Comparative Analyses of Costs and Survival Between Radical Cystectomy and Trimodal Therapy - Stephen Williams
September 11, 2019
Ashish Kamat invites Stephen Williams to present on costs and survival analysis between radical cystectomy and trimodal therapy. He shares results of two studies which patients from 2002-2011 were looked at in both studies focusing on methodology/statistical analyses and the outline of outcomes. This work found that trimodal therapy had significantly worse survival, these findings persisted across all stages except stage IV, and the same direction of association in both sensitivity and propensity score analyses. In terms of cost, Stephen shares outcomes supporting increased costs and decreased survival compared to radical cystectomy.
Biographies:
Stephen B. Williams, MD, MS, FACS, Chief, Division of Urology, Professor of Urology and Radiology (Tenured), Robert Earl Cone Professorship, Director of Urologic Oncology, Director of Urologic Research, Co-Director Department of Surgery Clinical Outcomes Research Program, Medical Director for High Value Care, UTMB Health System, Galveston, Texas
Ashish Kamat, MD, MBBS, is a Professor of Urology and Wayne B. Duddleston Professor of Cancer Research at MD Anderson Cancer Center in Houston, Texas.
Biographies:
Stephen B. Williams, MD, MS, FACS, Chief, Division of Urology, Professor of Urology and Radiology (Tenured), Robert Earl Cone Professorship, Director of Urologic Oncology, Director of Urologic Research, Co-Director Department of Surgery Clinical Outcomes Research Program, Medical Director for High Value Care, UTMB Health System, Galveston, Texas
Ashish Kamat, MD, MBBS, is a Professor of Urology and Wayne B. Duddleston Professor of Cancer Research at MD Anderson Cancer Center in Houston, Texas.
Related Content:
Abstract: Comparison of Costs of Radical Cystectomy vs Trimodal Therapy for Patients With Localized Muscle-Invasive Bladder Cancer.
Read: Discerning Patterns and Quality of Neoadjuvant Chemotherapy Use Among Patients with Muscle-invasive Bladder Cancer - Beyond the Abstract
Watch: Radical Cystectomy vs Trimodal Therapy for Patients With Localized Muscle-Invasive Bladder Cancer - Stephen B Williams
Download: Video Slides
Abstract: Comparison of Costs of Radical Cystectomy vs Trimodal Therapy for Patients With Localized Muscle-Invasive Bladder Cancer.
Read: Discerning Patterns and Quality of Neoadjuvant Chemotherapy Use Among Patients with Muscle-invasive Bladder Cancer - Beyond the Abstract
Watch: Radical Cystectomy vs Trimodal Therapy for Patients With Localized Muscle-Invasive Bladder Cancer - Stephen B Williams
Download: Video Slides
Read the Full Video Transcript
Ashish Kamat: So welcome. It gives me great pleasure to have Dr. Stephen Williams with us today. He is a Director of Urologic Oncology and Director of Research as well as Chief of the Division of Urology and the Medical Director at UTMB Health System here in Texas in Galveston. Welcome Stephen, and it's a pleasure to have you.
Stephen Williams: Well, thank you for having me today.
Ashish Kamat: So Steven, if you could launch into the very impressive publications that you've had recently comparing cystectomy and trimodal therapy, that would be great.
Stephen Williams: Absolutely. So as you may know, as we looked at comparative analyses of POS and survival between radical cystectomy and trimodal therapy, the background to this, as we know is, the guideline-recommended treatment for muscle-invasive bladder cancer is neoadjuvant chemotherapy followed by radical cystectomy with no dissection.
However, in more recent years, trimodal therapy has been deemed an option and even included in the guidelines as an option for treatment of muscle-invasive bladder cancer consisting of the maximal transurethral resection of a bladder tumor followed by chemo, radiotherapy. The objective that I'm going to discuss today with both of our papers is looking at and comparing these two treatments according to survival and cost.
In our first study that we published in 2018, we used propensity score matching. Primarily what we were concerned with was determining overall and cancer-specific survival. In addition, we also looked at cost up to 180 days. In our second publication, which was published in 2019, we used inverse probability treatment weighting as well as a two-part estimator to look more granular into the costs up to one year following diagnosis.
The design of the first study, as I mentioned earlier, is the propensity score matching, and as you can see here, their cohort originally then was limited to 687 patients that either went trimodal therapy and also compared these to 687 patients that underwent radical cystectomy.
The clinical characteristics are summarized here, which made the groups comparable to one another. What we observed was worse overall and cancer specific survival among patients that underwent trimodal therapy. This persisted in both sensitivity as well as propensity score analyses, which is summarized here in these Kaplan-Meier curves. What we also observed at 30, 90 and 180 days was costs were significantly increased, particularly at the 90... 180 day interval.
What was also interesting in more recent years, trimodal therapy costs at 180 days plateaued, but radical cystectomy continued to rise. The conclusion of this study that we were able to deduce was trimodal therapy was associated with decreased overall and cancer-specific survival, however, had increased cost comparable to radical cystectomy. When extrapolated nationally, this was up to $335 million in increased costs. This study was published in JAMA Surgery, which brought us to our second publication, which was also published in JAMA surgery, looking more in detail at the cost in itself.
One of the things that we found from our first study was we wanted to not only look more at the granular level of cost but also control for the impact when you don't control for survival as well as the intensity of treatments, which are important to control for as both of these can augment cost. In this study, we use inverse probability treatment weighting, and as you could see from before, this retains the overall patient in the study so that they're well matched and you don't lose data, which is a critical factor when you're performing these analyses.
In addition, as mentioned earlier, we incorporated survival because increased cost to do these treatment... to do the differences in survival between two treatments need to be taken to account as well as the intensity of treatment, which is the cost accumulation of time when patients are alive. As you can see here, after inverse probability treatment weighting, the populations were comparable between one another, and what we observed up to one year following diagnosis was significantly increased cost associated with trimodal therapy.
When we looked at the granular level of the cost, as one would expect, radical cystectomy had significantly increased hospitalization costs, however, trimodal therapy, a large amount of costs were attributed to medication expenses as well as radiology expenses. When we did look at patients that underwent neoadjuvant chemotherapy followed by radical cystectomy, there were significantly increased costs associated with radical cystectomy than compared to trimodal therapy.
When we used a two-part estimator controlling for the impact of the effects of survival and intensity of treatment, there are significantly increased costs associated with trimodal therapy. We extrapolate these findings in 2017 results in excess spending of $468 million associated with trimodal therapy. Moreover, we also found that overall cancer-specific survival was significantly decreased in trimodal therapy patients.
So in conclusion from the prior study as well as this more recent study, trimodal therapy was associated with increased and decreased survival, particularly when we also use the inverse probability treatment weighting. In addition, the differences in costs are largely attributed to medication and radiology expenses in trimodal therapy. Then when we account for the effects of survival and intensity of use, using robust statistical analyses, we found excess spending of approximately almost $140,000 per patient associated with trimodal therapy, which then extrapolating could be $468 million in increased costs in 2017.
Ashish Kamat: So this is great work, Stephen, and very provocative data. I think this is something that a lot of folks have paid attention to, both reading it and on social media. You were drawn into discussions between two camps. Could you elaborate a little bit on your sense as to where this data should take us, as in, providers and patients moving forward?
Stephen Williams: Yeah, absolutely. I think the first thing is this is a retrospective cohort study. This is not level one evidence. However, with a lack of level one evidence, this is hypothesis-generating, which means that we need to have, and hopefully, this fosters randomized controlled trials comparing these treatments. I think trimodal therapy still has a place in the management of patients with bladder cancer. However, these findings suggest that when patients are comparable between offering either treatment, radical cystectomy was associated with improved survival as well as significantly decreased costs.
Ashish Kamat: So you bring up the issue of a randomized trial. But as you know, in this patient population, the only true randomized trials were back in the early '80s, '90s, and nothing henceforth has actually been successful. So I applaud you for stating that. But practically speaking, I don't think we're going to have a randomized study comparing radical cystectomy with the radiation. With that in mind, would you then... Recognizing that this is not a level evidence, use this as a counseling tool when you're talking to patients and family?
Stephen Williams: I do. This brings me to discuss if we developed a different nomogram if you will to preoperatively also cancel patients. So regards to their comorbidities, which as you know with this population, can carry quite significant competing risks. Using this study, particularly in the context of my new role as medical director for high value care, we're not only looking at the treatments that will of course improve the survival of our patients, but also the costs that are associated with these treatments that we're all going to have to be responsible more so in the current healthcare climate.
Ashish Kamat: Yeah, and you raise an interesting point. So cost is obviously very important in bladder cancer. It is the most expensive cancer to treat in humans. So if you factor that in, what exactly do you advise? Or what would you say for the community urologist that is counseling their patient? How could they use this data to talk to their patients about outcome survival and also recognize the impact on cost on the system?
Stephen Williams: Absolutely. There was a great study that was performed recently by Kulkarni over in Toronto where they used a multidisciplinary clinic to compare and use propensity score matching. These data, particularly for those patients that are in the community or community urologists, should in fact open their eyes in regards to not only the limitation of the randomized trial that we discussed, but using deep data and robust statistical methodology to in fact counsel patients that may be selecting a bladder sparing option, which definitely sounds more appealing.
As I mentioned in this study and our limitations, this is not a cost-effectiveness study because we don't take into account quality of life years. When we're looking at cost and our cost assessment for 365 days, this is... may not account the total global cost of care. So it is an initial snapshot, and we're looking into even being more granular to understand what costs are directly associated with bladder cancer. Unfortunately using these large population-based data, we cannot do that. However, there are other multicenter databases where we hopefully can elucidate what are the actual direct costs associated with these treatments.
Ashish Kamat: Yeah, I really appreciate you cautioning our viewers and listeners that this is not level one data, but it is impressive data. Again, the granularity and the clarity that you offer is very useful. Any closing thoughts, Stephen, with regards to either your study or just the overall choice that our patients face between radiation therapy and cystectomy?
Stephen Williams: Absolutely. I think the first thing, it's a shared partnership. These data are retrospective. They don't account entirely for the inherent selection bias or determine you what treatment, the considerations of the patient, the quality of life concerns. But I think when it's determining optimal treatments for patients that are candidates for radical cystectomy, these findings with support proceeding with radical cystectomy, and would also caution patients.
I still run a multidisciplinary clinic, and quite often I have patients that are considering trimodal therapy as an option... is actually one of our co-authors on this paper, have them go see him as well to have that discussion. But even with that being said, I think these findings support proceeding with radical cystectomy and until we're able to definitively show that trimodal therapy is a superior treatment option.
Ashish Kamat: Yeah, I think you summarized that well. It's always worth remembering that it's not about just the disease, but also the patient and our patients [inaudible 00:13:04]. For some of them, clearly, bladder sparing options are worth considering. For the others, we want to do personalized medicine and radical cystectomy. Once again, Stephen, thank you very much for joining us today. Really appreciate all the work you've done in this field over the years, and excited to see what else you have to bring to our attention and publication, and I'm sure it'll be pretty soon. Thank you.
Stephen Williams: Absolutely. It was my pleasure. Thank you for inviting me.
Ashish Kamat: So welcome. It gives me great pleasure to have Dr. Stephen Williams with us today. He is a Director of Urologic Oncology and Director of Research as well as Chief of the Division of Urology and the Medical Director at UTMB Health System here in Texas in Galveston. Welcome Stephen, and it's a pleasure to have you.
Stephen Williams: Well, thank you for having me today.
Ashish Kamat: So Steven, if you could launch into the very impressive publications that you've had recently comparing cystectomy and trimodal therapy, that would be great.
Stephen Williams: Absolutely. So as you may know, as we looked at comparative analyses of POS and survival between radical cystectomy and trimodal therapy, the background to this, as we know is, the guideline-recommended treatment for muscle-invasive bladder cancer is neoadjuvant chemotherapy followed by radical cystectomy with no dissection.
However, in more recent years, trimodal therapy has been deemed an option and even included in the guidelines as an option for treatment of muscle-invasive bladder cancer consisting of the maximal transurethral resection of a bladder tumor followed by chemo, radiotherapy. The objective that I'm going to discuss today with both of our papers is looking at and comparing these two treatments according to survival and cost.
In our first study that we published in 2018, we used propensity score matching. Primarily what we were concerned with was determining overall and cancer-specific survival. In addition, we also looked at cost up to 180 days. In our second publication, which was published in 2019, we used inverse probability treatment weighting as well as a two-part estimator to look more granular into the costs up to one year following diagnosis.
The design of the first study, as I mentioned earlier, is the propensity score matching, and as you can see here, their cohort originally then was limited to 687 patients that either went trimodal therapy and also compared these to 687 patients that underwent radical cystectomy.
The clinical characteristics are summarized here, which made the groups comparable to one another. What we observed was worse overall and cancer specific survival among patients that underwent trimodal therapy. This persisted in both sensitivity as well as propensity score analyses, which is summarized here in these Kaplan-Meier curves. What we also observed at 30, 90 and 180 days was costs were significantly increased, particularly at the 90... 180 day interval.
What was also interesting in more recent years, trimodal therapy costs at 180 days plateaued, but radical cystectomy continued to rise. The conclusion of this study that we were able to deduce was trimodal therapy was associated with decreased overall and cancer-specific survival, however, had increased cost comparable to radical cystectomy. When extrapolated nationally, this was up to $335 million in increased costs. This study was published in JAMA Surgery, which brought us to our second publication, which was also published in JAMA surgery, looking more in detail at the cost in itself.
One of the things that we found from our first study was we wanted to not only look more at the granular level of cost but also control for the impact when you don't control for survival as well as the intensity of treatments, which are important to control for as both of these can augment cost. In this study, we use inverse probability treatment weighting, and as you could see from before, this retains the overall patient in the study so that they're well matched and you don't lose data, which is a critical factor when you're performing these analyses.
In addition, as mentioned earlier, we incorporated survival because increased cost to do these treatment... to do the differences in survival between two treatments need to be taken to account as well as the intensity of treatment, which is the cost accumulation of time when patients are alive. As you can see here, after inverse probability treatment weighting, the populations were comparable between one another, and what we observed up to one year following diagnosis was significantly increased cost associated with trimodal therapy.
When we looked at the granular level of the cost, as one would expect, radical cystectomy had significantly increased hospitalization costs, however, trimodal therapy, a large amount of costs were attributed to medication expenses as well as radiology expenses. When we did look at patients that underwent neoadjuvant chemotherapy followed by radical cystectomy, there were significantly increased costs associated with radical cystectomy than compared to trimodal therapy.
When we used a two-part estimator controlling for the impact of the effects of survival and intensity of treatment, there are significantly increased costs associated with trimodal therapy. We extrapolate these findings in 2017 results in excess spending of $468 million associated with trimodal therapy. Moreover, we also found that overall cancer-specific survival was significantly decreased in trimodal therapy patients.
So in conclusion from the prior study as well as this more recent study, trimodal therapy was associated with increased and decreased survival, particularly when we also use the inverse probability treatment weighting. In addition, the differences in costs are largely attributed to medication and radiology expenses in trimodal therapy. Then when we account for the effects of survival and intensity of use, using robust statistical analyses, we found excess spending of approximately almost $140,000 per patient associated with trimodal therapy, which then extrapolating could be $468 million in increased costs in 2017.
Ashish Kamat: So this is great work, Stephen, and very provocative data. I think this is something that a lot of folks have paid attention to, both reading it and on social media. You were drawn into discussions between two camps. Could you elaborate a little bit on your sense as to where this data should take us, as in, providers and patients moving forward?
Stephen Williams: Yeah, absolutely. I think the first thing is this is a retrospective cohort study. This is not level one evidence. However, with a lack of level one evidence, this is hypothesis-generating, which means that we need to have, and hopefully, this fosters randomized controlled trials comparing these treatments. I think trimodal therapy still has a place in the management of patients with bladder cancer. However, these findings suggest that when patients are comparable between offering either treatment, radical cystectomy was associated with improved survival as well as significantly decreased costs.
Ashish Kamat: So you bring up the issue of a randomized trial. But as you know, in this patient population, the only true randomized trials were back in the early '80s, '90s, and nothing henceforth has actually been successful. So I applaud you for stating that. But practically speaking, I don't think we're going to have a randomized study comparing radical cystectomy with the radiation. With that in mind, would you then... Recognizing that this is not a level evidence, use this as a counseling tool when you're talking to patients and family?
Stephen Williams: I do. This brings me to discuss if we developed a different nomogram if you will to preoperatively also cancel patients. So regards to their comorbidities, which as you know with this population, can carry quite significant competing risks. Using this study, particularly in the context of my new role as medical director for high value care, we're not only looking at the treatments that will of course improve the survival of our patients, but also the costs that are associated with these treatments that we're all going to have to be responsible more so in the current healthcare climate.
Ashish Kamat: Yeah, and you raise an interesting point. So cost is obviously very important in bladder cancer. It is the most expensive cancer to treat in humans. So if you factor that in, what exactly do you advise? Or what would you say for the community urologist that is counseling their patient? How could they use this data to talk to their patients about outcome survival and also recognize the impact on cost on the system?
Stephen Williams: Absolutely. There was a great study that was performed recently by Kulkarni over in Toronto where they used a multidisciplinary clinic to compare and use propensity score matching. These data, particularly for those patients that are in the community or community urologists, should in fact open their eyes in regards to not only the limitation of the randomized trial that we discussed, but using deep data and robust statistical methodology to in fact counsel patients that may be selecting a bladder sparing option, which definitely sounds more appealing.
As I mentioned in this study and our limitations, this is not a cost-effectiveness study because we don't take into account quality of life years. When we're looking at cost and our cost assessment for 365 days, this is... may not account the total global cost of care. So it is an initial snapshot, and we're looking into even being more granular to understand what costs are directly associated with bladder cancer. Unfortunately using these large population-based data, we cannot do that. However, there are other multicenter databases where we hopefully can elucidate what are the actual direct costs associated with these treatments.
Ashish Kamat: Yeah, I really appreciate you cautioning our viewers and listeners that this is not level one data, but it is impressive data. Again, the granularity and the clarity that you offer is very useful. Any closing thoughts, Stephen, with regards to either your study or just the overall choice that our patients face between radiation therapy and cystectomy?
Stephen Williams: Absolutely. I think the first thing, it's a shared partnership. These data are retrospective. They don't account entirely for the inherent selection bias or determine you what treatment, the considerations of the patient, the quality of life concerns. But I think when it's determining optimal treatments for patients that are candidates for radical cystectomy, these findings with support proceeding with radical cystectomy, and would also caution patients.
I still run a multidisciplinary clinic, and quite often I have patients that are considering trimodal therapy as an option... is actually one of our co-authors on this paper, have them go see him as well to have that discussion. But even with that being said, I think these findings support proceeding with radical cystectomy and until we're able to definitively show that trimodal therapy is a superior treatment option.
Ashish Kamat: Yeah, I think you summarized that well. It's always worth remembering that it's not about just the disease, but also the patient and our patients [inaudible 00:13:04]. For some of them, clearly, bladder sparing options are worth considering. For the others, we want to do personalized medicine and radical cystectomy. Once again, Stephen, thank you very much for joining us today. Really appreciate all the work you've done in this field over the years, and excited to see what else you have to bring to our attention and publication, and I'm sure it'll be pretty soon. Thank you.
Stephen Williams: Absolutely. It was my pleasure. Thank you for inviting me.