The Role of Nephrectomy with Immunotherapy in Metastatic Renal Cell Carcinoma - Dimitrios Makrakis
February 29, 2024
Dimitrios Makrakis presents a significant meta-analysis exploring the role of cytoreductive nephrectomy in patients undergoing immune checkpoint inhibition for metastatic renal cell carcinoma. This study, a substantial effort involving 2,319 patients, sought to address the gap in understanding post-CARMENA, focusing on outcomes in the era of immune checkpoint inhibitors. The analysis, utilizing individual patient data (IPD) for a deeper, more nuanced examination, reveals that patients who underwent cytoreductive nephrectomy alongside immune checkpoint inhibitors exhibited notably improved survival outcomes. This research underscores the potential benefits of cytoreductive nephrectomy in select patients with favorable risk profiles, advocating for randomized trials to further clarify its role in contemporary treatment paradigms.
Biographies:
Dimitrios Makrakis, MD, Jacobi Medical Center, Albert Einstein College of Medicine, Bronx, NY
Daniel George, MD, Medical Oncologist, Professor, Departments of Medicine and Surgery, Duke Cancer Institute, Duke University, Durham, NC
Biographies:
Dimitrios Makrakis, MD, Jacobi Medical Center, Albert Einstein College of Medicine, Bronx, NY
Daniel George, MD, Medical Oncologist, Professor, Departments of Medicine and Surgery, Duke Cancer Institute, Duke University, Durham, NC
Read the Full Video Transcript
Daniel George: Hi, I'm Dr. Dan George from Duke, and it's my pleasure on behalf of UroToday here at GU ASCO '24 to introduce you to my guest today, Dimitrios Makrakis from Jacobi Einstein Hospital in New York City, who just presented his poster on behalf of his co-authors on a meta-analysis with individual patient data on the role of nephrectomy, specifically cytoreductive nephrectomy, in patients on immune checkpoint inhibition. Dimitrios, walk us through a little bit your data and why you did this study and what you found.
Dimitrios Makrakis: Thank you for having me. It's a great pleasure to be here. We were looking at the recent evidence on cytoreductive nephrectomy in targeted therapy for metastatic renal cell carcinoma. Then we realized that we are not sure there's enough data and information on the role of CN in patients that are treated with immune checkpoint inhibitors. We decided to do this meta-analysis that uses individual patient data, we're going to talk about this, to gather the evidence and see where we stand currently on the information we have for this.
Daniel George: Okay, great. Just to put this in context, a few years ago we had the CARMENA data that showed that, hey, you know what, now in the setting of really effective targeted therapy, maybe there isn't necessarily a need and role for cytoreductive nephrectomy. Tell us a little bit about your methodology and what you found here.
Dimitrios Makrakis: We conducted a meta-analysis after a systematic review, of course, in the classical sense. We took the individual patient data approach. So far, most meta-analyses that come out, they use what we call aggregate patient data (APD). The idea is that in aggregate patient data, you have abstract numbers. You have metrics, such as medians, means, you have hazard ratios, but you don't have access to the data at the individual patient level. Now, the IPD offers a much more granular approach to that and has some very good advantages. Some of them are, you have access to the full dataset, you can reconstruct all the different datasets into a common pool.
Daniel George: Mm-hmm.
Dimitrios Makrakis: You can run your own analysis on those. You can validate the analysis of the authors. You can also look at the groups and identify possible biases and errors that APD meta-analyses may not pick up. This was the rationale. There are some experts right now in statistics that actually have said that we should move away from the APD and move towards the IPD. Now that we're submitting more and more complete datasets with our papers and journals, this is becoming easier. I think it's a very good time to do this.
Daniel George: That's great. This really dovetails with a larger use of real-world data, but in a situation like this where there are so many different variables, it's helpful to have this kind of methodology to harmonize across a lot of different settings, this data. That's great. How big a study was this?
Dimitrios Makrakis: It was a big study. We started in the classical way. We screened three big databases: MEDLINE, Embase, and Web of Science using two independent reviewers. We screened for abstract and title to see if studies were relevant or not. The ones that were picked, we looked at the actual text. Our inclusion criteria included retrospective studies, possible controlled trials. We excluded case series and case reports. We also wanted papers to have Kaplan-Meier curves because we actually took the data and reconstructed common curves. We screened several thousand papers and we eventually went down to eight studies, all of them were retrospective. It gives you an idea of the fact that there is data, but not plenty of it. With these, we conducted the analysis.
Daniel George: Interesting. Great.
Dimitrios Makrakis: We ended up with 2,319 patients, which is a hefty cohort.
Daniel George: Absolutely. That's a lot of patients. These are patients all that underwent either they all had ICIs.
Dimitrios Makrakis: Exactly.
Daniel George: But some underwent cytoreductive nephrectomy, and some did not. Did they all get the cytoreductive nephrectomy first, or did some of them get it kind of on therapy?
Dimitrios Makrakis: No, they got it on therapy. They started with ICIs, and then at various times, and we can talk a bit about this, the cytoreduction.
Daniel George: Got it. Top-level data, what did you find?
Dimitrios Makrakis: Initially looking at the dataset, because that's an important thing, and I want to stress it out, the two groups were balanced in terms of numbers. The patients that received cytoreduction were 1,200, 1,264 to be exact, and then 1,055 patients that received immune checkpoint only. Now, when we started looking at different parameters like age, IMDC scores, the histologies, the metastatic disease burden, we started seeing some differences that were statistically significant. Patients that got cytoreductive nephrectomy and ICIs were younger, relatively younger, they had either low or favorable IMDC scores while the others were having more poor risk IMDC scores. They had a smaller metastatic disease burden. They were more likely to have clear cell histology, and their performance status was also better.
Daniel George: Mm-hmm.
Dimitrios Makrakis: These were important numbers that will actually come up later, so put an asterisk on that. Then when we proceeded to reconstruct the survival data, we saw that the patients that received cytoreduction and checkpoint inhibitors, they did have better survival outcomes. This was statistically significant in several analyses. In the simple analysis that we did and the landmark analyses, two of them that we performed for six months and 12 months to control for immortal time bias, this was consistent. We also did something that only the IPD can actually allow you to do, which was subgroup analysis. We looked at patients that received checkpoint inhibitors as a first line and then did or did not receive cytoreduction. That was a big plus of the IPD.
Daniel George: Okay, great. Great. What did you find? Did the people get cytoreductive therapy, do better, did worse? Was there mixed results?
Dimitrios Makrakis: Yeah, we did find that they do better. It was significant. Also, when we conducted a survival expectancy analysis, we saw that in the one year, the survival expectancy is better by one month for the ones that receive cytoreduction. When you go to three years, it actually jumps to seven months. You're expected to live seven more months if you have received cytoreductive nephrectomy.
Daniel George: Interesting. The longer these patients survive, the greater return on that cytoreductive therapy?
Dimitrios Makrakis: Yeah, I guess you can say that, yes.
Daniel George: Yeah, interesting. Were there any particular surprises in these findings? Was there any group that maybe you thought was going to do well and didn't, or vice versa?
Dimitrios Makrakis: When we looked at the other two groups and we saw all these differences in the demographics, the better risk factors that the patients who received cytoreduction had, we realized that probably these were not two very well-balanced groups, which brings us to the main limitation of the study and also a big strength of the IPD analysis, that it can give you insight into these pitfalls. You can know that, "Okay, my data may have some biases that will affect my results in the very end." The second limitation is also that all these studies are retrospective and they are subject to all the biases of retrospective analysis that a prospective trial wouldn't have. Yeah, I don't think that was a huge surprise because all these patients, they had factors that were more favorable risk factors.
Daniel George: Got it. Got it. Let's talk a little bit about the biases. I think these are important. This is real-world data.
Dimitrios Makrakis: Yeah.
Daniel George: It's not a randomized trial. There should be biases. As clinicians, we should be thinking about who we think might benefit from this nephrectomy or maybe shouldn't get this nephrectomy. Help us understand how we can take what we've learned from this study now and apply it to our everyday practice.
Dimitrios Makrakis: This is real-world data, and it wasn't that surprising to me that people who have a more favorable risk profile, they get nephrectomy more often. They're the patients that can tolerate it, it's safer to do it. Eventually, they're kind of selected by a natural process. Now, how we can think of applying these results into an everyday clinical practice, I think the study definitely hints at a possible role for cytoreductive nephrectomy in patients that have a favorable risk profile. Now, that being said, these results need to be interpreted cautiously. I think they stress the fact that we need randomized trials for the role of cytoreductive nephrectomy in patients that have received checkpoint inhibitors. We just need to do those trials and see what happens.
Daniel George: Absolutely. Absolutely. I think that's really the key takeaway point, but in addition to recognizing there's a role today, even without this data, as we await those trials to get conducted and mature, there's a role today for cytoreductive nephrectomy. Maybe not across the board, but in these selected patients, particularly with favorable risk factors.
Dimitrios Makrakis: Exactly.
Daniel George: Well, Dimitrios, thank you so much for conducting this, sounds like an incredible body of work that you and your colleagues have put together. Really helpful for those of us in the field to have some data to lean on and to talk with patients about, as to why we are or are not recommending this important but potentially risky procedure in these patients. Thank you so much for this. Any last parting thoughts on your study that you'd like to share with the audience?
Dimitrios Makrakis: No, thank you for having me. It's been a great honor to be here. It's been an absolute privilege to work with this team. Dr. Msaouel, Dr. Eshaghian, and Dr. Karam, they're all amazing colleagues. It's been a great opportunity. I hope we have more soon to bring up on this subject.
Daniel George: I'm sure you will. You've got a great future ahead of you. Thank you so much.
Dimitrios Makrakis: Thank you so much.
Daniel George: Hi, I'm Dr. Dan George from Duke, and it's my pleasure on behalf of UroToday here at GU ASCO '24 to introduce you to my guest today, Dimitrios Makrakis from Jacobi Einstein Hospital in New York City, who just presented his poster on behalf of his co-authors on a meta-analysis with individual patient data on the role of nephrectomy, specifically cytoreductive nephrectomy, in patients on immune checkpoint inhibition. Dimitrios, walk us through a little bit your data and why you did this study and what you found.
Dimitrios Makrakis: Thank you for having me. It's a great pleasure to be here. We were looking at the recent evidence on cytoreductive nephrectomy in targeted therapy for metastatic renal cell carcinoma. Then we realized that we are not sure there's enough data and information on the role of CN in patients that are treated with immune checkpoint inhibitors. We decided to do this meta-analysis that uses individual patient data, we're going to talk about this, to gather the evidence and see where we stand currently on the information we have for this.
Daniel George: Okay, great. Just to put this in context, a few years ago we had the CARMENA data that showed that, hey, you know what, now in the setting of really effective targeted therapy, maybe there isn't necessarily a need and role for cytoreductive nephrectomy. Tell us a little bit about your methodology and what you found here.
Dimitrios Makrakis: We conducted a meta-analysis after a systematic review, of course, in the classical sense. We took the individual patient data approach. So far, most meta-analyses that come out, they use what we call aggregate patient data (APD). The idea is that in aggregate patient data, you have abstract numbers. You have metrics, such as medians, means, you have hazard ratios, but you don't have access to the data at the individual patient level. Now, the IPD offers a much more granular approach to that and has some very good advantages. Some of them are, you have access to the full dataset, you can reconstruct all the different datasets into a common pool.
Daniel George: Mm-hmm.
Dimitrios Makrakis: You can run your own analysis on those. You can validate the analysis of the authors. You can also look at the groups and identify possible biases and errors that APD meta-analyses may not pick up. This was the rationale. There are some experts right now in statistics that actually have said that we should move away from the APD and move towards the IPD. Now that we're submitting more and more complete datasets with our papers and journals, this is becoming easier. I think it's a very good time to do this.
Daniel George: That's great. This really dovetails with a larger use of real-world data, but in a situation like this where there are so many different variables, it's helpful to have this kind of methodology to harmonize across a lot of different settings, this data. That's great. How big a study was this?
Dimitrios Makrakis: It was a big study. We started in the classical way. We screened three big databases: MEDLINE, Embase, and Web of Science using two independent reviewers. We screened for abstract and title to see if studies were relevant or not. The ones that were picked, we looked at the actual text. Our inclusion criteria included retrospective studies, possible controlled trials. We excluded case series and case reports. We also wanted papers to have Kaplan-Meier curves because we actually took the data and reconstructed common curves. We screened several thousand papers and we eventually went down to eight studies, all of them were retrospective. It gives you an idea of the fact that there is data, but not plenty of it. With these, we conducted the analysis.
Daniel George: Interesting. Great.
Dimitrios Makrakis: We ended up with 2,319 patients, which is a hefty cohort.
Daniel George: Absolutely. That's a lot of patients. These are patients all that underwent either they all had ICIs.
Dimitrios Makrakis: Exactly.
Daniel George: But some underwent cytoreductive nephrectomy, and some did not. Did they all get the cytoreductive nephrectomy first, or did some of them get it kind of on therapy?
Dimitrios Makrakis: No, they got it on therapy. They started with ICIs, and then at various times, and we can talk a bit about this, the cytoreduction.
Daniel George: Got it. Top-level data, what did you find?
Dimitrios Makrakis: Initially looking at the dataset, because that's an important thing, and I want to stress it out, the two groups were balanced in terms of numbers. The patients that received cytoreduction were 1,200, 1,264 to be exact, and then 1,055 patients that received immune checkpoint only. Now, when we started looking at different parameters like age, IMDC scores, the histologies, the metastatic disease burden, we started seeing some differences that were statistically significant. Patients that got cytoreductive nephrectomy and ICIs were younger, relatively younger, they had either low or favorable IMDC scores while the others were having more poor risk IMDC scores. They had a smaller metastatic disease burden. They were more likely to have clear cell histology, and their performance status was also better.
Daniel George: Mm-hmm.
Dimitrios Makrakis: These were important numbers that will actually come up later, so put an asterisk on that. Then when we proceeded to reconstruct the survival data, we saw that the patients that received cytoreduction and checkpoint inhibitors, they did have better survival outcomes. This was statistically significant in several analyses. In the simple analysis that we did and the landmark analyses, two of them that we performed for six months and 12 months to control for immortal time bias, this was consistent. We also did something that only the IPD can actually allow you to do, which was subgroup analysis. We looked at patients that received checkpoint inhibitors as a first line and then did or did not receive cytoreduction. That was a big plus of the IPD.
Daniel George: Okay, great. Great. What did you find? Did the people get cytoreductive therapy, do better, did worse? Was there mixed results?
Dimitrios Makrakis: Yeah, we did find that they do better. It was significant. Also, when we conducted a survival expectancy analysis, we saw that in the one year, the survival expectancy is better by one month for the ones that receive cytoreduction. When you go to three years, it actually jumps to seven months. You're expected to live seven more months if you have received cytoreductive nephrectomy.
Daniel George: Interesting. The longer these patients survive, the greater return on that cytoreductive therapy?
Dimitrios Makrakis: Yeah, I guess you can say that, yes.
Daniel George: Yeah, interesting. Were there any particular surprises in these findings? Was there any group that maybe you thought was going to do well and didn't, or vice versa?
Dimitrios Makrakis: When we looked at the other two groups and we saw all these differences in the demographics, the better risk factors that the patients who received cytoreduction had, we realized that probably these were not two very well-balanced groups, which brings us to the main limitation of the study and also a big strength of the IPD analysis, that it can give you insight into these pitfalls. You can know that, "Okay, my data may have some biases that will affect my results in the very end." The second limitation is also that all these studies are retrospective and they are subject to all the biases of retrospective analysis that a prospective trial wouldn't have. Yeah, I don't think that was a huge surprise because all these patients, they had factors that were more favorable risk factors.
Daniel George: Got it. Got it. Let's talk a little bit about the biases. I think these are important. This is real-world data.
Dimitrios Makrakis: Yeah.
Daniel George: It's not a randomized trial. There should be biases. As clinicians, we should be thinking about who we think might benefit from this nephrectomy or maybe shouldn't get this nephrectomy. Help us understand how we can take what we've learned from this study now and apply it to our everyday practice.
Dimitrios Makrakis: This is real-world data, and it wasn't that surprising to me that people who have a more favorable risk profile, they get nephrectomy more often. They're the patients that can tolerate it, it's safer to do it. Eventually, they're kind of selected by a natural process. Now, how we can think of applying these results into an everyday clinical practice, I think the study definitely hints at a possible role for cytoreductive nephrectomy in patients that have a favorable risk profile. Now, that being said, these results need to be interpreted cautiously. I think they stress the fact that we need randomized trials for the role of cytoreductive nephrectomy in patients that have received checkpoint inhibitors. We just need to do those trials and see what happens.
Daniel George: Absolutely. Absolutely. I think that's really the key takeaway point, but in addition to recognizing there's a role today, even without this data, as we await those trials to get conducted and mature, there's a role today for cytoreductive nephrectomy. Maybe not across the board, but in these selected patients, particularly with favorable risk factors.
Dimitrios Makrakis: Exactly.
Daniel George: Well, Dimitrios, thank you so much for conducting this, sounds like an incredible body of work that you and your colleagues have put together. Really helpful for those of us in the field to have some data to lean on and to talk with patients about, as to why we are or are not recommending this important but potentially risky procedure in these patients. Thank you so much for this. Any last parting thoughts on your study that you'd like to share with the audience?
Dimitrios Makrakis: No, thank you for having me. It's been a great honor to be here. It's been an absolute privilege to work with this team. Dr. Msaouel, Dr. Eshaghian, and Dr. Karam, they're all amazing colleagues. It's been a great opportunity. I hope we have more soon to bring up on this subject.
Daniel George: I'm sure you will. You've got a great future ahead of you. Thank you so much.
Dimitrios Makrakis: Thank you so much.
