newly approved first-line therapeutic options for mRCC (ie, avelumab/axitinib, pembrolizumab/axitinib, nivolumab/ipilimumab); factors influencing whether to use a VEGF tyrosine kinase inhibitor

DR LOVE: Before we hear about the other combinations, atezo/bev, just to go back to Toni, in terms of choosing between the approved options at this point, can you talk a little bit about how you do that?

DR CHOUEIRI: I mean, I think it’s a discussion with patients. Today, if a patient comes, luckily, both Tom and I continue to have trials in the first-line setting, but it’s going to be first if it’s not a trial. If it’s a trial, great. If it’s not a trial, it’s going to be the exclusion.

We talk about, does this patient have significant cardiovascular disease? Recent strokes? I mean, we see these patients, and we see them a lot. We don’t hear about them in clinical trials. They’ll never be eligible. Maybe I want to avoid a VEGF TKI and I go with nivo/ipi.

If a patient has some sort of an autoimmune disease where I think the consultant said that “checkpoint blocker may be okay,” I certainly don’t want to go with two of them, and I may go with one of them and I may go with pembro/axi.

First is, is there a trial? Second, if there’s an exclusion criteria. And (3), which I would have by that time is the IMDC risk group, I could discuss with the patient if they’re good risk nivo/ipi, but I’d probably go with a VEGF/IO combination. And after that, it’s a discussion with the patient about the pros and the cons.

With nivolumab and ipilimumab, it’s the CR — how much this is a chance of CR — and I wish we do have, Tom and Neil, predictors of that CR, of 10% to 12%. What really predicts CR or near CR? We don’t have that yet.

And second is the schedule. With nivolumab and ipilimumab, after 4 infusions of ipilimumab you are done, and your nivolumab is maintenance every 4 weeks. With pembrolizumab and axitinib, it’s visit every 3 weeks. Also, there are talks about pembrolizumab moving to a less frequent schedule, and with axitinib it’s continuous treatment, although this TKI is quite well tolerated in general and has a short half-life, so you could adjust with it.

They’re different. But again, it’s one of those situations, I think, Neil, probably they are closer, I think, in prostate cancer when we had abiraterone and enzalutamide at the same time. And folks in prostate cancer tried to make the difference that perhaps enzalutamide, because abiraterone comes with 5 mg twice a day of prednisone, which I looked at and said, “That’s really not a huge deal, because you could do 5 once a day, which is great, physiologic.” But we arrive to that degree. That’s my 2 cents in general.

DR LOVE: Yes, that’s a great analogy. Just to clarify 1 thing you said though, Toni. Did you say that if people have a history of cardiovascular disease, you shy away from TKIs? I don’t think I’ve heard that.

DR CHOUEIRI: Yes. I did have, actually recently, a patient that had months before — and that’s how his metastatic kidney cancer was discovered — a massive stroke with deficit, but, lucky enough, he was near a medical center, and it was reversible. But in this patient, especially with what VEGF tyrosine kinase inhibitors can do, with potential effects on the vasculature, at least I could stay — I stay away a bit from VEGF tyrosine kinase inhibitors if the patient just recently had an event, a myocardial infarction, ST elevation, MI, a stroke or even if patients have uncontrolled CHF, congestive heart failure.

We have published in the past, was the era of sunitinib and sorafenib, that the risk of these drugs in patients on trial — we polled trial — may be higher. The absolute risk, I would say, remains small. So that’s important. But again, if I have an alternative, that’s great, and I can use these later on.

DR LOVE: I’m amazed at how many things I haven’t heard before, and that’s one of them, but really interesting.

Perspective on the results of the Phase III IMmotion151 trial: Atezolizumab with bevacizumab versus sunitinib in patients with previously untreated mRCC

DR LOVE: Let’s round this out in terms of where things are and where things are heading with the atezo/bev story, Tom. Can you talk about that data?

PROF POWLES: Yes, I can. The atezolizumab/bevacizumab program has probably been the most thoughtful of all the programs. They did a Phase I trial, then a randomized Phase II study. In the randomized Phase II study, they had 3 arms of sunitinib, atezolizumab and then atezolizumab and bevacizumab. Bevacizumab is a monoclonal antibody to the ligand. Atezolizumab is a PD-L1 inhibitor, which, obviously, is distinct from nivolumab and pembrolizumab, which are PD-1 inhibitors.

The randomized Phase II data showed in the biomarker-positive group defined with a 142 antibody that there seemed to be a benefit for the bevacizumab/atezolizumab combination in those individuals who overexpressed the biomarker compared to sunitinib. And for that reason, that arm was taken forward as the primary endpoint of their randomized Phase III study.

The randomized Phase III had all comers. PFS was the primary endpoint in the biomarker-positives, and then OS was the primary endpoint in the ITT population.

They hit the PFS endpoints — they have a positive randomized Phase III trial — but when you looked at the updated OS signal, the most recent signal that was published in The Lancet about 3 months ago — and this work was led by Brian Rini and Bob Motzer — this work showed the hazard ratio for survival was in the 0.9s. That’s more mature than the data we’ve seen for axi/avelumab and for axi/pembrolizumab — it’s more like an 18-month, 2-year survival cut.

And I think at that point it suggests to us that although we’ve hit PFS, which has been a great endpoint for drugs like sunitinib and pazopanib and axitinib in the past, remember, the axitinib/sorafenib data hit PFS but didn’t have an OS advantage. In the past, we’ve been happy with PFS as an endpoint. But because ipi/nivo has got an overall survival hazard ratio of 0.63, which is a 37% reduction in the risk of death, and because axi/pembro half the risk of reduction of death, 47% reduction in the risk of death, it makes that bevacizumab/atezolizumab look uncompetitive, because patients will clearly want to pursue the survival signal.

Why is the bev/atezo combination at this point not performing like the axi/pembro combination? And we’ve had that debate already today. We don’t know the answer to that. One could say is it because the bevacizumab is less active. We know the bevacizumab/atezolizumab response rate is not the same as axitinib or avelumab or axitinib and pembrolizumab. It’s, like, 40% versus 60%.

Bev/atezo might not be quite as strong as the VEGF TKIs, and so we might be losing something there from a response perspective, which may be important. And then the second chapter to the story is, are PD-1 inhibitors and PD-L1 inhibitors the same? We don’t know the answer to that question.

My gut feeling on this is, bevacizumab doesn’t seem to be performing as well as axitinib in the front-line setting by indirect comparisons that we shouldn’t do just looking at response rates. And under those circumstances, that may be contributing a little bit.

But as it currently stands, this combination doesn’t have a pivotal randomized Phase III in the front-line space. And so we are waiting for final survival analysis, which we don’t yet have. But if that survival analysis is not positive, I don’t think this combination will end up being widely used.

Ongoing investigation of anti-PD-1/PD-L1 antibody-based therapy in the (neo)adjuvant settings

PROF POWLES: The one caveat I’d say to that is, there’s an atezolizumab renal cancer adjuvant study, an adjuvant study called IMmotion010 — has finished its recruitment and is really exciting and may be our first adjuvant study to read out. It may be that we don’t see atezolizumab in the front-line space, but we may see a positive adjuvant study we may be using as a monotherapy as adjuvant therapy in the future.

DR LOVE: That’s really interesting, Tom. What other major Phase III adjuvant trials with checkpoint inhibitors are going on right now?

PROF POWLES: Toni leads one. Toni, why don’t you talk about that?

DR LOVE: What’s yours, Toni?

DR CHOUEIRI: Yes, thank you. Thank you, Neil and Tom. The 010, the IMmotion, has finished accrual actually and is an exciting study. The other ones, there is an adjuvant study of 950 patients with pembrolizumab versus a placebo. And that study also allows a very interesting population of M1 NED. I don’t know how many patients we’re going to end up with M1 NED, because that’s a question that oncologists in the community always ask me about. This patient that had kidney cancer and had a liver met or a pulmonary met that was resected. This patient is Stage IV. And they come, and what is the treatment? We do have 2 randomized trials, very small trials, I think both of them Tom knows, both of them were oral presentations at ASCO this year and the year before. Very small, one from Italy and one from ECOG. And there is no disease-free survival or overall survival benefit with sorafenib or pazopanib. Now we have several adjuvant studies allowing M1 NED. That’s very important.

Another study is with the combination of nivolumab and ipilimumab. And that study, the interesting thing about the study, all high-risk renal cell cancer, is that it limits the treatment to 6 months, while the atezolizumab and pembrolizumab limit the therapy to 1 year.

Two other studies I want to highlight, academically led. One study was one of my colleagues from ECOG and Dana-Farber, Dr Lauren Harshman, which is the PROSPER study. And that study is no treatment versus nivolumab, the PD-1 inhibitor, before nephrectomy and after nephrectomy. It’s not an adjuvant study. It’s a perioperative study that involves a lot of academic centers.

And finally, a study that just started with leadership from the Medical Research Council in the UK, for which Tom is heavily — he’s the Vice Chair, and I’m on the Steering Committee. It’s a study, actually, with 3 arms. The patients don’t get any therapy versus they got durvalumab — durvalumab is a PD-L1 inhibitor — versus the combination of tremelimumab and durvalumab. And this study is a large study of more than 1,750 patients. I believe, and Tom can correct me, the therapy duration here with systemic treatment is 1 year, I believe, in the combination arm. Very exciting. One of them finished.

And then, Neil, we may get into the same problem. Imagine if one of the studies reads positive, while one of the studies is still enrolling. But, however, our experience with adjuvant studies, even in high risk, that these studies need time to read and be powered for the primary endpoint.

These studies may change how we treat kidney cancer. And suddenly from overnight, what is the best combination? Is it nivo/ipi? Pembro/axi? Avelumab/axi? Pembrolizumab plus lenvatinib? Cabozantinib/nivolumab? What should we do? But when the adjuvant studies are positive, then you have a population, either cured or progressed, but progressed after a PD-1 or PD-L1 inhibitor.

The data for first-line untreated IO-naïve patients may not apply. I think that’s why Tom and I continue always with our research we have in clinic, investing in new assets in this post-IO space. Because this is coming in case any of these adjuvant studies end up being positive.

DR LOVE: If you think about it too, you already have that situation in melanoma, where people are getting checkpoint adjuvantly. In a way, you have it in non-small cell, when it’s being used in locally advanced disease. To me, that’s consolidation. It’s kind of an adjuvant therapy also.

Toward increasing cure rates for patients with RCC

DR LOVE: It sounded like you were talking about what I’ve heard described as a window of opportunity trial, whatever — you give a drug, and somebody is going to go to surgery, Tom. I’ve seen that in lung. You can call it neoadjuvant. I guess it’s neoadjuvant.

I think it’s been done in bladder also. Has it been done at all in renal? People are going to go for nephrectomy, getting a couple of weeks of therapy.

PROF POWLES: That’s a really interesting question, Neil, because one of the goals that we need to talk about — and I don’t know if we’re going to get time to do it today — but we’re going to have to talk at some point about how we’re going to cure half the patients. That should be our next goal. I think that Toni mentioned it earlier, with axi/pembro and with ipi/nivo. Maybe 1 in 10 patients are going into a long-term durable remission. Maybe more, maybe 1 in 8. But we need to get to a position where we’re really making a big dent in patients who are dying from this disease.

And the best way of doing that may be moving the drugs earlier in the disease setting, which I think is really important. And the other piece is looking at new combinations to try and improve our cure rate in the metastatic setting.

And certainly one of the more attractive approaches, because we’ve struggled, actually — in kidney cancer we’re being successful with ipi/nivo and axi/pembro. If you look at other cancers, lung cancer, bladder cancer, head & neck cancer, we’ve struggled with the CTLA-4 PD-L1. It’s not being universally used. And that’s why, again, renal cancer does seem different from the other cancers where the combinations have been successful.

The question is, if we bring them earlier into the disease setting, can we cure half of those high-risk patients? And that would make a massive dent in this. And that’s why the adjuvant studies, including the ones that Toni’s talked about, are really important.

The neoadjuvant studies, we haven’t been as aggressive. And there are some neoadjuvant studies and window of opportunity studies that are ongoing.

There are 2 attractive points to that. Number 1 is, because the primary tumor is still in situ, they are, obviously, shedding a huge number of neoantigens. There’s easier recognition, broader recognition, you would imagine, by having a bigger cancer burden. And, therefore, the potential to generate better memory for the future in case relapsed occurred.

I think that’s the first point. But if you remove that bulk tumor and there’s only micrometastasis left behind, it may be much more difficult for immune checkpoint inhibitors to be effective.

That’s the hypothesis or one of the hypotheses why it may occur. The second bit, Neil, we just haven’t talked about enough is, remember that if you give neoadjuvant therapy, you could start treatment really quickly. In the real-world in the adjuvant setting, you wait a couple of weeks for your nephrectomy, maybe 6 weeks for a nephrectomy, and maybe between 2 and 6. I don’t know. But it’s not — doesn’t happen the next day.

And then you have your surgery, and then you wait another 6 weeks. And maybe 12 weeks down the line you’re getting your adjuvant treatment, whereas actually neoadjuvant, you’re starting the next week. You’re starting quickly and you’re getting effective drugs in really early, effective immune drugs that are designed potentially for long-term tumor recognition, we hope, obviously, and eradication, generation of activated T cells. And so then you do the debulking surgery, and then you give the period of adjuvant therapy after neoadjuvant therapy. And under those circumstances, it’s possible we could induce longer-term durable remissions.

I actually find the neoadjuvant approach attractive. We have seen data in bladder cancer. We’ve seen data, terrific data, actually, in brain cancer published in Nature Medicine, 3 publications there. We’ve seen data in lung cancer published in The New England Journal of Medicine, and, obviously, data in melanoma as well, with Christian Blank’s great work.

And so I think this piece is really exciting. There’s also terrific biomarker work to be done there, because you can look at dynamic changes to biomarkers, associated response and resistance. We’re going to see data in this space in kidney cancer in the not-too-distant future. And I personally, when we are treating these patients in 10 or 15 years’ time, I actually hope we are doing neoadjuvant therapy.

I remember talking about neoadjuvant therapy before I became an oncologist in breast cancer. And it seems we’ve struggled a little bit in the other tumor types. It does make a lot of sense to me. And I think in the future we’ll be doing neoadjuvant therapy, surgery, biomarker analysis and then tailor-made therapy for the future, I hope.

Approach for patients with no evidence of disease after nephrectomy

DR LOVE: Toni, before, you mentioned the issue of the patient who has NED after a metastatic lesion is resected. I’m just kind of curious what you do in your practice with these patients.

DR CHOUEIRI: Yes. In these patients, I do not use systemic therapy on them. We know now for many, many years, over 20 years, 20, 25 years, there’s a subset of patients — we don’t know who are these patients. We have an idea — that when you resect their metastatic disease, they can have a long-term remission, and some of them are cured.

And who are these patients? It seems like the evidence is showing patients with lung metastases, for example, compared to brain metastases. Although in my practice, I have patients with single brain metastasis doing well 5 years later, resected.

When you have a p-value and a hazard ratio and an odds ratio that is significant, it doesn’t mean that applies to 1 patient. We have that — lung metastasis.

When the disease-free survival or the disease-free interval is prolonged, usually over 2 years, ideally 5 years, so you have someone that had a Stage II renal cell cancer you’re following. Six years later, you have this pulmonary nodule that is growing, growing. You go back, and scans — maybe it was there, but there is no other primary. So this is the same cancer. Maybe it was there and took 6 years to grow, 1 or 2 in the same lobe. You resect them. Those patients could also be cured.

Those are examples where certainly you discuss with patients. And with these patients, actually, I do not use any systemic therapy. Same thing with patients where the indication to resect was borderline but you were able, at least in the short term, to get them to be NED. We have no idea if giving systemic therapy when we don’t see things on scans in renal cell cancer is better than when we see things on scans.

The hope is that with some of these adjuvant studies that involve and enroll patients with M1 NED — for example, on the pembrolizumab KEYNOTE-564 study, we are allowing M1 NED within a year, so at the time of diagnosis or maybe 7 months later, 8 months later, you have a resected pulmonary nodule. As long as your time from nephrectomy is less than a year, you’re allowed. Hopefully we have enough patients to have some meaningful conclusion, because we see that a lot in practice.

Predictors of benefit with immune checkpoint inhibitors and anti-angiogenic agents

DR LOVE: I want to go on and go through a few of your cases, but 1 more question before we do, Tom. Just looking back over these critical trials that we just talked about, as well as prior studies, can you talk a little bit about what we know in terms of predictors of benefit from immunotherapy, particularly PD-L1 levels, how they’re done in terms of different ways they’re done and also other assays, such as tumor mutation burden.

PROF POWLES: Yes. Thank you.

I think the question’s quite a difficult one. I think when we look back, and if we go back perhaps 10 or 15 years and we look at historical biomarkers, we started with IL-2 and biomarkers with immunohistochemistry, like CA9, which are hypoxic-type biomarker. They have not stood the test of time, and we didn’t develop any successful biomarkers to sunitinib or pazopanib, axitinib or indeed any of the targeted therapies.

With hindsight, actually, with RNA sequencing, you can identify angiogenic signatures. And that work was done in IMmotion150 and IMmotion151. And in IMmotion150, there was a study where we compared atezolizumab, which is a PD-L1 inhibitor, with sunitinib, the VEGF TKI. And, actually, you could show that those patients who had the angiogenic signature did better with sunitinib than those that had the T effector. Interferon gamma-type signature did better with atezolizumab.

And so with RNA sequencing, you can identify angiogenic signatures. And, actually, when you look at good-risk patients, guess what? They have more of the angiogenic signature, and that’s why in the ipi/nivo trial, the good-risk patients don’t respond as well to ipi/nivo, because they’ve got many of the angiogenic signatures. And when you look at the poor-risk patients, they have overexpressed immune signatures and don’t have a high expression of the angiogenic signature. And that’s why in those patients, the ipi/nivo is outperforming the sunitinib.

That’s quite a nice piece. And, actually, if you then move forward into Toni’s most recent work with avelumab and axitinib, he’s actually confirmed a lot of those findings. One of the really important findings is, number 1 is, those RNA signatures are really important. It looks like they are discriminatory and can be validated. And indeed they have been validated in IMmotion151 as well.

Number 1 is, I think RNA signatures are important. They’re not here today. We’re not using them today, but I think we will in the future.

The second piece is around tumor mutational burden. Tumor mutational burden is low in renal cancer. We’ve done a lot of work on this. And, actually, it doesn’t matter how high your cut point is — you don’t seem to have a cut point which individuals respond to therapy.

But the third piece is around PD-L1. And the really interesting thing about PD-L1 is, in the second-line setting with nivolumab, it appeared to be prognostic and not predictive for both drugs. But in the front-line setting, there does appear to be a degree of prediction when you compare ipi/nivo versus sunitinib and, also, actually, axi/avelumab versus sunitinib. But the strange thing is not that the ipi/nivo patients who are biomarker-positive are doing so much better. The strange thing is, sunitinib does not appear to work well on PD-L1-positive, biomarker-positive renal cancer patients.

It’s actually the sunitinib patients who are biomarker-positive who are doing really badly. And that comes back to that piece around if you’re positive, you have a less-strong VEGF signature. And that’s that piece that we’re beginning to unravel now, and Toni said this previously, and I’m going to repeat it, the IDMC score is not important in its own right. If you have anemia, this, the anemia is not driving the fact that ipi/nivo is working better than sunitinib. It’s just a surrogate for the biology of the disease. And the biology of the disease is, those patients with more aggressive disease have less clear angiogenic signatures — therefore have poor-risk disease. And because they have these more powerful T-effector signatures, those patients are responding better to immune therapy than drugs like sunitinib.

Case (Dr Choueiri): A man in his mid-60s with clear cell mRCC who receives nivolumab/ipilimumab and for whom nivolumab is held briefly because of a number of rare immune-related AEs ultimately attains a durable near CR

DR LOVE: I want to go through a few of these cases that were submitted. Why don’t we start out, we were talking about ipi/nivo, and Toni, you have a 64-year-old man. Can you just briefly present what happened with him?

DR CHOUEIRI: This man came to my attention because he had systemic symptoms, and that was in March. And there was discovered the patient had fever, weight loss. And soon, for like 3 weeks, as soon as he had hematuria, he went to the ER and was discovered to have a large, large kidney mass, a right kidney mass.

He underwent nephrectomy that showed clear cell renal cell carcinoma with around 40% sarcomatoid feature. And at that time, this man did not have any metastatic disease and was not referred for any further management, which is appropriate, I would say.

However, almost a year later presented with the same symptoms that triggered imaging and was found to have metastases in the pancreas and the spleen. He underwent pancreatectomy and splenectomy. And on his follow-up scans — he had them in February — he had paraesophageal nodules, omental nodule, biopsied again, clear cell renal cell carcinoma.

He underwent at that time 4 cycles of nivolumab and ipilimumab with best response being PR. And when he finished ipilimumab and was on nivolumab, nivolumab was held because of Grade 2 epididymitis and hypoglossitis. And we ruled out any potential infection. Was it the epididymitis? Was referred to urology. Got an ultrasound. Got a urinary exam and was hypoglossitis. Also ruled out any potential fungal infection or anything.

And when you go back in time, and this happened over 2 years, actually, and you start asking the patient about symptoms specific to the epididymitis and hypoglossitis, some of them were there and got worse with time. That tells you, important to see the patient.

The patient then was restarted after his symptoms. He had low-dose steroids. The symptoms got suddenly better. Gave you a potential diagnosis here. And we continued nivolumab, actually, for 2 years. The patient achieved a CR, so now we are almost 4 years later from the time that started nivolumab and ipilimumab. And the symptoms started to come again.

Last month we had a joint decision to completely stop therapy. The patient is in near CR, and I continue following him with serial imaging.

Clinical experience with and management of rare immune-related AEs in patients receiving checkpoint inhibitors

DR LOVE: Continuing along the theme of things I’ve never heard of before, Tom, epididymitis and hypoglossitis with checkpoint inhibitors. Have you ever heard of that or seen it, Tom?

PROF POWLES: It’s a good question. We actually have seen hypoglossitis. We’ve seen that with high-dose IL-2. We used to be 1 of the 2 UK centers that gave high-dose IL-2, and I remember treating a patient 10 years ago who got hypoglossitis associated with high-dose IL-2.

DR LOVE: Wow.

PROF POWLES: It's a recognized side effect in immune therapy. And epididymitis is entirely reasonable within those categories, though I’ve not seen epididymitis before.

What I would say is, yes, I wouldn’t call that unexpected. And I think one of the really important points, Neil  — I’ve said one thing about toxicity already about the importance of emergency department, the importance around patient and doctors’ information. The second thing I’d like to say, which isn’t talked about enough, is that we use CTC criteria to measure toxicity, and those CTC criteria are really based around chemotherapy, and that’s what the C stands for.

Now, so, actually, immune therapy, what we’re seeing is these rare side effects that Toni’s described here that’s happening to 1% or 2% or 3% of patients is not happening frequently enough for us to write them all down and to report them all and for them to be frequent enough in each trial. But what we are seeing is 1% or 2% of patients getting really rare side effects, and I’ve got a patient who got really bad pan fungus of the leg, a really bad skin rash. We did the biopsy, and he’s out with pan fungus. We thought — they were actually on VEGF targeted therapy — at the time, we thought it was VEGF TKI therapy. It wasn’t. It was actually really bad immune-related pan fungus, and you see that.

I remember a case we saw of demyelinating brain disease, where the patient had demyelinating brain disease and got an MS-like picture, which was extremely unusual. Of course it’s 1 in 1,000 cases, so it’s not happening in all the trials. But if you add all of this up together, what you’re getting is the unexpected. And so if you’re treating these patients with immune therapy, what I would say is, expect the unexpected. If you’re seeing things that you’ve not read about in the books and you can’t find online but you haven’t got to the bottom of it yet, a neuropathy of some description, an unusual skin rash…

Insulin-requiring diabetes, one thing, our institution might have been the first to describe that. We saw a case of insulin-requiring — someone pitched up their emergency problem with diabetic ketoacidosis. It was 4 or 5 years ago now. But that at the time, people didn’t — of course it’s immune related. Now much more recognized when we see it maybe in 1% of patients.

But do expect the unexpected. And you probably need to treat 200 patients before you can be confident that you’ve seen some of these rarer side effects. If it is rare, get together in multidisciplinary settings. Reach out to people.

And then the last point I wanted to make about toxicity, which I also think is important, is, at the beginning of this journey, we just phoned up the gastroenterologists or we phoned up the dermatologists or the endocrinologists. But if they’re not experiencing these drugs, they’re no better at treating it than we are. And what we were doing at the beginning was, we were doing investigations. Actually, if you think it’s immune-related toxicity and it’s there, you need to start steroids first and investigate second, not the other way around. Because if you let these toxicities — like pneumonitis, colitis and transaminitis — if you let those escalate unchecked while you’re doing the investigation, they are a whole load more difficult to treat when they’re more severe.

DR LOVE: I want to hear, Tom, about your 32-year-old woman in a second, but I was just flashing on Julie Brahmer has this thing. She says anything that ends in an “itis” you can see. But I’m just kind of curious, the hypoglossitis that this patient had, what symptoms, and how did it present clinically?

DR CHOUEIRI: The patient actually simply had tongue soreness. Similar to, if you want, mucositis that is functional. Sent him to dentistry. We sent him to an oral surgeon. We looked at things going on, and nothing there. Think about it: It’s almost a gland, just around the tongue, being affected. And the other thing when you rule out other situations in the hypoglossitis, potentially the first thing you think about is thrush — fungal infection.

When you give them steroid, which should get better if these are fungal, and immediate improvement over 24 hours in symptoms, especially if they’re mucositis. Put that diagnosis there.

We’ve seen, as Tom had said, myocarditis. Any “itis,” you can name it. We’ve seen people with salivary gland problems that end up having a Sjogren-like. And most recently I’ve seen 2 patients actually with blepharitis. They creep up on us.

DR LOVE: Wow.

DR CHOUEIRI: And since blepharitis and having a chalazion or anything are common, we don’t always pay attention. And we tend to have sometimes a knee-jerk reflex and send the patient to a specialist. And the specialists come back to us. I don’t think they have more experience than we do. I mean, dealing with immune-related adverse events, it’s almost like now doing robotic prostatectomies. Once you see 200, 300, you’re fine with it. And the oncologists end up, and the nursing team, the folks with the most experience.

And I want to add 1 small thing. It’s about the steroids. There’s a lot in the literature there. And I would say it is controversial, and I’ll tell you my own view. Where if you use a corticosteroid that inhibits the T-cell function in general, you may blunt the immune response for treatment. I think this is a very controversial statement, including with folks recently that published in the Journal of Clinical Oncology and others. I think when you need to use steroid, you need to use steroid. And, actually, you may need to use them with a very slow taper.

I think in many of these studies, when you address for the patient’s baseline characteristic, you see their disease is actually really aggressive. In my book, when you need to use a steroid, you use steroid. Is it possible that steroids make the disease grow and lose that response to immunotherapy? I think it’s less likely, in my book. I think an immune-related adverse event, then you need to handle. Is very important.

Luckily, though, there is mounting evidence, perhaps not as much in renal cell, because I don’t think this is tumor dependent, that immune-related adverse event may be associated with more responses and more durable responses. So you feel that this is not a patient that had already progressive disease and you’re adding steroid on it. Luckily, sometimes mutually exclusive.

I want to make sure that that point is well taken. I don’t know what Tom thinks, but I’m very liberal with the use of steroids.

DR LOVE: I just was flashing on this case we had — I think, Tom, you might have been there. It was, I think, ASCO GU — of a patient, had very bad Sjogren’s syndrome related to checkpoint therapy, then had this incredible response in the lung to metastatic disease.

Case (Prof Powles): A woman in her early 30s with newly diagnosed intermediate-risk mRCC receives nivolumab/ipilimumab

DR LOVE: Why don’t we get into another case, though, and Tom, you have a 32-year-old lady, very young patient. What happened there?

PROF POWLES: This was a lady we actually saw quite recently, and she had a small child and came with her husband to clinic having had a CT scan after an episode of hematuria. She’d been really quite well in the past with irritable bowel syndrome but nothing of note and wasn’t on any regular medication.

The CT showed a T2N0M1 tumor. The metastases were in the pancreas but also in the bone, and the bone was problematic — there was more than 1 site in the bone, and that, obviously, worries me a great deal. No one likes any metastases. Pancreatic metastases tend to grow slowly, in my experience. But the bone metastases are a real problem.

That was worrying me a great deal. She was anemic as well. And she, therefore, because of the treatment-free — the interval being short, she scores a point there, and she scores a point for her anemia. She’s intermediate-risk group, with 2 points. And this is the group, for me, that in the past we probably would have done a nephrectomy on with CARMENA. We don’t do that anymore, in my opinion.

We’ve actually been, as Toni described earlier, if you want to do a nephrectomy, in my opinion you need a multidisciplinary approval of that. Surgeons performing nephrectomies on patients without a wider discussion, I think, is probably the wrong approach. I know that there are very experienced surgeons on both sides of the Atlantic, but my current approach to this is to say, if you get a multidisciplinary group of people to agree this is the right thing to do who are familiar with the CARMENA data, under those circumstances, do a nephrectomy. But we’re not doing nephrectomies on patients with intermediate-risk disease and bone metastases. We’re actually much more keen to intervene.

We actually, in the UK, currently we have access to ipi/nivo. It now has European and NICE approval, driven by its great survival advantage. And so for her, with intermediate-risk disease, a year ago it would have been sunitinib and pazopanib. But now there’s a discussion for her about what treatment she should have, and clearly the European and American guidelines with intermediate-risk disease would support both ipi/nivo and axi/pembro. Both treatment options are attractive.

I’ve described already axi/pembro, less mature follow-up. And probably more impressive response and PFS data, and a hazard ratio for survival is 0.53. I’ve talked about ipi/nivo, more mature data, perhaps less impressive PFS data. But we know about the long-term durable remissions and quality of life.

I actually would be happy with both approaches. I could see myself giving this patient axi/pembro. She actually got ipi/nivo, because the axi/pembro is not yet EMA approved.

She had ipi/nivo. And she’s actually in the middle of treatment, and she’s doing well, and she’s responded well to therapy.

One of the questions, which one of the surgeons puts to me the other day was, “A few months down the line, should we be cutting out her kidney now?” Because you’ve got control of her metastatic disease, and the pancreatic metastasis is smaller. The bone mets are the same. They don’t respond to therapy. But should we be cutting out this 7-cm tumor that’s now 4 centimeters?

And my opinion of that is, we shouldn’t be doing that as it currently stands. If we’re 2 years down the line and there’s no other evidence of disease or a year and a half down the line, I’m happy to have that debate. But number 1, it’s currently too early. And number 2 is, there’s no evidence that interval surgery is a benefit to these patients. I don’t know what Toni’s opinion is of what I’ve just said.

DR CHOUEIRI: I agree. I think because of the age, even if I had axi and pembro, I would have done nivo/ipi. There was no contraindication for it. And look at that potential CR. It’s a great case, Tom.

I would have also, I mean, sent this patient parallely to — and I’m sure you did — to a genetic counselor and looked at the histology, again, to make sure this is not translocation renal cell cancer in a young female like this, knowing that it’s uncommon but more common in these younger females.

The question is, with a response rate of 42% in patients with intermediate/poor-risk disease, Tom, if this patient doesn’t respond, the most important question: What’s your second-line treatment? Is it going to be TKI alone? Or, now you have pembro/axi approved. And lets say it’s approved independent of the line of therapy. Would you use a TKI, whatever that TKI is — cabozantinib, axitinib — or would you continue with the checkpoint blocker and use pembro/axi? And this is a question I cannot answer in my practice. But if I feel a patient progressed on a PD-1 inhibitor-based therapy, like nivo/pi, I don’t know if there is any evidence that patients that progressed, let’s say, on nivo, have any chance to respond to pembro. But I cannot rule out the possibility of synergy between pembrolizumab, let’s say, or a PD-1 inhibitor and a TKI, even after you progress on a checkpoint blocker. And hopefully you and I are able to answer this important question in a randomized trial.

Case (Prof Powles): A man in his late 60s with intermediate-risk, clear cell mRCC experiences low-grade immune checkpoint inhibitor-associated side effects with nivolumab/ipilimumab

DR LOVE: Just to maybe touch a little bit more in terms of toxicity, you had a couple of cases here to just probe a little bit. I was just kind of curious what happened. Tom, you have a 68-year-old man who was on ipi/nivo and developed diarrhea and fever. What happened there?

PROF POWLES: This was one of the first cases we treated, actually, and it’s a good case, because the patient actually wasn’t that unwell. And they had some Grade 1 symptoms. And had they been on sunitinib and phoned our nurse 2 years ago, they probably would have said, “Stop the tablets for a couple of days and see how you feel.” And that would have been reasonable.

We weren’t hauling these patients in for blood tests at that point. But when we started immune therapy, we have started doing that. And, actually, when we looked at this patient, we brought them in and they had a transaminitis. Their platelet count had also dropped. We see that with immune therapy toxicity, you see the platelet count drop, and those two made me nervous. And because of the diarrhea, we get a CT scan, and we showed a bit of pneumonitis and we showed a bit of colitis.

As you actually look at this patient, he’s actually saying, “I’m feeling okay,” but they’ve got a pneumonitis. They’ve got colitis. They’ve got transaminitis. The drug’s affecting their platelet count.

This is actually a patient who’s got multiorgan failure with immune therapy. And these are the patients that make me extremely nervous. And I think that this underlines the important nature of that and really swift, early intervention.

I’m really pleased I was given the opportunity to reply to what Toni said earlier about the steroids. I’m liberal with the steroids. Obviously this patient needs steroids. But the story really is, for this, he’s around the early intervention.

And then the second piece, is if you look, his symptoms actually recover very quickly on the steroids, although the transaminitis is slower. But the key is, you need to reduce the steroids slowly. This is not a case you say, “Okay, his symptoms are better. We’re day 7. The scan looks okay. Let’s stop the steroids.” Because what happens is, his bounce is back. And then what happens when it does bounce back? Then it’s harder to treat a second time. And then you don’t get the patients back on the immune therapy, because everyone gets nervous, saying, “The transaminitis is going to come back.” We stop the steroids.

The key to this is early intervention, be nervous and particularly — and the reason I chose this case is, some of the sunitinib-type toxicity in the past, which we ignored, is superimportant with immune therapy. And when you do start, don’t stop the steroids early. Taper slowly. And what you can do is start the immune therapy 4 weeks or 5 weeks later. Don’t be panicked to restart the immune therapy. It has a very long half-life. And make sure the toxicity has gone back to normal before you jump in with more treatment, otherwise you’re going to get the original problems you started with.

DR LOVE: Yes, that’s really great advice.

Case (Dr Choueiri): A man in his mid-50s with clear cell mRCC receives avelumab/axitinib on a clinical trial

DR LOVE: Let’s finish out, Toni, with your 56-year-old man. Always intrigued if I see the term “ICI-induced myocarditis.” What happened?

DR CHOUEIRI: Yes. So this is a young patient doing well. Had a radical nephrectomy and had clear cell RC and 10% sarcomatoid feature. And upon staging was found to have lung metastases and bone metastases. And we started him on a trial with avelumab/axitinib at that time. And he got 2 infusions but with the second infusion had a Grade 3 infusion reaction with shortness of breath and chest pain.

Now, with avelumab, which is an IgG1 wild type was the initial trial, there were reported a lot of infusion reactions. Very few, if any, Grade 3. This was unusual. That’s why in the label there is premedication with acetaminophen and H blockers. We stopped avelumab completely, and the patient got better and got a scan, after only 2 cycles, and each cycle was actually 2 weeks here and had a partial response.

DR LOVE: Wow.

DR CHOUEIRI: And totally partial response.

We decided at that time, let the patient get better, and we’ll see when to restart axitinib. The patient presented to the ER a month later with shortness of breath, orthopnea. A workup was done and was found to have an injection fraction of 29% and luckily was — and this is by pure luck — on the cardiologist service, and the attending that was hounding — was an attending, and we have a huge cardiology service at the Brigham, was interested in cardio-oncology. So this is not — was a consultant whose background is cardiac catheterism or echocardiogram or anything. It was that.

Immediately an MRI was done that showed diffuse myocarditis from a toxic cause, and steroids were actually started that day. And a biopsy was obtained later, after the steroid by a few days. It was negative, but we know very well that those endomyocardial biopsies sometimes are patchy.

We tapered the steroids, and we repeated. The patient got better. The ultrasound and the ejection fraction went up to 50%.

Myocarditis, again, like Julie was saying, any “itis” is possible. Luckily, myocarditis are very, very rare with all the Phase I, II or III with checkpoint blockers in renal cell and bladder cancer. I have seen two, and one was a referral, and this was the second one. But they could happen. And if you have a high index of suspicion, you need a referral.

DR LOVE: Just curious. What’s the patient’s current situation?

DR CHOUEIRI: With more follow-up, the patient actually has a complication we believe from the myocarditis. Developed a significant pericardial effusion around a month later that needed an admission initially to the ICU. And this was complicated by a tamponade. And around that time, I repeated the staging. Because as you know, with oncology patients, metastatic coming, with every intervention, the question by the consultant team is the prognosis and where we are, to know how much you’re going to push. And we repeated the scans at that time — remember, the patient had a partial response after 2 cycles — and at that time had progression, slight progression in the bones but had brain metastases.

Based on that, the patient got better and was started on cabozantinib and was planned for a rod nailing for one of the painful metastases. And he underwent surgery and, unfortunately, around extubation he developed a PEA arrest, which, of course, he doesn’t have the best organ function here, and he survived that PEA arrest. He had a decrease in his ejection fraction. But even with that, we started him later on cabozantinib. He did well for a few months, then had more disease progression.

Six months later, unfortunately, the patient passed away from disease progression for sure but also was all the background comorbidities that happen initially with the myocarditis. I do not think this helped the patient.

Choosing between pembrolizumab/axitinib and avelumab/axitinib

DR LOVE: I asked Tom before about this issue of avelumab as opposed to pembro plus axitinib. You very much, you looked at this issue of avelumab. Do you see a current role? Is there a situation where you use avelumab and axitinib right now, Toni?

DR CHOUEIRI: There is no situation I can think of. I participated, my disclosure and all the axitinib/avelumab studies, so I have patients now that have a great response that are transitioning to standard of care, because they don’t see me every 2 weeks. I try not to take this personal. So they go on axitinib/avelumab, a standard of care, where I perhaps can see them every 4 weeks, 6 weeks and they’re doing very well.

I think with the pembrolizumab/axitinib overall survival benefit, actually, it’s in our guidelines. I really hope we have an embarrassment of choices where we have many of these events and these discussions with you, Neil, and try to find out what are the situations where one combination over the other makes sense?