The Uromigos Episode 148: Paper of the Month—Stereotactic Radiotherapy for Advanced Renal Cancer

Chad Tang discusses his oncology paper recently published in Lancet.

Episode Transcript

Brian:
Hey, welcome everyone to our Paper of the Month podcast. Tom and I are here with Chad Tang from MD Anderson Cancer … Published an interesting paper about the role of radiotherapy and the management of oligometastatic renal cell carcinoma.

Chad, welcome. Why don’t you introduce yourself? Use whatever titles you’d like, and then give us a little, short summary of what you did and what the rationale was and maybe a couple of the high-level results.

Chad Tang:
Sure. Thank you. This is awesome to be on this podcast with our urologic friends here. I’m a big fan, so I’m really, really excited.

Tom:
You’re joining us. We’re super excited, too.

Brian:
“Urologic friends” is the English translation of Uromigos, Tom. Just so you know.

Tom:
Rather than Spanish, we should change the name to Urologic Friends. I think that’s much better.

Brian:
Anyway, go ahead, Chad.

Chad Tang:
Yeah. So, I am a radiation oncologist at MD Anderson Cancer Center. I specialize in GU malignancies, of which I see predominantly prostate and a large number of kidney cancer, increasingly. We did this one trial, me and actually, Pavlos Msaouel, who’s my kind of co-conspirator in this as basically fellows coming out of our training. We got together over coffee one day, and we’re like, well, wouldn’t it be neat if we did radiation therapy instead of systemic therapy for these metastatic patients, being all naive. Then, we thought of big grand trials to design, and this got pretty much whittled down to a 30-patient trial that we decided to be a feasibility/kind of phase II trial.

In this trial, we tested whether instead of doing systemic therapy, which is really often an alternative in these patients that are coming to our clinic, whether we could just sequentially do local therapy with a twist, that it would just be radiation as local therapy.

We know that patients get sequential mastectomies all the time, but sequentially radiation therapy in, and then try to keep them off the drugs for as long as possible. We presented this, and we decided we should try it as feasible because, at that time, there’s a lot of pushback that radiation would even work in renal cell carcinoma. This was 4 years ago.

We then built into it sequentially as we do just some PFS endpoints. Then we looked into this with this approach. In the beginning, we enrolled 1 or 2 patients, but over time, it kind of blew up a little bit, and then when you add the second cohort and more patients.

But the first cohort, as presented here basically showed it was quite feasible. We could treat sequentially with radiation therapy, often multiple sites. We learned a lot and we thought that radiation, generally, was well tolerated. It had favorable progression-free survival and very nearly all of our … [crosstalk 00:03:11]

Tom:
Which patients did you pick, Chad?

Chad Tang:
For the most part, these were patients that were on, actually, surveillance after … no treatment surveillance after their nephrectomies, or they would be coming off of systemic therapy and then something grew. Then we would be wondering, hey, should we be doing local therapy or systemic therapy?

Tom:
Patients with local progressive disease?

Brian:
No.

Chad Tang:
Basically, they would have 1 or 2 little lung nodules because they’re all metastatic, and they’d be watching them for a long time or doing mastectomies. Then it would grow again, and they would do a biopsy usually. Then we would say, okay, let’s try radiation in this patient who’s, for the most part shown … and this high-selected group of patients that they’ve been pretty indolent, not spreading disease everywhere. That’s usually in lieu of what we would’ve done for these patients, which would’ve been, in these patients, specifically, systemic therapy for many of them or mastectomy.

Brian:
Chad, I have a question about sort of patient selection, which is, I think, what Tom’s getting at. So, in the CONSORT diagram, it talks about sort of 45 patients at the top, and then that whittled down to 30. But, in terms of patient characteristics, as you’re saying, this is clearly a highly selected group.

Chad Tang:
Yes.

Brian:
Four years since nephrectomy, the vast majority IMDC, zero, 1 risk factor, 40% had had prior local therapy, and two-thirds had only 1 site of mets. Do you have an idea of how many advanced kidney cancer patients at MD Anderson over that same timeframe, and does this represent 2% of patients, 5%, 10%? I mean, in our observation work, I’ve always said that I think it probably applies to 10% or 15% of patients, so that’s kind of what I’m getting at is what’s the real denominator here?

Chad Tang:
Great question. In the beginning, I mean, we very rarely got these patients, but they started becoming a lot more frequent later on as a trial progressed as we saw we could do this. But, to answer your question specifically, probably the denominator is probably, yes, 10 times this at least.

We do see a lot of kidney cancer patients at MD Anderson, and it is a selected group, and I must say not all med-ons … because we have a lot of med-ons who do kidney cancer … MD Anderson have a hundred percent bought into this. So, we’re not probably even getting a hundred percent of the group enrolled. But, yes, it’s selected group of patients that have demonstrated, over time, good biology. Yes.

Brian:
You said these are often patients who were you’re sort of watching and have small volume disease and then maybe 1 or 2 area grows, right. Those are the patients who you then treated. Is that accurate?

Chad Tang:
Exactly. So, these, for the most part, were patients … There are 2 camps of patients. One is those who are kind of watching and, and once they’re of a site growing, it’s really hard to tell a patient, hey, we’re going to watch you some more because sites are growing. The patients are biting at the bit to do something. That’s one group.

The second group of patients is those in which we have low volume disease on systemic therapy, which is a minority of patients, and the toxicity is getting too much for them. They still have sites for disease, so they’re stopping it. We’re watching them with the interim scan before their enrollment, and then we’re treating them and keeping them off systemic therapy. Those are kind of the 2 main groups.

Tom:
Difference between those pretreated groups and the untreated group?

Chad Tang:
I don’t think it is. I think that, honestly, those patients who were on systemic therapy and then came off it had low volume to begin with, and we could have done it before. But they just happened to be on systemic therapy when they came into the door.

Tom:
Chad, any difference between VEGF-targeted therapy and immune therapy. I kind of guess your numbers are too small. Is that right?

Chad Tang:
It is. Most of them were [inaudible 00:06:57] honestly.

Brian:
I have a question about sort of rate of growth. You said these patients, whether on systemic therapy or not, had some sort of growth prompting the intervention, right … prompting the radiation. I know there was a … I think a bar graph in your supplementary appendix about sort of growth before and after, but these patients have low-volume disease, so the percentages might be misleading.

I guess I’m wondering how much absolute growth … I mean, if a lung nodule went from 7 to 8 millimeters, is that growth, and would you radiate that, or was there some sort of threshold to say, well, gee, this has grown enough that it’s actual growth, and we should treat it?

Chad Tang:
We weren’t rigorous in that we mandated a certain amount of millimeter growth before we did it. It was generally not taking off, as you said, so it would be a couple millimeters, 2 or 3 millimeters to a lung nodule or a lymph node. It wasn’t that much growth, but there was usually growth before we started treatment.

Tom:
Chad, is it reasonable to say that radiating small lesions may not alter the whole systemic disease?

Chad Tang:
That is very reasonable to say, Tom. I totally agree, and it may not. But kidney cancer’s a funny one, right, where we’re doing nephrectomies on patients, and that’s supposedly the increases their OS in an old data and the … right.

I want to say that local therapy may stem for dissemination, and I felt like that those data with nephrectomy is we’re kind of indicating that potentially. Why else would they be causing an OS benefit in those older data? But now we’re going to basically reduce the disease burden by maybe another log order, and the thinking is we may be stopping dissemination and doing all sorts of other fun things maybe with the immune system.

Brian:
I don’t want to dwell on debulking nephrectomy, but I think the difference is when you’re doing a debulking nephrectomy, and you’re removing a big chunk of tumor, right, and patients ultimately die from an accumulating tumor burden, right … a lethal tumor burden. So, radiating an 8 here and an 11-millimeter lung nodule there, that’s not a lethal … That tumor burden isn’t anywhere close to being lethal, but removing a 13-centimeter renal mass, I think they’re just very different.

Chad Tang:
But we can debulk again and again and again, right. We can radiate them so that, by conventional imaging, by our best imaging, we make it so it’s only microscopic disease. So, we’re probably taking down a couple logs in terms of [crosstalk 00:09:40].

Tom:
I mean, Chad, you’re really into this by the sound of things.

Brian:
Eventually you’ll turn patients to dust, though. You can’t radiate every single met that pops up, right. I’m just saying, I think there’s a difference, biologically, between removing a big primary and separate radiation to small mets. That’s my only point.

Tom:
Chad, I wanted to talk about the radiation technology you’re using and how that’s changed in the last 5 to 10 years, and what technology …

Chad Tang:
We have all the toys over here in our institution, all sorts of different advanced imaging. We have CT-on-rails with MR-Linacs. We have conventional cone beam CT of protons, so we’ll use whatever we think is most appropriate to deliver the least dose possible but still cover the tumor. I could get more specific, but it might get you bored if I go into that.

Tom:
Which do you use? Which type was most commonly used in the trial?

Chad Tang:
Generally, for lung, we generally used SBRT with what we call a cone beam setup, which is these little arms that come out, and we get a pretty reasonable CT scan because it’s on a lung. For abdominal tumors, we’ll use something called a CT-on-rails. We have a full-on CT scanner attached with a rail system to a Linac that we’ll use to basically make sure the bowel’s not in the way. Then for things a contralateral kidney lesion, which we didn’t have in this group, or, increasingly now, ITP thromboses, we’ll use MR-Linacs. We’ll use the right tool for the right indication. I use them all pretty much as needed.

Brian:
Are those available broadly at academic institutions only or at most major hospitals, just to get a sense of feasibility of applying this broadly?

Chad Tang:
For lung nodules, for the cone beam CT, that’s the standard order variant system. You should be able to get that anywhere. CT-on-rails is a little bit of specialty, and same with the MR-Linac. Those are probably only available on a handful of places. You probably can get away with a cone beam CT just fine to treat abdominal lesions, and I would do it. But we have these toys, and I can get much better visualization to see even finer …

Brian:
Sure.

Tom:
There were 2 papers shortlisted for Paper of the Month. There was your paper, and there was also an HLA genotyping paper, which was shortlisted for this particular month. I read the paper and feel I’d like to ask you questions about that anyway.

Chad Tang:
Yes.

Tom:
People tell me that if we irradiate lesions, we can stimulate the immune system and generate an abscopal effect and have a magic effect around the body. Is that something that you bought into?

Chad Tang:
We all kind of bought into this in the beginning to a certain degree. We did some trials in this at MD Anderson, which I was a part of, and we were not amazed with the results. Now, is there some sort of immune stimulation that happens with radiation potentially? Is that enough to involve gross disease? Not in the majority of cases, probably. Is it enough to help hinder microscopic disease that has not built up a microenvironment that’s conducive to that? Maybe. I don’t know, but we are going to do a lot of translational work on these patients where we do have their blood and their tissues collected.

Tom:
Do you know I’ve got …

Chad Tang:
I’ve been working on that right now.

Tom:
I would love to look at ctDNA or methylation … Even the methylation signatures … I’d love to see how much it goes down. I mean, my concern around this, speaking frankly … I’m pretty negative about going off to isolated lesions with a systemic progression. I agree completely in renal cancer there is a subgroup, maybe 1 in 10, maybe 1 in 20, maybe 1 in 30 patients that have these isolated problems where it seems like systemic therapies overkill and, less particularly in the [inaudible 00:13:43], where we weren’t curing these patients.

With immune therapies where [inaudible 00:13:47] cure more patients, it becomes more debatable, and I’m happy to have that debate. My concern has always been that where is the tipping point between what needs systemic therapy and what you can get away with surgery or radiation therapy.

What I would love to see as brought with the nephrectomy issue is if you are giving radiation therapy, the ctDNA, or the methylation signature by 90%, you can say, listen, we are making a big difference here whereas if you’re doing that and actually no effect, it comes back to that theory about, well, what actually … and you can see something on a scan, and you go and hit it. That’s not as sophisticated as saying we’re looking at the systemic picture and by targeting that [inaudible 00:14:37], we [inaudible 00:14:39] biology of the disease.

Brian:
I mean, I think what you’re saying, Tom and Chad, with your input is I think if they’re giving local therapy and treating the majority of the disease, you feel better about it, right, i.e., the ctDNA lowers, et cetera. If there’s a lot of subclinical disease lurking and you do local therapy, and then 3 months later, 4 more lung nodules appear, you probably haven’t helped that patient, right. That’s the issue.

Maybe that’s the question to you, Chad, is we can’t apply this to everyone, so how do you select patients? How do you know … I’m sure there are patients you’re helping and others where giving the radiation may not alter their disease course. Even if you delay their systemic therapy, it’s just really not altering the biology. Do you have any idea of truly selecting patients for this approach?

Chad Tang:
That is a million-dollar question and the subject of pretty much all of my grant proposals based on this same thing. We need biomarkers, and we think radiology is crude. So ctDNA is a wonderful idea. I just don’t know … and we do collect a fair number of Streck tubes with that intent in mind. We just do not have a very good ctDNA panel to apply, in my opinion.

Tom:
I’ve got a panel, Chad. I’m up for it. I’m up for it. Let’s touch base after this. I’ve got a brilliant idea. You’ll love it.

Chad Tang:
All right. Let’s touch base. Yes. Collaboration in action. I love it.

Brian:
So, you have more sophisticated biomarkers, even just clinically, because I think you’d agree … Like you say, this might, if the denominator was 10X, so it’s 3 to 5 percent of patients you might consider for this … rough numbers. How do you find those patients? Even among those 3 to 5 percent, it’s probably only a subset of those where you’re really truly altering their disease course. Even clinical tools or anything you might use to select …

Tom:
I would love to see a radiology picture. I’d love, on top of that, where we stratify patients into high, intermediate, and low, and I’d love to see a systemic test looking at the … not, obviously, the IMDC score with all respect to Danny. But not that score. I’d love to see a blood test looking at systemic burden of disease, and I’d love to pull those two together and then pull out a subgroup of patients who have both visible disease that you can target that’s kind of measurable of what’s going on in the body. That indication that you need to go in and potentially address that target. I can see, actually, that sort of model being really, really exciting for the future. I’d love to work on that.

Brian:
No, I’d love to see it, too, Tom. I was asking Chad … Tomorrow, when you see somebody, [crosstalk 00:17:17].

Tom:
Can I …

Brain:
I wasn’t asking you to … [crosstalk 00:17:19]

Tom:
I think you’ve hijacked the podcast … [crosstalk 00:17:21]

Brian:
Tomorrow, when you see somebody in clinic … I mean, is there a cutoff number of lesions, size of lesions, location? I mean, there’s just so many clinical factors to consider. Do you have a framework in your mind? I’m not sure. I think you kind of know it when you see it, but I’m just maybe asking you to sort of elaborate on what’s that clinical framework on whom you might apply this?

Chad Tang:
I think the best indicator is basically a CT scan done before, 3 months earlier, off the systemic therapy and one done now. You compare the 2 scans, and you get a goody good picture of what’s the trajectory of this patient. If he’s popping up lesions everywhere, then bunch of new lesions in between, like 1 to 5; 1 to 3 months ago, 5 now, that’s probably a bad patient, even though they still have 5 lesions or less.

Tom:
Chad, I think …

Chad Tang:
But if they have one … Yeah.

Tom:
I think that’s terrific advice. I mean, I remember Tim Eisen. I was in one of these GU education sessions maybe 10 or 15 years ago, and it was one of those sorts of cases here. There was this really complex conversation. Tim, who, as you know, is super bright, just stopped the conversation and said listen, everyone, the right thing to do here … The patient’s not in harm’s way. Just repeat the CT scan in 2 months’ time. If there are 5 new lesions, we’re done. We know where we are. But if this is the only lesion, we can then attack it from there. I think you’re absolutely right. I always remember that conversation that Tim had. It sort of stopped the whole debate, and it was really incisive. I think you’re absolutely right.

Chad Tang:
Let’s be really fair about this. This is a 30-patient, phase II trial. It’s going to open the door, I hope, to more exploration in this matter, which I think is sorely needed. Let’s also be very frank in that this is not a pharmacy-funded trial. We scraped this together with zero money, and basically just got by with whatever we have right now. All right, let’s be very frank. This was the university-sponsored study where we basically scraped together the money with a department grant here and department grant there. We’re able to get 30 patients from kind of bootstrapping the funding. It’s not a pharmacy-sponsored study where you get to enroll a couple hundred patients.

I think it’s important to do these kinds of studies and to compare them towards the pharmacy-funded strategies potentially so that we can, in the right patients, get these patients on a de-escalation therapy.

Tom:
Listen, Chad, this is really important. Investigator-initiated trials like this are hard to do. We’ve all done them, and these are the hard yard. It’s much more difficult to do this than being involved in a big global pharmaceutical trial. These are the hard yard trials. I know it’s only 30 patients, but each one of those patients takes up a huge amount of effort from an institution, and you guys need to be congratulated for that.

I’ve been critical of radiation therapy trials, and I continue to be critical of them, but you could say that for the whole of investigator-initiated trials. We haven’t done terribly well in terms of practice changing investigator-initiated trials in kidney cancer over the last 5 to 10 years. They’ve mainly been pharma-sponsored trials. I think this is a first and very important step, and the fact it’s published in Lancet Oncology is testament to your hard work and the importance of the work because there’s so little of it. If there were hundreds of these trials, it wouldn’t have ended up in Lancet Oncology. It’s because there’s not enough of these trials, and we need to see more of these studies. I don’t know how to-

Chad Tang:
Because they’re practice impacting, right. These are the patients we see in clinic, right.

Tom:
Absolutely, and, at the moment, we haven’t got any [inaudible 00:20:47] to support this, so I just say don’t do it.

Brian:
Well, don’t say that. I guess maybe my last question is Chad, what do you think about … I could make an argument that instead of radiating all these small lesions, I’m just going to wait. I’m going to let them grow, and then, at some point, I’m going to start them on regimen X. Do we know that intervening earlier with stereotactic radiation is better than just waiting and starting systemic therapy with a higher tumor burden?

Chad Tang:
Yeah. That’s a great question.

Brian:
Tom, that’s 3 great questions by me this podcast.

Tom:
I wasn’t counting …

Brian:
Three.

Tom:
I’m not sure that’s right.

Brian:
Go ahead.

Chad Tang:
You can go back and count, but what I was … so these patients, they have growing lesions or a new lesion potentially, and they’re not going to be … and we don’t know when to stop the surveillance in these patients. So, they’re going to want some … They’re not going to be happy you saying you grew, let’s watch it further.

Brain:
Well, hold on. Hold on. I’m going to stop you right there.

Chad Tang:
Okay. Sure.

Briand:
If I have a patient who has a lung lesion grow by 2 or 3 millimeters, I routinely tell those patients, hey, great news, your lung lesion only grew by 2 millimeters. We’re going to watch you, and patients are more than happy to do that. So put aside patient desire, right, because I think we can argue that one patient wants it, another doesn’t. As you well know, you can talk them in or out of most of waiting or not waiting.

Tom:
No great question there, Brian. I got notice.

Brian:
No, so it’s not … Let’s say the patient will do whatever you say, right?

Chad Tang:
Yeah.

Brian:
So, again, how do you know that intervening on that 2 millimeters of growth is doing something?

Chad Tang:
How we know is we have to do a phase III trial, right. We are going to have to test this versus the frontline therapy because they’re well selected and you just go back to historical standards and you’re like, oh, they do better because you selected a good patient. We’re never going to answer that question by looking at the 30-patient or a hundred patients if I enroll them on a single arm trial. We’re going to have to do a phase III trial where we compare this first front line at the OS, and that’s …

Tom:
But how do that?

Brian:
Do you think that can happen? I agree with you, and I’d love to do it. I’m on board.

Chad Tang:
Pharma’s never going to pay for that.

Tom:
We can do it. We can do it. If we all get behind it, it will get done. But we just need to all get behind it. Look, I’ve been saying for ages … I sit in many forums. I don’t [inaudible 00:23:11] radiation therapy to isolated lesions until we have more data. There are hundreds of people … What about giving yogurt and altering the … Well, the lot do, and, theoretically, that might help. Vitamins … All sorts can make a difference. What’s the data to actually support this? What we have is a machine. It looks like it can make the lesions smaller. It looks promising, but there’s just not enough data. We need to generate that, and we need to do these randomized trials. It’s really important. I’d love to try and help with that.

Chad Tang:
Yeah. No, that’s what needs to be done, I think, to answer your question, Brian. Otherwise, I wave my hands, and you’re going to just smack them down every time.

Brian:
I think it is possible. I think it is possible because, again, you could make systemic therapy dealers’ choice where it doesn’t have to be drug X or Y or regimen or whatever. I think ensuring adequacy of SBRT and the technology and the application, right. That’s always a challenge in radiation trials. It’s the technical equivalence and you know way more about that than I do. But I don’t think it’s impossible, and I think there’s enough equipoise in the community. I mean, we get a lot of traction around the observation data, and I think people still observe people a lot. These are, obviously, overlapping populations, right. Observations, and that’s …

Tom:
I’m sure I can [crosstalk 00:24:26] my patients. I’m sure we can get …

Chad Tang:
I totally agree. We’ve actually had a little Twitter debate on this and the endpoint that we … Endpoints are really hard for such a trial. You have to probably go for OS because what does PFS mean? I mean, just radiate a new spot? Then we were like, well, PolyMet progression, but what does that mean? So, it’s OS or nothing.

Brian:
I think that’s right, but difficult.

Chad Tang:
It is, but it’s very difficult.

Tom:
But, hold on. But hold on. We can put some intermediary endpoints.

We can put endpoints which actually give us the information along the way saying it’s pointing in the right direction. So, time and treating failure … You can put in identification on new lesions, or there are lots of endpoints that we can … and we can allocate a tiny amount of alpha, likely to … [crosstalk 00:25:12]

Chad Tang:
If you’re going to use sequential radiation, then what does time and treatment failure mean? It means how much until I lose my kind of cool about treating more spots, and that’s going to be different for every people. When do you start systemic therapy, I mean, in these patients? That’s going to be very dependent. That was the main criticism that we got in the reviews … Stopping local therapy and switch to systemic therapy is a provider’s choice.

Brian:
Right. As it always should be, right. That’s just what it should be.

Chad Tang:
As it always should be. I know with Brian’s primary endpoint in his trial, which was my counterargument. Brian Rini published oncology [inaudible 00:25:50], a paper that looked at [inaudible 00:25:52] therapy. So, therefore, it should be valid.

Brian:
Did that carry any weight?

Tom:
I’m very surprised.

Chad Tang:
They were like what paper? We’ve never read that.

Brian:
Well, we don’t have to design the whole trial here, but I think we agree that it’s necessary for the field. I think it’s possible although exceedingly difficult and, you’re right, pharma won’t be … [crosstalk 00:26:09]

Tom:
[crosstalk 00:26:09] initiated trials can go on for a long period of time. It’s okay. It’s okay.

Brian:
Yeah. Agreed.

Chad Tang:
For a cooperative group, but, honestly, taking this to cooperative group would be very difficult because a lot of GU cooperative groups centered … Well, NRG is very prostate heavy. So, we lose a lot … We burn a lot of our capital on getting big, big prostate cancer trials through. Kidney has not traditionally been a very radiation kind of … It’s not been as a … and then when you do [inaudible 00:26:37] or something that you all just don’t. I mean, I don’t know. They don’t really like radiation that much, I think. I feel like they kind of stomped it down a lot.

Brian:
Here we go.

Chad Tang:
Sorry. I love my urologists. I love you all.

Brian:
We’ll delete that out.

Tom:
Chad …

Brian:
I think this is …

Tom:
It could be our last podcast of the year. It’s possible.

Chad Tang:
Wow.

Brian:
It probably will be.

Chad Tang:
Thanks, guys.

Tom:
It probably will be. Chad, this has been terrific. Well, you have a fantastic Christmas.

Chad Tang:
Thanks, you all, too.