Clinical Outcomes With Anti-PD(L)1 Immune Checkpoint Inhibitors in Patients With Advanced Urothelial Carcinoma – Dimitrios Makrakis
February 1, 2022
Dimitrios Makrakis joins Alicia Morgans and Petros Grivas in a discussion on a BJU International publication titled The Association of prior local therapy and outcomes with PD(L)1 inhibitor in advanced urothelial cancer looking at a data set trying to understand the important potential interaction between exposure to checkpoint inhibitors and local therapy for patients at the time of diagnosis assessing biomarkers for patients that receive checkpoint inhibitors.
Biographies:
Dimitrios Makrakis, MD, Postdoctoral Research Fellow, University of Washington, Seattle Cancer Care Alliance, Seattle, WA
Alicia Morgans, MD, MPH, Genitourinary Medical Oncologist, Medical Director of Survivorship Program at Dana-Farber Cancer Institute, Boston, Massachusetts
Petros Grivas, MD, Ph.D., Associate Professor, Clinical Director of Genitourinary Cancers Program, University of Washington, Associate Member, Clinical Research Division, Fred Hutchinson Cancer Research Center.
Biographies:
Dimitrios Makrakis, MD, Postdoctoral Research Fellow, University of Washington, Seattle Cancer Care Alliance, Seattle, WA
Alicia Morgans, MD, MPH, Genitourinary Medical Oncologist, Medical Director of Survivorship Program at Dana-Farber Cancer Institute, Boston, Massachusetts
Petros Grivas, MD, Ph.D., Associate Professor, Clinical Director of Genitourinary Cancers Program, University of Washington, Associate Member, Clinical Research Division, Fred Hutchinson Cancer Research Center.
Read the Full Video Transcript
Alicia Morgans: Hi. My name is Alicia Morgans, and I'm a GU Medical Oncologist at Dana-Farber Cancer Institute. I'm so excited to have here with me today Dr. Dimitrios Makrakis, who is a Postdoctoral Research Fellow at the University of Washington and Seattle Cancer Care Alliance, and the illustrious Dr. Petros Grivas, who is a GU Medical Oncologist and Associate Professor at the University of Washington. Thank you both so much for being here with me today.
Petros Grivas: Thank you, Alicia, for having Dimitrios and me.
Alicia Morgans: Wonderful. I wanted to talk with the two of you about a recently published paper in BJU International, where you looked at a data set and tried to understand the important potential interaction between exposure to checkpoint inhibitors and local therapy for patients who had nonmetastatic disease at the time of diagnosis. I'd love to hear why you thought that this was an important piece of information to investigate and then how you performed this study.
Dimitrios Makrakis: Hello Alicia. Good morning. And it's very nice to be with you today. Thank you for having us discuss our findings. What we tried to do in this study was again, to assess biomarkers for patients that receive checkpoint inhibitors. As you know, checkpoint inhibitors come with a lot of risks for patients. We have immune-related adverse events that we always need to account for and think about in our clinical practices. And we also know today that, at least in advanced urothelial carcinoma, they have a modest effect, and responses can be modest for patients. We are constantly in the look for biomarkers, things that we can take into consideration to assess the efficacy. In this paper, we tried to assess whether a prior history of local treatment can be associated with differences in outcomes for patients with AUC that got checkpoint inhibitors.
Alicia Morgans: Wonderful. And can you tell us a little bit, Dr. Grivas, about this data set, how you put it together and how you used it to answer some of these questions?
Petros Grivas: Great question, Alicia, and thanks again for having us. It's very important obviously, to have data from clinical trials, which is the golden standard. And along with meta-analyses, is the main vehicle of how we make progress in urothelial cancer and other malignancies. At the same time, I think, both you and I recognize the importance of what we call real-world "data" that comes from patients that are not part of a clinical trial that can complement, can add to the data fields in the field of urothelial cancer. And that, I think, is a relevant point because we can take patients who are not, let say, necessarily being under the filter of a clinical trial and can give us some real-time data of how the checkpoint inhibitors work in this "real-world" setting. We reached out to colleagues from different institutions, and right now we have more than 25 centers most in the US, but also in Europe, and we have constructed a database along with our colleague, Dr. Ali Khaki, who is an Assistant Professor at Stanford University.
Dr. Khaki was a fellow at the University of Washington when he started this work about three years ago. And this idea in collaboration with Dr. Khaki has become this enormous database. Now we have about, I would say 1,300 patients. That is a huge resource for us. All these patients have received immune checkpoint inhibitors for advanced, locally advanced resectable, or metastatic urothelial carcinoma. And we can go back and mine this database and ask different questions that can be relevant to clinical practice and even directly or indirectly help our discussions when we designed clinical trials or indirectly help the discussions about decision making in the clinic. So in that context, one of the questions that Dimitrios wanted to ask was can we evaluate the impact of the prior historical local therapy, either in form of definitive radical surgery or prior definitive radiation? Two, the response [inaudible 00:04:12] counts on checkpoint inhibitors that were administered for metastatic urothelial cancer. Whether this historical therapy has any impact on the outcomes of those patients.
Alicia Morgans: And it's a very important question because certainly there can be some discretion around whether we treat that local disease and Dimitrios, can you tell us what did you find?
Dimitrios Makrakis: Okay, so Alicia, let me go through a bit about how we did this analysis. What we did, we collected data on... we started with 1,000 patients. Then we, [inaudible 00:04:46] we fell down to 554 people, and we analyzed... firstly, we compared patients that had received local treatment compared to those that received absolutely no local treatment in the past. We stratified our analysis by the line of treatments. A first-line compared to actually no first-line and then second-line or greater for checkpoint inhibitors. We adjusted our analysis for two scores. One is the well-known Bellmunt score. And the other was an internally developed score that has been developed by Dr. Khaki. And it includes some more factors such as the NLR and the presence of liver mets as well.
And what we found. What we found was that in the first-line setting, we did not find any difference between patients that had received local treatment compared to those that had not received treatment. So our results were not statistically significant. In the second line or greater, we saw a difference in patients that had received local treatment in the past, either in the form of radical surgery, which was the majority of our patients, or for some patients who received radiotherapy. We saw a difference in all outcomes, [inaudible 00:06:05] objective response rate, progression-free survival, and overall survival, we saw a difference in favor of patients that had a history of local treatment.
Alicia Morgans: That's really interesting. And I wonder Dr. Grivas, why do you think this is happening? Do you have a biological reason or maybe a difference in patients who received local treatment versus those who did not? And why would this come out, maybe in the second-line or greater setting?
Petros Grivas: That's a great question, Alicia. As Dimitrios said we try to account for different confounding factors. And in the first-line setting, we used our previously published paper by Dr. Khaki, myself, and others. And we looked at prognostic factors in the frontline setting, like neutrophil-to-lymphocyte ratio, so-called NLR, albumin, liver mets, and ECOG Performance Status, which we have [inaudible 00:06:58] seem to be prognostic in the frontline setting in patients getting checkpoint inhibition, and also the Bellmunt risk score factors in the second line and beyond setting. And despite this, I would say the adjustment for different confounding factors that can still be unmeasured confounds, right? That may be inherent to the patient's overall status, fitness, or frailty as I call it. And these factors may potentially impact the outcomes of those patients, as well as the status of the immune response.
So one potential explanation of these unmeasured confounders is that patients who already had prior local therapy prior to definitive surgery or definitive radiotherapy could potentially be more fit. And that difference in this inherent fitness could explain the discord outcomes in the second-line setting, especially when the prognosis is very poor. Those factors may influence how patients do on salvage therapy. Those could be one of the explanations there. Obviously, again, we try to account for different prognostic factors, but it's almost impossible to account for all this and confounders, especially in a prospective cohort study, like this one.
Alicia Morgans: But still very important. And by the time we get to the second-line setting, it's possible that some of those factors have washed out, because I think you mentioned earlier, at least all those patients made it to that line of therapy. There may be some biological underpinning to why we're seeing this, which I think is so important. And also one of the things that I think you and your team have done with the [inaudible 00:08:30] factors and the other risk scores is that you have a data site where you can actually go back and look at that primary data. It's very different than claims-based data, which might be done in a setting where you can't go back to the original patient-level data in the same way and pull that information from the chart. So really important that you and the team took every step to try to get to the bottom of this and to understand it and really interesting information. And as we wrap up Dimitrios, I'd love to hear your take on where we go from here and what you do with this information, perhaps in future prospective studies.
Dimitrios Makrakis: That's a really nice question, Alicia, and you mentioned something that's very important. You said washed-out factors. I couldn't, say this in a better way. Yes, in the second line, some of the risk factors that may be associated with outcomes have washed out. That's why we see that difference. And I feel this is the main take-off of our work, and maybe it should be the take-home message. Where we go from now, I think the most important thing one needs to take into consideration from these findings is the fact that patient status plays an important role in responses to treatment.
It looks like that when patients have greater frailty if we can say that, they are at high risk of not responding to treatment. I think this can be taken into consideration in clinical trials or in the clinic. The fact that we need to act early to stop the disease before it progresses a lot, and before the patients get into a status where their health status is not that good. I think this is the main take-home message, that no matter how good of medications we may have, a patient that does not have a good health status is at an increased risk of not responding.
Alicia Morgans: I completely agree. And we've seen in other real-world data sets that we do have a drop off of patients receiving therapy by almost 50% or around 50% with each line of treatment. And I think it will be important for us as clinicians to support patients and make sure that they are fit enough as we do try to give them line after line of therapy and make a difference in their disease. As we wrap up Dr. Grivas, I wonder if you could give us a summary and where we go from here with this data.
Petros Grivas: Thank you so much Alicia for having us, and we will continue mining this database to come up with different answers and important questions, trying to account for those confounders. And as I mentioned before, try to at least inform discussions regarding clinical trial designs and also discussions about clinical decision making. For example, when you are in front of a patient, does the presence or absence of prior definitive therapy, should that impact your decision? And the answer is probably not. You should probably make this decision in the clinic regardless of the prior or definitive therapy or not. And then decisions about the respectability of disease in a borderline respectable setting should probably be taken in tumor boards, multidisciplinary tumor boards, it takes into account multiple different factors.
Thanks again for having us. And I appreciate the great work by Dimitrios, Ali Khaki, and all the other collaborators from 25 institutions. It is really exciting to work with all of them. Thank you.
Alicia Morgans: Thank you so much. And thank you so much for taking the time and sharing your expertise.
Alicia Morgans: Hi. My name is Alicia Morgans, and I'm a GU Medical Oncologist at Dana-Farber Cancer Institute. I'm so excited to have here with me today Dr. Dimitrios Makrakis, who is a Postdoctoral Research Fellow at the University of Washington and Seattle Cancer Care Alliance, and the illustrious Dr. Petros Grivas, who is a GU Medical Oncologist and Associate Professor at the University of Washington. Thank you both so much for being here with me today.
Petros Grivas: Thank you, Alicia, for having Dimitrios and me.
Alicia Morgans: Wonderful. I wanted to talk with the two of you about a recently published paper in BJU International, where you looked at a data set and tried to understand the important potential interaction between exposure to checkpoint inhibitors and local therapy for patients who had nonmetastatic disease at the time of diagnosis. I'd love to hear why you thought that this was an important piece of information to investigate and then how you performed this study.
Dimitrios Makrakis: Hello Alicia. Good morning. And it's very nice to be with you today. Thank you for having us discuss our findings. What we tried to do in this study was again, to assess biomarkers for patients that receive checkpoint inhibitors. As you know, checkpoint inhibitors come with a lot of risks for patients. We have immune-related adverse events that we always need to account for and think about in our clinical practices. And we also know today that, at least in advanced urothelial carcinoma, they have a modest effect, and responses can be modest for patients. We are constantly in the look for biomarkers, things that we can take into consideration to assess the efficacy. In this paper, we tried to assess whether a prior history of local treatment can be associated with differences in outcomes for patients with AUC that got checkpoint inhibitors.
Alicia Morgans: Wonderful. And can you tell us a little bit, Dr. Grivas, about this data set, how you put it together and how you used it to answer some of these questions?
Petros Grivas: Great question, Alicia, and thanks again for having us. It's very important obviously, to have data from clinical trials, which is the golden standard. And along with meta-analyses, is the main vehicle of how we make progress in urothelial cancer and other malignancies. At the same time, I think, both you and I recognize the importance of what we call real-world "data" that comes from patients that are not part of a clinical trial that can complement, can add to the data fields in the field of urothelial cancer. And that, I think, is a relevant point because we can take patients who are not, let say, necessarily being under the filter of a clinical trial and can give us some real-time data of how the checkpoint inhibitors work in this "real-world" setting. We reached out to colleagues from different institutions, and right now we have more than 25 centers most in the US, but also in Europe, and we have constructed a database along with our colleague, Dr. Ali Khaki, who is an Assistant Professor at Stanford University.
Dr. Khaki was a fellow at the University of Washington when he started this work about three years ago. And this idea in collaboration with Dr. Khaki has become this enormous database. Now we have about, I would say 1,300 patients. That is a huge resource for us. All these patients have received immune checkpoint inhibitors for advanced, locally advanced resectable, or metastatic urothelial carcinoma. And we can go back and mine this database and ask different questions that can be relevant to clinical practice and even directly or indirectly help our discussions when we designed clinical trials or indirectly help the discussions about decision making in the clinic. So in that context, one of the questions that Dimitrios wanted to ask was can we evaluate the impact of the prior historical local therapy, either in form of definitive radical surgery or prior definitive radiation? Two, the response [inaudible 00:04:12] counts on checkpoint inhibitors that were administered for metastatic urothelial cancer. Whether this historical therapy has any impact on the outcomes of those patients.
Alicia Morgans: And it's a very important question because certainly there can be some discretion around whether we treat that local disease and Dimitrios, can you tell us what did you find?
Dimitrios Makrakis: Okay, so Alicia, let me go through a bit about how we did this analysis. What we did, we collected data on... we started with 1,000 patients. Then we, [inaudible 00:04:46] we fell down to 554 people, and we analyzed... firstly, we compared patients that had received local treatment compared to those that received absolutely no local treatment in the past. We stratified our analysis by the line of treatments. A first-line compared to actually no first-line and then second-line or greater for checkpoint inhibitors. We adjusted our analysis for two scores. One is the well-known Bellmunt score. And the other was an internally developed score that has been developed by Dr. Khaki. And it includes some more factors such as the NLR and the presence of liver mets as well.
And what we found. What we found was that in the first-line setting, we did not find any difference between patients that had received local treatment compared to those that had not received treatment. So our results were not statistically significant. In the second line or greater, we saw a difference in patients that had received local treatment in the past, either in the form of radical surgery, which was the majority of our patients, or for some patients who received radiotherapy. We saw a difference in all outcomes, [inaudible 00:06:05] objective response rate, progression-free survival, and overall survival, we saw a difference in favor of patients that had a history of local treatment.
Alicia Morgans: That's really interesting. And I wonder Dr. Grivas, why do you think this is happening? Do you have a biological reason or maybe a difference in patients who received local treatment versus those who did not? And why would this come out, maybe in the second-line or greater setting?
Petros Grivas: That's a great question, Alicia. As Dimitrios said we try to account for different confounding factors. And in the first-line setting, we used our previously published paper by Dr. Khaki, myself, and others. And we looked at prognostic factors in the frontline setting, like neutrophil-to-lymphocyte ratio, so-called NLR, albumin, liver mets, and ECOG Performance Status, which we have [inaudible 00:06:58] seem to be prognostic in the frontline setting in patients getting checkpoint inhibition, and also the Bellmunt risk score factors in the second line and beyond setting. And despite this, I would say the adjustment for different confounding factors that can still be unmeasured confounds, right? That may be inherent to the patient's overall status, fitness, or frailty as I call it. And these factors may potentially impact the outcomes of those patients, as well as the status of the immune response.
So one potential explanation of these unmeasured confounders is that patients who already had prior local therapy prior to definitive surgery or definitive radiotherapy could potentially be more fit. And that difference in this inherent fitness could explain the discord outcomes in the second-line setting, especially when the prognosis is very poor. Those factors may influence how patients do on salvage therapy. Those could be one of the explanations there. Obviously, again, we try to account for different prognostic factors, but it's almost impossible to account for all this and confounders, especially in a prospective cohort study, like this one.
Alicia Morgans: But still very important. And by the time we get to the second-line setting, it's possible that some of those factors have washed out, because I think you mentioned earlier, at least all those patients made it to that line of therapy. There may be some biological underpinning to why we're seeing this, which I think is so important. And also one of the things that I think you and your team have done with the [inaudible 00:08:30] factors and the other risk scores is that you have a data site where you can actually go back and look at that primary data. It's very different than claims-based data, which might be done in a setting where you can't go back to the original patient-level data in the same way and pull that information from the chart. So really important that you and the team took every step to try to get to the bottom of this and to understand it and really interesting information. And as we wrap up Dimitrios, I'd love to hear your take on where we go from here and what you do with this information, perhaps in future prospective studies.
Dimitrios Makrakis: That's a really nice question, Alicia, and you mentioned something that's very important. You said washed-out factors. I couldn't, say this in a better way. Yes, in the second line, some of the risk factors that may be associated with outcomes have washed out. That's why we see that difference. And I feel this is the main take-off of our work, and maybe it should be the take-home message. Where we go from now, I think the most important thing one needs to take into consideration from these findings is the fact that patient status plays an important role in responses to treatment.
It looks like that when patients have greater frailty if we can say that, they are at high risk of not responding to treatment. I think this can be taken into consideration in clinical trials or in the clinic. The fact that we need to act early to stop the disease before it progresses a lot, and before the patients get into a status where their health status is not that good. I think this is the main take-home message, that no matter how good of medications we may have, a patient that does not have a good health status is at an increased risk of not responding.
Alicia Morgans: I completely agree. And we've seen in other real-world data sets that we do have a drop off of patients receiving therapy by almost 50% or around 50% with each line of treatment. And I think it will be important for us as clinicians to support patients and make sure that they are fit enough as we do try to give them line after line of therapy and make a difference in their disease. As we wrap up Dr. Grivas, I wonder if you could give us a summary and where we go from here with this data.
Petros Grivas: Thank you so much Alicia for having us, and we will continue mining this database to come up with different answers and important questions, trying to account for those confounders. And as I mentioned before, try to at least inform discussions regarding clinical trial designs and also discussions about clinical decision making. For example, when you are in front of a patient, does the presence or absence of prior definitive therapy, should that impact your decision? And the answer is probably not. You should probably make this decision in the clinic regardless of the prior or definitive therapy or not. And then decisions about the respectability of disease in a borderline respectable setting should probably be taken in tumor boards, multidisciplinary tumor boards, it takes into account multiple different factors.
Thanks again for having us. And I appreciate the great work by Dimitrios, Ali Khaki, and all the other collaborators from 25 institutions. It is really exciting to work with all of them. Thank you.
Alicia Morgans: Thank you so much. And thank you so much for taking the time and sharing your expertise.