Read the Full Video Transcript

Pedro Barata: Hi there and welcome. My name is Pedro Barata. I'm a GU medical oncologist at Seidman Cancer Center, Case Western University in Cleveland, and I'm very, very happy to be joined by Dr. Andrew Hahn. Dr. Hahn is an assistant professor of the GU Medical Oncology Department at MD Anderson Cancer Center in Houston, and he's a friend and colleague, and he's been doing a lot of great stuff in the kidney cancer world. So Andrew, welcome.

Andrew Hahn: Thank you for the invitation, Pedro. I'm excited to be here and excited to have this discussion.

Pedro Barata: Absolutely. It was a pleasure to see you at this KCA back in Austin. One of the studies that I found really interesting was your presentation on cytoreductive nephrectomy for patients with metastatic sarcomatoid or rhabdoid RCC treated with immunotherapy-based regimens. So I found that very, very interesting.

Andrew Hahn: Thank you. Thank you.

Pedro Barata: I would like, maybe I'll start, I mean I know you put over 90 patients on that study and actually quite interestingly enough, I think about two thirds or so got nephrectomy. So that made me think, and I'm sure audience might be thinking the same way is, can you tell us a bit more about who were those patients as far as risk group, as far as RCC subtypes, and maybe what kind of IO they received, too?

Andrew Hahn: Yeah.

Pedro Barata: Actually a good portion of them are getting debulked nephrectomy.

Andrew Hahn: Thank you for the question. So you're spot on. 157 total patients who had either sarcomatoid, rhabdoid dedifferentiation, or the combination of having both sarcomatoid and rhabdoid dedifferentiation present. Those patients retrospectively collected coming from here at MD Anderson or at Memorial Sloan Kettering. Dr. Kotecha and the colleagues added in a large number of patients to the group as well.

And so when you're looking at the group, a majority of patients had sarcomatoid dedifferentiation present, 78 of the 157. 49 had rhabdoid dedifferentiation present alone, no sarcomatoid dedifferentiation present. And then 30 had actually both sarcomatoid and rhabdoid RCC. And so if you're looking at kind of WHO/ISUP grading, all of these patients are WHO/ISUP grade 4. For primary histology present, the overwhelming majority of the patients had clear cell RCC with sarcomatoid and/or rhabdoid dedifferentiation in the background. About 89% had clear cell. If you actually look, pretty much all of the rhabdoid alone had clear cell, and that's been our experience is it's rare to see rhabdoid dedifferentiation outside of clear cell.

About 65% of patients had no prior lines of therapy, and then most of the remaining patients had one prior line of therapy before receiving immune checkpoint therapy. The majority of patients had intermediate or poor risk disease similar to what you'd be seeing in other cytoreductive nephrectomy status studies.

Then in terms of type of immunotherapy that patients received, 40% received nivolumab plus ipilimumab, a CTLA-4 inhibitor, about a third of patients ended up getting immunotherapy with a TKI, and then actually about 28% of patients got immunotherapy alone. And I think that just reflects the change in practice over time where you saw the initial introduction of nivo as a monotherapy before you started seeing all the doublets come down the pipeline. So you're kind of seeing that whole gamut there.

Pedro Barata: Got it. This is very, very helpful Andrew, and maybe that's a great opportunity for me to highlight what I think how the study started. Because I think you presented this at ASMO, and it sounds like you were able to expand the cohort, getting Sloan on board, which is amazing, because you actually asked these important clinically relevant questions by getting a fairly larger number of patients, if you will, and that's awesome. So kudus to you because that's a lot of we do is collaborations, so that's amazing. Kudos to you for doing that-

Andrew Hahn: Thank you.

Pedro Barata: ... And for promoting that as well. So I guess, let me ask you. I know what the results are, but can you summarize the findings regarding the role of the bulk nephrectomy for the study cohort?

Andrew Hahn: Yeah. I mean, so main takeaway is we did not see a significant difference for cytoreductive nephrectomy either improving treatment duration on immune checkpoint therapy, so kind of time from beginning immune checkpoint therapy to discontinuation, for any reason. The hazard ratio for that was right around one very inconclusive 95% confidence interval. Yeah, we did see a two month longer median treatment duration in the group who got cytoreductive nephrectomy, but it was not statistically significant at all and was very inconclusive.

For overall survival, we again did not see a significant improvement in overall survival with cytoreduction nephrectomy. But there's a little bit of nuance when you're looking at that group because when you look at the Kaplan-Meier curves, there is a pretty visible separation between the two groups. And if you look at the median reported overall survival between the two groups, median overall survival in the cytoreductive nephrectomy arm was about 30 months. In the group without cytoreductive nephrectomy, which was a small number of patients, 39 total, that was 17 months. And the hazard ratio did favor having a cytoreductive nephrectomy, but again was not statistically significant.

Pedro Barata: Got it. That's very, very, very important. It's interesting, because I think in your original presentation we did see a benefit favoring those who got surgery. Maybe that's a good segue for my next question, which would be did you get the chance to explore dataset, the role of upfront nephrectomy versus deferred nephrectomy? Because some would argue, some time of systemic therapy as induction, allowing us to select the good actors compared to bad actors, and then thinking about local control later on could be a very thoughtful strategy. I mean, is that something you were able to explore in your dataset?

Andrew Hahn: Yes, we were. And there's a lot to unpack in your question, so I'll try to unpack it in a couple parts so we may have to come back and forth.

So the first one is you're absolutely correct and you've done your homework. At ESMO, when we had about 90 something patients from MD Anderson, we saw a significant improvement. There's one big asterisk next to that, though, which is how we did the analysis, and it wasn't published at the time. But we did not account for immortal time bias in that analysis. And immortal time bias, meaning that when a patient is starts immune checkpoint therapy and you're looking at overall survival, for patients who have a delayed cytoreductive nephrectomy, getting at your question, they are immortal compared to patients who've had upfront or no cytoreduction nephrectomy until the time that they have their cytoreduction nephrectomy. They can't die during that time period by definition of having a delayed cytoreduction nephrectomy.

So we ended up, I did all this working with Dr. Misal and we ended up accounting for this in this new analysis where we actually, we added more patients in. We thought we were going to find this really powerful signal, and then lo and behold we didn't see significant improvements when both, accounting for lead time bias, and adding in another institution and another way of doing it.

With the question of upfront versus delayed cytoreductive nephrectomy, what type of benefit we're seeing, so actually the majority of our patients who had reductive nephrectomies underwent an upfront cytoreductive nephrectomy. And not to give away too much, but there is definitely a difference in institutional approach at the time. I do think what you've described is what you're seeing us do, and actually I think when we talk with Sloan, they are doing the same approach now too, which is you're looking. You have a very effective therapy for patients with sarcomatoid disease. You're looking for that really nice response. And then I know our group's published retrospective data that when you see that 10% or more response, those patients are the ones that are likely to derive benefit from the cytoreductive nephrectomy approach.

So we ended up, we did the same question from an immortal time bias. We did a subsequent analysis where we just excluded patients who underwent a delayed cytoreductive nephrectomy. That's a different approach that kind of cuts down on your total number of patients. And actually when we did that, we saw a stronger signal favoring cytoreductive nephrectomy. It was still not statistically significant, but the hazard ratio for OS went from 0.79, a 21% improvement, to a 39% improvement, 0.61, which was surprising to us as well. When we actually took out the delayed patients, we saw the signal get stronger, and we're not certain what to make of that.

I think finally we all believe, and this is a message that we're finalizing the manuscript as we speak, and we're trying to convey this, is that you're clearly seeing that there are patients deriving benefit from cytoreductive nephrectomy who have this very poor prognosis aggressive disease. And it's even at institutions like Sloan and Anderson where you have the highest volume surgeons in the world, there's still difficulty in selecting those patients out. So there's a lot of work to do there.

Pedro Barata: Great, fantastic. That's very helpful, Andrew. Thank you for that. And maybe one other question, last question perhaps. So you are selecting, in your study, you are selected the good actors, right? Because you're taking the sarcomatoid rhabdoid tumors and then you offer them an IO-based approach, which as we know, increases the chances of success. But at the same time, on your baseline characteristics, you have patients with intermediary, even poor risk patients.

So I guess my question to you is the fact that you're selecting based on one marker, does that overcome what we historically know that it's not a good idea to think of surgery for patients with more than one risk factor, for example, although that data derives from obviously the TKI you ran. So if you are thinking of sarcomatoid, if you will, and now is that still a role for the bulk nephrectomy? Can you say that based on your dataset, even if you have a poor risk patient for example? What are your insights from what you were able to analyze?

Andrew Hahn: Yeah, I think when you dive into the dataset, you see, yes, there likely is a role in well-selected patients for cytoreduction nephrectomy, even in patients with poor risk, and I think we've all seen those patients in clinic. And I think that was actually what brought us into this study initially was the idea that sarcomatoid de-differentiation, it's totally changed how we think and approach those patients in clinic with the introduction of nivolumab plus ipilimumab with CR rates approaching 20%, objective response rates over 60%. And these patients have really durable responses when they have responses. But oftentimes, the last place we're seeing responses is in the kidney. And so if you have kind of residual disease in the kidney, it ends up just being very natural to take those tumors out. And I think that's where you're likely seeing some benefit.

Now, the question of clone production and if you do an upfront cytoreductive nephrectomy or you're removing some resistant subclone production, I don't know. And I think there's a lot of work that needs to be done in there in terms of exploring the biology behind what's happening from the primary tumor versus distant sites when the primary tumor is in place and with these aggressive features.

Pedro Barata: Got it. And I guess final, promise you, final, final question, did you understand, did you identify any signals depending on the type of IO-based approach patients received?

Andrew Hahn: No.

Pedro Barata: Because the reason I'm asking is we get more primary tumor responses with when a TKI is involved compared with IOIO. I'm just wondering your thoughts on that.

Andrew Hahn: It's a really good question. We did not see any treatment effect heterogeneity in our multi-variable analysis when we incorporated dedifferentiation status, looking at sarcomatoid, rhabdoid, or none, as well as looking at type of therapy received. So, we didn't see any difference there. Ultimately this is retrospective. The numbers still end up being small. And for us, I think the total number of cytoreductive nephrectomy patients is pretty big when you compare to historical precedent. It's actually our no cytoreductive nephrectomy group that's smaller, and that's kind of how we ended up in this analysis. So I think that's a place for improvement on our end.

Pedro Barata: Got it. That's awesome. Of course, we're going to know more in a perspective level, because these questions are actively being addressed, not specifically to my knowledge for tumors with sarcomatoid or rhabdoid features. But I'm just thinking, for example, probe asking this question in a prospective manner about role of debulked nephrectomy post-induction. And there are others out there. I know there's one from Mark Stein as an IT kind of investigated the same thing, asking the same question. So, Andrew, this has been amazing. Thank you so much. Do you have any final thoughts before I let you go?

Andrew Hahn: I think final thought is just kind of reinforcing what you said, which is I think our data is thought provoking, it's interesting. What it does is it points to the need for everyone to kind of contribute to the probe trial and the importance of the probe trial in answering some of these questions. And the beauty of that is is it's a cooperative group. You're going to have high quality GU pathologists reviewing that and you're going to have faith in what they're finding from their sarcomatoid subsets, which I'm sure they're going to be investigating when they're releasing that data. So that'd be my final thought and thank you for having me on, Pedro. This has been a lot of fun.

Pedro Barata: No, this was very amazing. And by the way, good luck with the submission. I'm sure it's going to land as a very good paper, so I'm looking forward to read the full manuscript. Thank you so much for joining us and I'll see you soon.

Andrew Hahn: Awesome. See you soon. Thanks, Pedro.