Zachary Klaassen: My name is Zach Klaassen, urologic oncologist at the Georgia Cancer Center in Augusta, Georgia. We are at the PCF Scientific Retreat in Carlsbad, California. I'm delighted to be joined by one of the young investigators for PCF, Dr. Chris Dee, who is a radiation oncology resident at Memorial Sloan Kettering Cancer Center. Chris, thanks for joining us on UroToday.

Edward Christopher Dee: Thank you for having me.

Zachary Klaassen: So you have some great work for your YIA looking at prostate cancer global disease burden and looking at some of the health system levers we can use to improve that. So before we get into the work that you've done, just tell us the problem. What's the prostate cancer global burden?

Edward Christopher Dee: Absolutely. I think one of the main things to know is prostate cancer is one of the most common cancers in the world, as we know, and the past several truly decades of advances have demonstrated how well we can control prostate cancer. As a radiation oncologist, we see patients, some of whom have pretty aggressive disease that we can literally cure, and there's so much excitement about what we're able to do both in the localized and even in the metastatic setting. But what we also I think have realized is yes, there's disparities and differences in terms of the burden of disease by ways of epidemiology and incidence, but what we also know is that prostate cancer is one of the most disparate cancers in terms of outcome. Overall survival, yes, there's some kind of differences in the quality of the data can range from upwards of 90 plus 95% in North America and Australia and Canada. But in contrast in other parts of the world where there's not as much access to whether it's early detection or treatment, overall survival could be in the sixties and seventies. There's a recent Lancet Commission that talked about that and really ...

Zachary Klaassen: It was eye opening, wasn't it?

Edward Christopher Dee: it was.

Zachary Klaassen: I saw that as well.

Edward Christopher Dee: Excellent piece, super comprehensive, but really I think highlights how our advances, although impressive, commendable, and truly life-altering for patients, don't reach patients equitably. And that's something that we wanted to ask more about in this arm of research.

Zachary Klaassen: And just tell our listeners what a health system lever is. I know there's a lot of examples of that.

Edward Christopher Dee: Absolutely.

Zachary Klaassen: And so I think just to level set before we get into your work, just tell us what that means.

Edward Christopher Dee: The way I think of this is we can think of health systems as just a block of moving parts and they're extremely complex. But when we were conceptualizing this paper and several other related studies, we said, okay, what levers, how can we break the idea of a health system down into things that are potentially learnable, things that we can figure out and into it perhaps without any truly causal data, ways in which those could potentially be leveraged to improve outcomes. I am trying to be careful with words because with retrospective data, we can't assume causality, but it's hard to really, you can't really run a trial to prove a causal association with the health system. So we said, okay, what are things that could play a role, whether it's GDP per capita, whether it's healthcare workforce, those are all factors that I think of as levers, things that could be potentially leveraged to potentially change outcome.

Zachary Klaassen: Great. So for your work, you gave a talk at the retreat on the update of your work. Just tell us the highlights of that talk and the work you've done.

Edward Christopher Dee: Absolutely. So the talk was twofold. One body of work that the group has done, although it's not the main focus of our conversation today is really descriptive epidemiology. One of my personal areas of interest is cancer in Southeast Asia, that's where I'm originally from. And we provide the kind of descriptive data looking at the burden of all cancers in that region. It's 700 million people, 11 countries, and I think an area that is often overlooked. And so, one of our earlier papers with the group was really just a descriptive epi piece looking at the burden of cancer in that part of the world. And we dug deeper. One of the, actually now physicians in the group, I was going to say medical student, he just graduated, Dr. Columbus, is a Filipino-American guy who led our study looking specifically at genitourinary cancers within Southeast Asia demonstrating what we know was going to be an increase, but really quantifying that and showing how in particular prostate cancer plays a pretty significant role in the region. So those are a pair of papers that helped us to quantify the problem more broadly.

But the crux of the study that I presented yesterday is what we're talking about today, which is the question of really which health system levers could potentially be associated with better or worse and in what direction prostate cancer mortality to incidence ratio. Now one caveat and just two caveats. Number one is we specifically looked as macroscopically as we could. We said we're going to use each country's health system as a data point. And in so doing one limitation and caveat in the work is we overlook the within country disparities, but we said if we zoom out, which few have done before, can we learn something that's unique about prostate cancer? The second piece is the outcome. So we chose mortality to incidence ratio, which is a proxy for overall survival. At the upper ends of mortality to incidence ratio, so countries where approaches 90, a hundred or 0.9 to one, it probably is less good at resolving differences amongst countries. But when you look through the breadth of outcomes that we see across the world, it becomes a more important metric. We know that there are countries where overall survival can be in the sixties and seventies range. So we said, okay, we'll get something that's a bit more of a sledgehammer and allows us to get that degree of resolution.

And so step number one was to collect all the data. We basically mined WHO, UNDP for factors that could be related, surgical workforce per unit population, radiotherapy centers per unit population, and so on and so forth, and try to get those data for as many countries as we could. Our base data for the outcomes was based on GLOBOCAN, which is one of the main aggregate data sets for cancer epidemiology. And that data set has estimates for 185 countries. So we said of that set of 185 countries, which of these have sufficient data to make these claims? So collected all that data for 10, 11 metrics, and then we ran univariable linear models first to see what associations we could find, and essentially everything we could find, GDP per capita, health spending as a percentage of GDP. All of that was significantly associated really in the expected direction with mortality to incidence ratio. And so we said, okay, we'll use forward selection and then take everything that was significant on univariable analysis and put that into a multivariable model to see, to really ask the question of which of these factors that we know are interrelated. We know these are intimately and very in a complex manner intertwined and tried to ask the question of which of these would still be independently associated one from the other. There's a bit of correction for multicollinearity, and so Human Development Index or HDI was removed, that moves in a collinear manner with a lot of the other variables.

And what we found was specific to prostate cancer, universal health coverage service index, GDP per capita, surgical workforce per unit population, and radiotherapy centers per unit population, were all independently associated with mortality incidence ratio, which we thought was a really cool finding. I think this speaks to number one, the complexity of health systems and how at a global level there really are several components to that story. Number two is I think it speaks to the availability of different modalities of treatment being important for patients actually mitigating that mortality burden. But I think the biggest piece of the story is a universal health coverage service index association. That's a UNDP-derived metric that really asks the question of to what degree are health services available with minimal risk of financial catastrophe for patients in each country? And that was a rock-solid association. And what that tells us is that there's something to be said for mitigating the risk of financial toxicity, mitigating the risk of cost of care actually impacting patient's ability to access that care. And that association tells us that that's something that potentially could inform policymakers as they recognize the burden of prostate cancer and then work to find ways to alleviate that burden.

Zachary Klaassen: That's great. No, phenomenal. That area of the world, that approach is just so fascinating. I have two follow up questions just listening to you explain that. We're blessed with a SEER database here, 1973. We've had phenomenal data. How does the granularity, obviously not as granular as SEER, how does it compare to some of these countries? Follow up question is it possible to look at PSA screening rates in some of these countries?

Edward Christopher Dee: Very, very good questions. Number one is in the US I think we have amongst the best population-level data sets out there. I think a couple European countries have pretty good data as well. The UK does, for example, but SEER and to a lesser extent, NCDB really are incredible in terms of the granularity of available data. That is not available really anywhere else to my knowledge, other than, again, a few caveat countries. The data sets that we use from GLOBOCAN are inherently limited, and we do our best to acknowledge that in the paper's text. But going back to the Southeast Asia descriptive epi paper, for example, five or six of the 11 countries in Southeast Asia don't actually have registries that work and are qualified and were therefore estimates based on neighboring countries that share similar demographics and also health systems. And so it's an inherently limited data set, which is why we felt comfortable using relatively more sledgehammer type metrics, including mortality to incidence ratio. Those are variables that move with no subtlety. And so yeah, I think it would be a dream to have something like SEER throughout the world, and I think that that's forthcoming in a few studies. One shout out to a group I work with in Hong Kong that uses essentially similar data to what the NHS has to study access to medications, but also even things such as metformin use and the risk of prostate cancer. Those are studies that the group has published and been part of before, but there's small pockets of places where there's good data, but at the population level, there's nothing that comes close. The second question?

Zachary Klaassen: PSA screening.

Edward Christopher Dee: That is something that we struggled with in the paper and did not ultimately include in the paper itself. The reason I say that is it is not as binarily defined, I think because even if a country were to have PSA screening available that the uptake would be unclear and the follow-through would be unclear. I suspect, yes-no recommendation would probably be associated, but would probably be a marker more broadly of access to healthcare and ...

Zachary Klaassen: Be collinear with a lot of the variables.

Edward Christopher Dee: Be collinear with a lot of the other variables. Yeah. There's an interesting paper from a breast group. Duggan is the first author in Lancet Oncology that actually inspired this where they asked a similar question, but specific to breast cancer. And then their outcomes were, I think they looked at mortality in this ratio, I could be wrong, but they also looked at stage at presentation and they found that UHC index was associated number one, but also that the presence of a screening program was also associated. So I think there could be a hypothesis to chase there as well.

Zachary Klaassen: Sure. Just great work, I think.

Edward Christopher Dee: Thank you.

Zachary Klaassen: Ultimately, the next, let's say two to five years, how do we start to implement these changes? Obviously this is highlighting the situation of the problem.

Edward Christopher Dee: The next 20 to 50 years.

Zachary Klaassen: Next 20 to 50, let's put a zero behind those next 20 to 50 years. How do we implement some of this great work?

Edward Christopher Dee: Yeah, I think about this a lot. I think number one is the work that UroToday does. The work that we do here at PCF, actually getting the message out is really, really important. We had a beautiful talk by Dr. Abate-Shen yesterday talking about the importance of science communication with non-science people. I think that is often overlooked. There's so much out there that, there's so much within the work that we do as researchers, whether basic scientists or population researchers, that doesn't leave the confines of our academia. Oftentimes, we're happy that we're able to say something at ASCO or ASTRO where you're talking to 50,000 other people who care about science, but the world is much bigger than that. And I think a lot of the onus falls on us and the partners that we have in different industries to be able to advocate further and to be able to, and this is far easier said than done, to be able to advocate for our patients and for improving health systems outside the confines of academia. And I think the act of writing papers and the act of putting together manuscripts, I think that gives us some degree of credibility. But the crux of it lies in the actual implementation in going out there and actually sharing this information. So this work that you guys do is really important. But I think a challenging thing that really needs to be worked on in the future is we have all these amazing papers, like that Lancet Commission, for example, but that has to reach policymakers. We need to get people who think similarly into the positions of power that are able to make those changes. Our finding of the association with UHC index and cancer outcomes, that's a systemic thing. That's something that is so high up and outside of the confines of the radiation oncology.

Zachary Klaassen: It's not just prostate cancer.

Edward Christopher Dee: It's not just prostate cancer. But I think that this is the first drop in a bucket really, and I think the onus falls on us to advocate more. We actually recapitulated these results for a pan-cancer analysis, essentially finding the same thing, at least for UHC index and GDP per capita in a pan-cancer study. Similar methods, but looking at all cancers combined, and it tells the same story essentially. The UHC index is at the heart of a lot of this, and it tells us the index is just a number, but it's reflective of underlying systems of essentially healthcare financing that need to be addressed further.

Zachary Klaassen: That's great. You and I could talk about this for the next three hours. It's been a great conversation. Thank you so much for joining us.

Edward Christopher Dee: Of course. Of course.

Zachary Klaassen: Congratulations on the award and just being a leader in this field, so congratulations.

Edward Christopher Dee: Thank you for having me. One quick word, I just want to absolutely do a shout-out for my mentors.

Zachary Klaassen: Absolutely.

Edward Christopher Dee: This work is not possible without the folks who helped put the award together. Dr. Puneeth Iyengar, Dr. Sean McBride and Dr. Dana Rathkopf, but also my mentors of many, many years Dr. Paul Nguyen, Dr. Brandon Mahal and Dr. Vinayak Muralidhar. These are folks who have and continue to be instrumental in all these ideas. A lot of them are co-authors of the paper, and I'm grateful for them.

Zachary Klaassen: well said. Great way to finish off the conversation. Thanks, Chris.

Edward Christopher Dee: Thank you. Thank you.