The Uromigos Episode 116: Renal Cancer Surgery Controversies at EAU

Dr. Grant Stewart discusses renal cancer data presented at the 2021 EAU Conference, including partial versus radical nephrectomy, KEYNOTE-564 trial results, and other controversial topics.

Episode Transcript

Brian:
Hey, everyone. Welcome to our next podcast. We’re here with Grant Stewart. We’re going to talk about some urologic issues. This is a few podcasts we’re doing around the EAU meeting. Grant, welcome. Why don’t you introduce yourself?

Then, let’s kick off with discussion about radical versus partial nephrectomy, which is sort of come back, sort of an in and out in the urologic circles. Sounds like it’s back.

Grant Stewart:
Yeah, thanks Brian. Thanks, Tom. Thanks for having me. Grant Stewart’s my name. I’m professor of surgical oncology at the University of Cambridge and I’m a consultant urologist here in Cambridge. Pretty much just treating kidney cancer.

Radical versus partial nephrectomy came up again at the EAU meeting over the last couple of days. Surgery, we’ve been making decisions on for years, isn’t it? So, you would think that we have this all kind of sorted by now.

But I suppose robotics have infiltrated our practice, but actually with minimal high-level evidence. On the surface, there should be less impact on patients because it’s less morbid than open partial nephrectomy, but should we be doing an increasing level of robotic partial nephrectomy, pushing the indications over and above a relatively straightforward laparoscopic radical nephrectomy?

Surgeons like doing robotics. I love doing robotics. It’s techy, it’s minimal access surgery. It’s definitely less physical for me to undertake. But it’s expensive. I think we are seeing some of the complications, the specific complications of urine leak and bleeding that increase more often as indication creep comes in. We are seeing some odd recurrences as well so there’s something that needs a little bit further evaluation there.

Brian:
Grant, obviously Tom and I are non-surgeons. I thought the robot made surgery easier. Shouldn’t there be less complications, less recurrences, because it’s a better operation or is that not true?

Grant Stewart:
It definitely makes surgery easier. It definitely makes minimal access partial nephrectomy much easier. I mean, in my hands now, I think I can do a better partial nephrectomy with the robot than I can do by open surgery. I don’t think the argument’s there, Brian. I think the argument is more, should we be doing complicated robotic partial nephrectomy versus in somebody who’s got a normal contralateral kidney versus a radical nephrectomy.

Brian:
I know there’s been data around, obviously, renal function and then survival after each. It seems to have swung back and forth, whereas we definitely want to do partial and preserve as many nephrons as possible. But now it seems like maybe it’s swinging back. Is that correct?

Grant Stewart:
I think to a degree. The evidence, like a lot of these surgical questions, is patchy. We’ve got large retrospective data series which seem in favor of partial nephrectomy for those criteria you mentioned: renal function, overall survival, et cetera. But actually, the only clinical trial we have, which yes, was underpowered, didn’t show any additional benefits of partial nephrectomy on cardiovascular events or on survival. It’s remained the persisting debate.

Tom:
Grant, do you want to go through that trial a little bit? Because some of our listeners, including myself, do you want to just talk about what that trial, how it was done, why it was done, and what it showed?

Grant Stewart:
Sure, absolutely. A little bit of history now. This is a trial from the EORTC back in 2011, it was published. It was a study randomizing patients with tumors less than five centimeters. A slightly bizarre size for our current staging, but nonetheless, that was the data that was there when the trial started back in 1992. As I say, a bit of urological history here.

They followed patients up for almost 10 years and they found that both methods, so randomizing patients to radical nephrectomy versus partial nephrectomy, gave equivalent oncological outcomes. But the nephron-sparing surgery, partial nephrectomy, was less effective than radical nephrectomy in terms of overall survival.

There were small numbers of progressions and the power, and the study was slightly underpowered, but that is the only randomized data we have.

Brian:
Because I remember there were very few recurrences on either arm.

Grant Stewart:
Yeah, absolutely. I think it was just, I’ve got it here actually, only 12 of 117 deaths were due to renal cancer. The questions reemerging interestingly in the United Kingdom are government-funding source.

The NIHR has just put out one of their health technology assessment calls to readdress this question of partial nephrectomy versus total nephrectomy. And so, we’ve been pulling together a multi-institution trial around this and we’re going to look at all of these issues again, actually.

Not so much the very small, low hanging fruit partial nephrectomy patients where that is a no brainer that partial nephrectomy should be undertaken, but more the intermediate size, so stage 1B and above, or the very deep-seated endophytic tumors that are deep inside the kidney that are T1a. These are the patient groups that we want to evaluate called renal function.

Brian:
I just have a question. Is it the kind of thing that, you know some things just don’t lend themselves to randomized trials. A lot in the surgical world is in this. Is this the kind of thing that maybe just doesn’t and it’s always going to be dealer’s choice?

Grant Stewart:
Well, it’s going to be tricky, that’s for sure. Because the key thing is the endpoints and exactly what you should pick as the endpoints. Because otherwise, as we said, we’re going to end up with so few recurrences. If we look at oncological endpoints, then we’d need an extremely lengthy study.

The trial that’s being designed is actually looking at renal function gains and assessing those against harms of unusual recurrences that you might see with robotic partial nephrectomy. We’ve all got a story of a strange recurrence site, but equally we’ve got lots of patients who do exceptionally well from robotic partial ectomy. So, in measuring those arms.

Tom:
One of the issues, I guess, is arguing around whether axitinib or sunitib was a better drug and we did big trials. In the grand scale of things, it turned out to be not that important question in here. Where actually this isn’t that important a question.

The reality is the world’s moved on, we’re doing robotic partial nephrectomy, who’s going to do it differently. And so, why don’t we ask different questions? Are we in a position where we’ve moved on or is there still sort of equipoise from your community around this question?

Grant Stewart:
Good point. I think the point is to find the sweet spot where there is equipoise. Where there will not equipoise is in a three-centimeter exophytic tumor in a young fit patient. Where there will be equipoise is where you have a larger T1b tumor in somebody who is not that fit.

Otherwise, a laparoscopic radical nephrectomy may well be the way to go rather than pushing the envelope with partial. But I would just come back to your point, Tom, about is this an important question? I think these are sort of similar to some of the studies you guys will do around de-escalation. It’s around de-escalation of harms.

I completely agree with the sentiment of your question, which is, is this going to cure more people of kidney cancer? I think the answer is on the whole, no, but it’s around just nuancing the way that we treat people with surgery around this.

Brian:
So then, what else at EAU? I know that I’ve seen it with the adjuvant pembro data that was presented, that’s a topic. Talk around, sort of, your feelings on the data, if anything else was presented at EAU on that whole topic.

Grant Stewart:
Well, I think Tom, you gave the summary talk, so I’m kind of speaking in place of the expert, but I’ll give you some [crosstalk 00:09:01]

Tom:
No one had never said that before.

Grant Stewart:
Congrats to you and Tony for leading the keynote 564 study. Fantastic. I mean, as soon as I saw the press release, this seems to be the way of oncology now. It’s sort of told by press release. I was pretty excited when I saw that a few months and then really pleased to see the disease-free survival hazard ratio being under 0.7.

That always strikes me as a level that we’re getting to an effect that is going to be impactful and will prick up the ears of the regulators.

If you remember back to, obviously that we’ve really just finished the trials in the TKI era for adjuvant treatment with the source trial being the last one to report, but the only positive study there, they were all negative, except for extract had a hazard ratio of 0.76. So, we’re getting under seven.

I mean, I’ve got a few, there are clearly some outstanding questions to be answered, and it’ll be great to get some thoughts from Tom on this, but OS is clearly the big thing. This is a very early report. I think it was 51 events on the OS side out of what I saw was a requirement for 200 to give us the mature data. Is that right, Tom?

Tom:
I say, yeah.

Brian:
Grant, if there is no OS benefit, will you give the drug in the adjuvant setting?

Grant Stewart:
I think the other thing that’s missing at the moment, or I haven’t seen, is the quality-of-life data. If we have a disease-free survival signal, acceptable toxicity, and at the moment, from the presentation slides G3, G4 talks was about 20% and there’ll be some lifelong immune toxicity with that which is relevant. I would really like to see some quality-of-life data. That the patients who were randomized to pembro with a DFS advantage did have an improved quality of life.

Tom:
Grant, what did you think? OS is trending in the right direction. It’s not statistically significant. It had a ratio that’s actually very low, but as you said, they’re relatively few events as it currently stands. Would that OS trending in the right direction, would that influence your decision making at all?

Grant Stewart:
Of giving the drug? I think …

Tom:
Yes. What conversation are you going to have with a 55-year-old man who comes with three tumors with some sarcomatoid component to it. He says, “Is there any treatment for me now?” Let’s [Inaudible 00:11:45] may approve and you can give it, what conversation will you all you have with him?

Grant Stewart:
I will have the conversation based around the risk profile that he has. The Leibovich score would be the one that I would use. I’ll give him some pretty precise figures around his risk for recurrence. Then, I will lay the data out there. I’ll tell him, the DFS data that there isn’t currently OS data, if that’s the scenario you’re giving me.

I think the key thing, and this will be done in the UK as you know, this is done jointly and led by medical oncology colleagues rather than the surgeon, unlike some of some European countries. But absolutely giving them the data on the toxicity profile. I still would come back to what the experience of the people in the trials were around quality of life.

The other thing I just wanted to flag as well that’s interesting from the data that we have. So, unlike all of the other … well, like some of the other adjuvant trials that are ongoing at the moment, the population included M1, no evidence of disease patients. They’ve had a nephrectomy and they’ve had metastatic disease removed at the same time.

The hazard ratio there was very low. So, these patients do very well with a hazard ratio of 0.29; whereas for the M0 population, it was 0.74. I also want to learn a bit more before I get into those conversations with my patients, a bit more about how much that influences the overall result, where effectively you’re treating micro metastatic disease with a very good drug, and clearly, they do better than placebo. That’ll be interesting to hear some thoughts on here.

Brian:
Grant, if you have the same patient in front of you, where you want to give adjuvant therapy, let’s say you can and the patient wants it, would you just give adjuvant pembro, or you have a trial where the patient has a 50% chance of getting a placebo?

Which one are you going to offer the patient? I know the answer is always offer the trial, but which one are you going to … The patient says, “Doc, I just want the drug.” But you have this great trial that you want to complete, and more data is important. How are you going to navigate that?

Grant Stewart:
Well, Brian, obviously, you know what I’m going to say, which is I’m going to offer them the RAMPART trial. RAMPART’s a trial that we’ve all been involved with. The great thing about that study is it’s a multi-arm, multi-stage design, which means, like STAMPEDE, which is the archetypal study in that space, that we can change the control arm and we can change treatment arms as we go along.

So, even when we see the approval for pembro and we see the approval for, we’ll see PD1 approval, then we’ll be in a position to change the control arm of RAMPART, which is currently standard follow-up observation, not a placebo, and it will be possible to change it.

Now, I’m not saying that necessarily PD1 equals PD-L1 in renal cancer. I don’t think we necessarily know that. But the nimble design, and we’ve had a discussion about this, means that we could change it very, very quickly. I would hope still to offer them the trial, and then they would be offered durva in the case of RAMPART versus durva and treme. There would be still what they would get potentially from pembro, but plus the option of the CTLA for inhibition as well.

Brian:
You’re still comfortable offering a trial where placebo is an arm even in light of these data.

Grant Stewart:
Currently, yes. Because we’re a little bit away from that, in my opinion. I mean, the other thing, and again this isn’t my space per se, is that we are seeing overall survival … well, we’re seeing complete responses in metastatic disease to the combination treatments that you are now able to give.

Worst case scenario, and somebody ends up on the placebo arm and develops recurrence, there is still very, very good agents in the metastatic setting that we can offer them. We’re not yet seeing that overall survival advantage in the adjuvant setting, as Tom says. Hopefully with more events, we will continue to see that nice figure mature that we’ve got a clue for at the moment.

Brian:
I agree with you. I’m still offering patients adjuvant trial therapy, but I think this conversation will evolve as survival data comes out, maybe as quality-of-life data come out or subsets or whatever, but to be continued. And there are obviously other huge trials. I mean, this is just one piece of data.

There’re four or five other trials. You mentioned RAMPART, there’s several other industry-sponsored trials that are ongoing, that I think will further define the space. If you think about the TKI trials, the abstract trial didn’t really define the space. It was almost everything else, so we’ll see how this field evolves.

Grant Stewart:
And I’m assuming Brian, that you very rarely give sunitinib in the adjuvant setting in the U.S. We don’t have … It’s not like [crosstalk 00:16:50]

Brian:
I think that’s accurate. I think I’ve personally done it maybe twice or so. Some people do a little more, some don’t do it whatsoever. So yeah, I’d say sparingly used is probably the right term. I don’t think that’s the issue.

Tom:
Grant, kidney cancer? Surgeon kidney cancer?

Grant Stewart:
Well, you’re going to get a bit of a bias view from me. One topic I’m really interested in is screening. Whether, in fact, the is a role for early detection of people who are asymptomatic with renal cancer. We know that the best way of curing people of kidney cancer is to find it as early as possible and to treat them with surgery.

That’s going to be curative. If we can get those people and get them early. We also know that 90% of people with those smaller renal cancers present incidentally. How are we going to improve on that? That’s an area I’m really interested in.

Tom:
Grant, just to focus on that for a second. Are we doing too much operating on small renal masses from patients who don’t need it?

Grant Stewart:
I think the answer is, in the round, yes, to that Tom. It’s our job as urologist to really nuance that piece of work. I think the work that your colleagues at the Royal Free, Maxine Tran, Ravi Barod, are leading around biopsy. We need to use biopsy a lot more than we are doing.

I mean, it’s a contravention of the rules of surgical oncology that we don’t biopsy patients with renal masses more than we do currently. There are deaths every year in our country and in the U.S. of patients who have surgery for oncocytoma and succumb to those operations. That’s just not the way we should be practicing medicine.

Brian:
How do we get there? Is it more robust biomarkers? Is it some sort of gene expression work? Because with just the biopsy with the grade and the histology doesn’t seem to have convinced people, at least broadly, to maybe do more of an observation approach. Is that right?

Grant Stewart:
Well, I think there is a bit of that. I mean, we need the experienced uroradiologists who are very good at doing biopsies, hitting the viable part of the tumor, and giving us tissue that the pathologist can interpret. That’s an important starting point.

I very rarely regret doing a biopsy and my patients very rarely come to me and say, “Why did you recommend that?” Either because it was very uncomfortable, or they had a complication. Complications are very rare. Less than 1% of people will have any form of complication that they notice.

I think that’s a piece around the hearts and minds of urologists. And remember that a decade ago, it was very, very rare to do a biopsy. We’ve had to change this thinking around pretty quickly and it’s accelerating quite nicely now, I think, in terms of the way we think about it.

Tom:
Grant, there’s a story going around about seeding. Is this a part of the moon landing conspiracy or is there actually genuinely seeding associated with biopsy and kidney cancer?

Grant Stewart:
I noticed in the EAU guidelines, there’s a sort of little cohort study of seven patients that seeded and little anecdotes with each one of them. But, and so personally, I am not aware of any patients that seeded in my practice. But again, it comes down to expert radiology.

The use of a coaxial needle where you can have multiple passes without going through the skin is very important. I think you’re probably right, Tom. But once again, we don’t have the high-level data of hundreds of patients that were in a clinical trial. They were observed very closely where some had biopsy and others didn’t.

Tom:
Of the thousands of biopsies, I’ve sat through in MBT, literally there’ve been thousands, I’ve never really seen one. So, it’s quite complicated. With clear cell renal cancer, I think that there are obviously one the tumors where it does occur, but …

Brian:
Grant, is there a robust translational work around these biopsies? Are there trials or efforts to move that forward? Because I think if there were more robust biomarker data around renal biopsy, I think probably the movement would be towards doing it if it were really going to affect decision making.

Grant Stewart:
The first thing is getting those samples, Brian. You know what it’s like. Sometimes, the amount of viable tissue is very small. We both do neoadjuvant trials, and we’ll often get our research biopsies recalled for the diagnosis of a particular histological subtype rather than them being able to do it, use it for research. So again, it comes down to very good radiology, very good pathology.

And it comes down to the confidence of the radiologist to sample the tumor multiple times for research purposes, as well as the diagnosis. These are things we are doing in Cambridge with our neoadjuvant and window of opportunity studies.

Again, it’s just part of that culture change to make sure that it’s done well enough, that you can then do single cell association. You can do single cell RNAC metabolic and all this sort of stuff. It’s maturing and improving, but still a bit of work to be done.

Tom:
Brian, our 20 minutes is up. This has been fantastic. Brian, do you have a question or we’re good to go?

Brian:
No, this was great. Thanks for the update, Grant. Really appreciate it.

Tom:
It’s nice to chat to you. Thanks so much for your time today.

Grant Stewart:
Thanks for having me. Cheers, guys.