Assessing the Potential Cost-Effectiveness of the Addition of Atezolizumab to First-Line Platinum Chemotherapy in Advanced Urothelial Carcinoma - Ali Khaki
June 11, 2020
Ali Khaki, MD, Hematology/Oncology Fellow, Fred Hutchison Cancer Research Center, University of Washington, Seattle, Washington, United States
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.
Petros Grivas: Hello, I'm Petros Grivas. I'm a medical oncologist at the Seattle Cancer Care Alliance and Associate Professor at the University of Washington and Fred Hutchinson Cancer Research Center. I'm very pleased today to be joined by Dr. Ali Khaki. He's a superstar hematology oncology fellow, pursing a fellowship at the University of Washington and Fred Hutchinson Cancer Center. Ali, welcome.
Ali Khaki: Thank you, Petros.
Petros Grivas: We had a fantastic virtual meeting at ASCO, and I know you had a very active presence there. So, congratulations for that. And I want to focus our discussion on one of the posters you presented specifically regarding the cost-effectiveness of the chemotherapy atezolizumab combination in the frontline setting of advanced urothelial cancer. So before we go to specific details about this very nice and elegant poster, I want to ask you to give us a little bit of a landscape about the frontline setting of advanced urothelial cancer, how patients are getting treated, and then what led to the IMvigor130 trial, and eventually your poster.
Ali Khaki: Yeah, sure. Excellent question. So, for over 20 years now the first-line setting for metastatic or advanced urothelial cancer has been sort of cisplatin-based chemotherapy, which has very high response rates, but unfortunately for many patients the durability of that response is not very long. And more recently we've actually had many great breakthroughs with new medications. Most notably the checkpoint inhibitors, where we have five of those achieved regulatory approval in the U.S. And pembrolizumab was shown to be better than second-line chemotherapy in one of the KEYNOTE-045 trial. And so naturally the next question became, could the combination of chemotherapy with immunotherapy be better than the sequential use of those agents, especially because the toxicity profiles are complimentary. They're not too overlapping. And many patients don't go onto subsequent lines of therapy after platinum-based chemotherapy.
So for all these reasons, I think the IMvigor130 trial was a natural next step. And that basically investigated the combination of atezolizumab and the immune checkpoint inhibitors with platinum-based chemotherapy. The way the trial was initially designed was going to be in the cisplatin-ineligible setting. So it was looking at carboplatin-based chemotherapy with atezolizumab versus chemotherapy alone. They eventually liberalized the criteria and made it a three-arm trial, which included sort of platinum-based chemotherapy atezolizumab, atezolizumab alone, as well as platinum-based chemotherapy alone.
Petros Grivas: Thank you, Ali. Very nice to hear the context here, it's very important for the audience. So we know that the results of the IMvigor130 clinical trial, a very important trial in the frontline setting. I saw that the chemotherapy plus atezolizumab combination met the primary endpoint of improved, prolonged progression free survival. That was a median differential for about two months. You can argue whether it's clinically meaningful or not, but there was no significant overall survival benefit with a combination of which chemotherapy and placebo. So many of us commented that this is a fantastic trial to report, design, and conduct. However, it did not change practice because of the lack of overall survival benefit. So in this benefit-risk discussion and balance here, I think you had a significant interest to evaluate the cost-effectiveness of this approach of chemotherapy in atezolizumab combination. So the question is, how do we think about cost-effectiveness in this setting? And what methodology did you use in order to ask this important question?
Ali Khaki: Yeah, so I think you summarize it nicely. I think that the combination of atezolizumab with chemotherapy did sort of improve progression-free survival by about two months, but no sort of improvement in overall survival. And that sort of led to not many people adopting this as a standard of care practice. So I think that from that perspective, you could question whether the cost-effectiveness study needs to be done or not, but I think that there was still this anticipation that there could be an overall survival benefit in the future as the data matured. And this could become something that is frontline. Or if it's a cost-effective measure, then maybe even if it's not as clinically meaningful, it's something we could investigate in treating some patients with. And so we thought that there would still be value in looking at sort of what the cost-effectiveness of this combined approach would be compared to chemotherapy alone.
So the way we afforded with studying this as we use partitioned survival model, which is a three-state model commonly used for cost-effectiveness, especially in cancer research. And it basically consists of a state of progression-free, or at the beginning of treatment phase, and then a state for progression patients who have progressed on therapy, and then a state for death. And basically using the data from a clinical trial, we can calculate what proportion of patients are in each of these three states. And based on that, we can sort of calculate the costs associated with each state, as well as the health utility of each state, and estimate what the cost-effectiveness would be.
Petros Grivas: That's very interesting, Ali. And I know you have a significant interest in cost-effectiveness studies and outcomes research end-of-life care overall. Let me ask you regarding the question in principle here. Is there a literature benchmark, how we measure cost-effectiveness just for the audience? And what would constitute, in your opinion, meaningful difference, or at least a benchmark met for cost-effectiveness in that study?
Ali Khaki: Yeah. So to answer that question, I'll tell you a little bit about sort of what the output from a cost-effectiveness study can be. And so, typically we can measure any sort of outcome that's measurable in a cost-effectiveness model, but the common ones oftentimes in cancer again are life years and quality adjusted life years, as well as sort of the cost. And then using those outputs, we can calculate what's something called incremental cost-effectiveness ratio, the ICER. And that basically is just the cost per life-year, the cost per quality-adjusted life-year. And the benchmark sort of in terms of what would be considered cost-effectiveness, I think that varies based on different settings. Different countries, different societies have put different values on that. And some would argue that you can never put a value on a human life. But I think historically in the U.S. that we have sort of identified a willingness-to-pay threshold of what would be sort of an appropriate cost-effective intervention of $150,000 per quality-adjusted life-year.
And this sort of dates back to sort of the old dialysis literature, when the country was initially considering whether to pay for patients who need dialysis and that's where that willingness-to-pay threshold has come from. And there's a broad discussion in the cost-effectiveness literature, whether that's an appropriate threshold or not. Our threshold in the U.S. is much higher than in many other countries. And so some people would argue that's too high and others would argue it's too low. And so I think that there's a lot of discussion about that willingness-to-pay threshold. But as of this time right now, I think the most widely agreed upon one still in the U.S. is about $150,000 per quality-adjusted life-year.
Petros Grivas: Thank you, Ali. This is a very important context. So tell us a little bit more about the resolve specifically and how this will lead us to the next step related to this important research.
Ali Khaki: Yeah. So, for our three health states ... I'll tell you a little bit more about the methods and then we'll tell you about the results. We estimated our health utility for the different states, the progression-free, and the progression state based off the KEYNOTE-045 and KEYNOTE-052 data. And I mentioned that because those are not exactly the same setting as first-line cisplatin-eligible patients that are invested in investigating the IMvigor130 trial. But they were the best approximation that we have from the literature of what those health utility states are. And those health utility states help inform sort of how we calculate the quality adjusted life years. And so it's important to keep in mind that there is a little bit of estimating that we're using there from the literature.
But what we found is that both in terms of costs per life-year and cost per quality adjusted life-year, atezolizumab with chemotherapy was not cost-effective at the $150,000 willingness to pay threshold. And specifically, we calculated an incremental cost-effectiveness ratio, $629,755 per life-year, and $895,800 per quality-adjusted life-year.
What we also did is that a natural next question is our estimates are based on the annual sales price that's published by Medicare and Medicaid services. But oftentimes, health companies and either health insurance companies will have rebates. And so that may not be actually the price that the health system is paying for it, or patients are paying for it. And so what would be the reduction in the cost to make this quote, to make this cost-effective? And we calculate that cost-effectiveness ratio. So for the cost for quality of adjusted life-year $150,000 willingness-to-pay threshold, we would need a 33% reduction in the cost of atezolizumab chemotherapy combination.
Petros Grivas: Ali, I'm just thinking about this type of research which I think is very important when we think about decision making in clinical practice. We take into account efficacy, safety, level of evidence, and convenience, of course for the patients, frequency of administration. Cost-effectiveness has not been traditionally a widely accepted way of thinking, I would say in clinical practice and at the same time research has recently become, I think there's no doubt about it, more important. The question is how would you translate this into next steps? And how would you envision this work to impact future studies, especially in the context of new data coming up in the frontline setting and recently presented data at ASCO like the JAVELIN switch-maintenance trial?
Ali Khaki: Yeah. So, I think that as we said earlier in our conversation, this combination of chemoimmunotherapy was not something that had sort of adopted to become standard of care because of this modest improvement in being only a progression-free survival benefit only. I think that our data sort of further supports that sort of assessment with it being a pretty costly intervention with not much in the way of improvement in outcomes for our patients.
And I think that this is especially, I feel even stronger saying that because the JAVELIN Bladder 100 trial was just so impressive in its results that were presented as a plenary at ASCO, where we saw a significant improvement in overall survival for patients who were treated with avelumab switch-maintenance after frontline chemotherapy. And so it's an interesting approach because that one you spare patients the mixture of toxicity of chemotherapy with immunotherapy at the same time.
You sort of minimize the amount of medications they are getting all at once. And at the same time, you're seeing an overall survival benefit in that setting. And so I think that our data, as well as the JAVELIN Bladder 100, all support right now as of June of 2020, the switch-maintenance approach as being sort of a new standard of care for frontline of bladder cancer treatment.
Petros Grivas: I think it's really exciting to see this research going on. I agree with you, Ali, that this has been becoming more important and cost is actually going to impact more, in my opinion. I'm sure you agree we see this down the road, and this is already happening even more in Europe with the approval of an agent could potentially take to account costs. Something that is not happening in the U.S. I think the FDA is doing a fantastic job in approving agents based on merit and efficacy. At the same time, the cost is not part of that discussion. It's not part of the decision.
Ali Khaki: And I think now, in the same way that we sort of value the cost-effectiveness of the IMvigor130 trial results, I think that as we now have results in the JAVELIN study, and I think that a similar approach could be taken in that setting to look at sort of how cost-effective that measure is. And it'll be interesting to see what those results look like. And I think that might be something we work on as a next step.
I think that the other thing I mentioned about cost-effectiveness is that especially in models like the one we use that are projection models, there are limitations of that approach because you're limiting yourself to sort of the costs that are published in the literature, which may not actually be the true cost of any of the therapies. As I mentioned, there are oftentimes many rebates that sort of go into place in real practice. So with that, you could have an overestimate of the cost of an intervention.
At the same time, some of the measures you're putting into the model also can have their own limitations in terms of health utility, especially if you use literature-based health utility assessments. The good thing is the JAVELIN study and also the IMvigor study had their own quality of life adjustments that were calculated. And so I think that as those get reported out, we'll get more and more information about what is the health, the quality of life, and the different arms. And I think that can lead to a more rigorous assessment of the cost-effectiveness of these interventions.
Petros Grivas: I share with you the enthusiasm about this work. I also share with you the enthusiasm about the practice-changing data of the JAVELIN Bladder 100 trial, which were presented by Dr. Tom Powles at the ASCO Virtual Meeting 2020. Very, very exciting data. And to your point, so hard to generate practice-changing evidence and having overall survival benefit in the Phase II study, especially in urothelial cancer. So with that, I would like to thank you so much for your great work that you're doing all along in your career so far, and looking forward to more work from you, and hopefully having you in UroToday again in the future. Thank you, Ali.
Ali Khaki: Thank you, Petros.