Petros Grivas: Hello, I’m Petros Grivas, I’m an Associate Professor of Medicine at the University of Washington and Fred Hutchinson Cancer Research Center, and Medical Oncologist at the Seattle Cancer Care Alliance. I’m very fortunate today to be accompanied in this beautiful environment at ASCO by Dr. Ali Khaki. Ali is a superstar Fellow at the University of Washington, and he is doing great work in the field. Ali, welcome.
Ali Khaki: Thank you Petros, it’s good to be here. Thank you for inviting me.
Petros Grivas: So, it’s really exciting to be in a form like ASCO with so many datasets and findings that potentially can change the way we treat cancer down the road. I would like to focus our discussion specifically on the work you are presenting, with this very interesting poster looking at the performance status and response to immune checkpoint inhibition in patients with advanced urothelial cancer.
Ali Khaki: Yeah. Something that drew my interest is that as these checkpoint inhibitors have changed the landscape of treatment for urothelial cancer, there’s been more and more interest in using them in people who are not suitable for chemotherapy, as approved in the post-platinum setting, and for more vulnerable patients.
Early literature has shown that up to a third of patients with poor performance status may be getting these treatments, but the clinical trials did not include these patients. Only three of the eight trials had ECOG Performance Status of two, and no one had ECOG Performance Status worse than that. So I wanted to look further into, what are the outcomes in the real world for these sort of patients? So what we were able to do is we were able to get together 15 institutions to do a multi-institution chart review. We did a retrospective chart review of up to 400 patients, 402 patients total, and got data about their clinical history leading up to the checkpoint inhibitor use, as well as their outcomes on treatment.
Petros Grivas: Interesting.
Ali Khaki: In that cohort, we had complete data to be done for the analysis in 369 of those patients. About 52% were treated with first-line checkpoint inhibitors in the metastatic setting, and 48% were treated in the second-line or beyond in the salvage setting.
Petros Grivas: So it has half and half in the two, first and second line.
Ali Khaki: Yeah. In the majority of patients either got atezolizumab or pembrolizumab with a scattering of some of the other checkpoint inhibitors that are approved in that setting.
Petros Grivas: And those who got salvage therapy are maybe second line and beyond, right?
Ali Khaki: Correct. So the two things we want to look for in terms of clinical outcomes are the response rate, as well as survival. And while we didn’t see any differences in response rate, or we saw numerical differences but they were not statistically significant. So in the first line, it was 28% for those with good performance status, and 24% for those with poor performance status. And in the second line, salvage setting, it was 25% compared to 20%. Again, slightly higher, but not statistically significant.
What we did see, is we did see in the first line patients did have worse survival if you did have a poor performance status. So checkpoint inhibitor therapy is not necessarily salvaging our vulnerable patients and helping them live as long as patients with good performance status. That’s my major take-home from that first part of the analysis.
Petros Grivas: Interesting. So it sounds like if you have a patient with a very poor performance status, let’s say an ECOG PS of three, in your opinion, it’s unlikely to change the trajectory of disease by using checkpoint inhibition.
Ali Khaki: Exactly.
Petros Grivas: And it’s hard, of course, to predict what will happen with the individual patient, but it’s informing your decision-making process, it sounds like.
Ali Khaki: Exactly, yeah. So it would have been very helpful if we saw that the response rate was worse. From that perspective, we could help guide who should not get treatment, in the same way that we don’t give chemotherapy to patients who have poor performance status. We don’t have that strong of evidence from that perspective for this population, and that underscores the need for more and more biomarker work or other sorts of prognostic tools to be developed to help guide clinicians in that decision-making process.
Petros Grivas: Do you think that either worse survival vs. the response rate, either or both are relevant in the context of this discussion, especially with all the confounding factors for those patients?
Ali Khaki: Right, yeah. I mean I think that they are both relevant. I think that it helps guide that conversation. I think it’s a much more nuanced conversation now. And so if you have a patient for example who is young and is very very motivated, even with poor performance status is might be someone who you’re like, “Okay this is still something worth trying.” But having that sort of measured discussion of, “The outcomes aren’t necessarily as great for you if you’re pretty unfit moving into this.” But if you have an older patient who is pretty vulnerable already, and is not up for continued hospitalizations or continued treatment of any sort, this might be the time where you use this information to guide that patient to moving towards hospice at an earlier time.
Petros Grivas: So, like everything we do it’s some incremental knowledge and information that can help inform some individual discussion with the patient. Of course, it cannot be definitive or conclusive for each individual patient, but it can definitely help this informed discussion that can happen.
Ali Khaki: Exactly.
Petros Grivas: Do you want to talk a little bit about the end-of-life healthcare utilization, cost of care, and what are the findings from the study in that regard?
Ali Khaki: The comments I’m making about the difference between the older and younger patients are informed by that second half of our analysis. And so the next thing we did is we looked at utilization patterns for patients who died in the study. So about 215 of our patients were deceased at the time of data collection, and we actually had the site of death being hospital vs. not hospital for 140 of those patients.
So we asked the question of, how many of those patients who died were treated in the last 30 days; in the last 90 days of life? And what we saw is about 10% in both the 215 population and the 140 population were treated in the last 30 days of life, and about 35% were treated in the last 90 days of life. And when I’m saying treated, I mean the new initiation of treatment. And so we’re looking at starting this new line of therapy in the last 30 or 90 days of life. What we saw is that for those patients who died… we know the site of death, so that 140 patient population, there was an increased odds of you dying in the hospital if you had initiation in the last 30 days of life. That was statistically significant. So it suggests to that point I was making about, if you’re doing this you might end up having a more aggressive death. You may not be at home with your friends and family in a peaceful way. It’s just a small signal but sheds some light onto what those patients’ experiences may be.
And the other thing I was interested in, as the cost is becoming a bigger and bigger part of our healthcare system, is estimating how much are we spending on checkpoint inhibitors for these patients who are dying soon thereafter starting on this treatment. So, we estimated that using basically the average wholesale price, which is one of a number of different ways you could estimate this. So we got that published value and calculated what that cost would be per patient based off the type of treatment they got, being whichever medication they got, and how many cycles they received that last 30 or 90 days of life. We estimated that $1400 were spent per patient in the last 30 days of life, for these patients who were new initiations, and $2800 was spent for people who were treated in the 90 days, with a new initiation in the last 90 days of life.
Petros Grivas: That’s very very informative, and I think it’s very, I would say, insightful data because cost-effectiveness and value-based care is an everyday topic of discussion. And having this, even early data I would say can definitely be informative and very interesting to take into account when you have a patient in front of you, you want to do the best for the patient. And that takes into account, of course, the cost and expectations this patient has. But having it in your mind some context about the healthcare utilization and cost I think are important, especially healthcare systems that may not sustain significant cost and an unfavorable cost to value ratio.
Ali Khaki: Exactly. And if we are able to identify biomarkers or prognostic tools that help identify who is or is not going to benefit, this would really help underscore that reason of deciding not to treat people who are not going to benefit all the more.
Petros Grivas: And that’s important, and goes back to the urgent need for predictive markers that can actually help tease out which patients may benefit more from others. And in that current era, we need more biomarkers with proven clinical utility, prospective validated, especially in the context of this data.
Ali Khaki: Exactly, yeah.
Petros Grivas: And of course, in addition to the cost of drug acquisition right, to buy as a practice, a particular chemotherapy, immunotherapy agent, it is also the cost of managing side-effects. We did some work before regarding evaluation of the cost estimates for managing side-effects from chemotherapy and immunotherapy in patients in the first line suffering from advanced urothelial cancer, and we found some interesting results, I don’t know if you’ve had the chance to take a look at that.
Ali Khaki: Yeah, I don’t know if I have the exact numbers, but it was very insightful to see. We often talk about how beneficial checkpoint inhibitors are, and how well-tolerated they are, but I think we under-appreciate the toxicity, especially the immune-related adverse events that lead to hospitalizations and IV treatments, or even sometimes initial procedures to get diagnosis. And I think that you showed in particular toxicity, like pneumonitis and colitis, what was the extent of the cost, again with an estimate based off the in-patient hospitalization time, and some of these other costs that were associated with that. Which is very useful information because again I think the rhetoric and the narrative around these medications is how great they are, but we have to have a more measured perspective on what we’re dealing with here.
Petros Grivas: And we definitely need this data, both for chemotherapy and immunotherapy as you mentioned, and we saw corresponding costs for chemotherapy-related adverse events: dehydration, acute kidney injury, infection, sepsis, and so forth, and immunotherapy. So having these data are important because it can inform decisions among multiple stake-holders, how we are going to develop trials and prospective data to inform our decisions and share those data with those stake-holders and the patients because I think it’s relevant to our clinical practice. So moving on overall in your, I would say estimate, about next steps in the field of outcomes research, value-based care, cost-effectiveness, how do you envision that the field will move forward? How do you see the research in this particular field evolve, especially with the discussion of choose wisely, select wisely from ASCO, and other value-based care prepositions?
Ali Khaki: I think there’s a few different things that are interesting. One is identifying those low-value practices, that’s what choosing wisely does. That’s what some of the attempts of my work here was to do, is to look at things that if we know that they clearly are not benefiting patients, we can help guide clinicians and the practice of medicine away from those practices. So I think that’s one major point of it. Another big thing is how are we paying for things in general? And so I’ve been doing some work trying to compare the multi-insurance healthcare system in the US to the single-payer system in Canada, trying to quantify what are the differences in cost. Again, I have a lot of interest in end-of-life care patients and the different markers of utilization in that patient population, it’s very vulnerable.
Petros Grivas: Interesting. So, because I know it’s your personal interest, do you see this field growing in the future and across tumor types?
Ali Khaki: I think so, I think that healthcare is a growing cost for everyone, not only for those who are sick but also for those who are well, as it becomes a larger portion of our GDP for one thing. And then at our institution, there’s been other faculty who have reported extensively about the financial toxicity of a cancer diagnosis specifically, leading to increased risks of bankruptcy, and the harms both in terms of treatment adherence, as well as personal stress and burden that it’s all associated with. And so I think that as we better understand the larger harms, the larger toxicity that our treatment has on patients, it’s imperative that we begin to develop new ways of understanding this, and then hopefully finding ways so we can curb these practices.
Petros Grivas: Absolutely, I fully agree with you. And there is great work being done by Dr. Shankaran that you mentioned, Dr. Lampe and Dr. Cooper in that regard. And I think having these discussions in an open and transparent way is very very critical. Quality of care, outcomes, research, value-based care, sometimes may not be the first thing that we think in the clinic, but overall as a healthcare system are important cost durations. And having research tools to measure those metrics can go a long way.
Ali Khaki: And if I see a patient in my clinic and I tell them, “I want you to live longer and I want you to live better.” But I put them into bankruptcy, I don’t think that’s making them live better by any means. And so I think that we have to consider that we may help people live longer, but if we’re leaving them in a life debilitated by bankruptcy, then that’s not a quality of life that we want to leave them with.
Petros Grivas: You introduce a new term, called financial toxicity, which is important and cannot be neglected, and has to go with access to care. Many patients may not be able to have access to care. And of course, the healthcare disparities issue is an important component of research, and of course of consideration on how we can move the field forward and improve outcomes on the global level.
Ali Khaki: Yeah. And I’m excited both that ASCO as well as the work that’s being done at the HICOR Institute, as part of the Fred Hutch, to help begin to move that forward. Both in terms of financial disparities but also race disparities, geographic disparities, these are all important things that we hopefully better understand and make increased access for all our patients in terms of standard-of-care treatments, as well as clinical trials.
Petros Grivas: I fully agree with you, and the work being done in the Fred Hutchinson Cancer Research Center, University of Washington, and Seattle Cancer Care Alliance as a clinical cancer center I think is very relevant in that direction, and we need more people like you who have this interest and focus. Hopefully, we will have the opportunity to talk more about that in the near future.
Ali Khaki: Thank you, yeah. It’s been a pleasure, it’s an exciting area of work to go into because I feel like it has such potential to really change patient care, but also change the state of our society and the cost that we’re all enduring, so I’m excited to keep working in this field.
Petros Grivas: And collaborate with many other leaders in the field and work together.
Ali Khaki: Exactly.
Petros Grivas: Ali, thank you so much for coming here today and spending time with us, and I would like to also thank the audience for their attention.
Ali Khaki: Thank you so much.