Dr. Meredith Shiels — Health in Numbers

>> Diego (narration): Almost 50 years ago, in the late 70s, NIH scientists studying cancer rates in the US discovered that some cancers varied significantly across the population. They plotted cancer death rates by county and found that there were cancer mortality hotspots. Prostate cancer, for instance, was higher among nonwhites, kidney cancer was more prevalent in people of German, Scandinavian, and Russian descent, and bladder cancer was higher in New England.

These initial findings spurred further investigations aiming to understand what could account for the differences in cancer prevalence. In the case of higher bladder cancer incidence in New England, a separate team of NIH scientists showed years later, that part of explanation was arsenic in the drinking water.

[SOUND OF WATER FAUCET OPENING AND CLOSING]

This is just one example that illustrate the clear impact of epidemiology on public health. In simple terms, epidemiology is the study of how diseases or health-related events spread and affect people in particular populations. Epidemiologists are like watchful guardians of health; they collect and analyze data to track the status quo. And when there are deviations, they crunch the numbers to understand who is getting sick, where, how, and why. 

>> Dr. Shiels: Descriptive epidemiology mostly focuses on surveillance. And by surveillance, I mean monitoring populations generally through data collected by public health departments to understand how patterns of disease and death have either changed over time or differ across different groups of people.

>> Diego (narration): That’s Dr. Meredith Shiels, an epidemiologist and senior investigator at the National Cancer Institute, or NCI—not coincidentally the same institute as the NIH scientists who mapped out cancer mortalities in the 70s. Following in their footsteps, Dr. Shiels keeps her eye on the cancer landscape to discern what populations might be at higher risk, figure out ways to mitigate those risks, and evaluate whether those measures are working.

 >> Dr. Shiels: I think of descriptive epidemiology or surveillance as like the beginning and the end of everything. So, it allows us to identify populations at risk or, you know, alarming patterns. Like if some cancer starts increasing, something is not -- something's going wrong, right? That's the beginning. And then at the end, the goal is to reduce disease and death. So, we need to know that whatever public health efforts we're putting into place, we need to check to see that at a population level, we are making an impact.

>> Diego (narration): Surveillance research might not sound like the flashiest type of research, but it’s importance can’t be understated. Without the work of epidemiologists like Dr. Shiels, major health threats would remain hidden in the data. 

In this episode, I talk to Dr. Shiels about the stories the stats can tell. Stories that sometimes lend credence to what we might already believe to be true, and other times, lead to surprising, even counter-intuitive, discoveries.

[MUSICAL TRANSITION]

>> Diego (interview): Since descriptive epidemiology is all about discerning trends in the numbers or patterns from the data, I thought it might be fitting to start by discussing a series of numbers that I’m sure you’ll be familiar with, and that’s 50, 25, 2.3 and 2.7.

These numbers describe the goal and the status of President Biden's Cancer Moonshot. His administration is aiming to reduce cancer rates by at least 50% over the next 25 years which would require a 2.7 rate of decline, but currently it's only 2.3.

>> Dr. Shiels: Yup.

>> Diego (interview): And that's something you found. You led that study that projected that rate. And in doing so, you identified gaps that could be narrowed to improve those margins and address cancer health inequities. So let’s talk about those.

>> Dr. Shiels: Sure. So, the project that you're referring to was really to understand this moonshot goal of a 50% decline in age-adjusted mortality rates over 25 years. There's been a lot of progress in cancer mortality over time. And so what we did was basically say, if we progress at the same rate, how close or far are we from that goal? And so, what we estimated was, as you said, at a 2.3% decline which is where we are now, per year, that will get us to 44% instead of to 50%. If we get to 44%, that's actually like a tremendous success in and of itself. That's a lot of lives saved.

But in order to reach the 50%, there needs to be an acceleration of these declines. And so what we did was we looked at the leading six causes of cancer death which together cause 57% of all cancer deaths. So that's lung, pancreas, prostate, breast, colorectal and liver. And we identified opportunities to accelerate declines in these cancers. And for some cancers, it's more promising than others. So lung cancer, for example, causes about between 20 and 25% of all cancer deaths. But there's been tremendous success against lung cancer over time. A lot of that is because the prevalence of smoking has declined tremendously. Lung cancer lags behind by about 20 to 30 years when you compare it to smoking prevalence. Because it takes a long time for lung cancer to form, the success that we're seeing now in reductions in lung cancer due to smoking are actually from the progress against smoking that happened--

>> Diego(interview): --In the 90s and early 2000s with the no smoking campaigns.

>> Dr. Shiels: Yeah. So, we think there's a lot of progress against lung cancer already baked in, so to speak. And there's also been developments with lung cancer treatment that have been very successful for specific types of lung cancer that have also contributed to declines in mortality. There's also lung cancer screening which became available maybe 10 years ago. And so each of those three things, so that’s prevention, early detection which is screening, and then treatment are modalities available to reduce mortality, but there are disparities in all of them. So, accelerating progress against cancer mortality and addressing health inequities have to go hand in hand. Because if we're saying like we need to reduce smoking further, well, who still smokes? The biggest gradient of smoking prevalence is across education. There's a huge difference between the prevalence of smoking of somebody with a graduate degree and somebody with a high school degree. So that is a huge disparity there. And then the same with screening and treatment. So, I think the goal of reaching a 50% decline can only be achieved by expanding access to the things that already exist and work.

>> Diego (interview): Right. Well let’s get into those disparities a little bit more. Because a line of your research looks at how emerging health problems can have different effects on different socioeconomic and ethnic groups. And one example that stuck out to me while I was reading up on work was how you made a concerted effort to distinguish Native Hawaiian and Pacific Islander from a more generalized Asian American grouping while studying excess death rates due to cancer. When you separated them out, you found a significant difference. Can you describe what that was and what was the motivation to delineate between the groups in the first place?

>> Dr. Shiels: Well, within our broad categories of racial groups, there's obviously a lot of differences, right? And our analyses are limited by the data that are available essentially. And a lot of that work that you're referring to uses death certificate data and those groups were grouped on death certificates for a number of years. In the early 2000s, the standard death certificate changed in the US. But because we are a nation of states, it took the states from 2003 to 2017 for every state to adopt this new death certificate. So, it took until the whole country was on board before we could separate these groups. In the newer death certificate, I say newer but it was introduced 20 years ago, you can separate these groups. And there are profound differences between Asian Americans and Native Hawaiian and Pacific Islanders in terms of disease rates, and in this case, mortality rates.

What was before the Asian and Native Hawaiian Pacific Islander group grouped together, is often seen as like the healthiest group. But Native Hawaiians and Pacific Islanders only make up about 3% of that aggregated group. So when you look at the group together, the patterns really only apply to this broad Asian American catergory. And so that population of Native Hawaiian and Pacific Islanders was completely hidden in prior statistics.

>> Diego (interview): Ah ha.

>> Dr. Shiels: So now that we're able to separate it, it's really important to look at these things differently.

And so, we have the one paper on cancer mortality rates that have separated Asian Americans and Native Hawaiian and Pacific Islanders. And then there's several fellows here as well as outside collaborators that are really digging into, you know, disaggregating these two groups to really understand where the health disparities are or else they can't be addressed, right?

And in that study of cancer mortality, we actually found that in the youngest age group that we look at, young adults, I can't remember the exact age range, but in young adults, Native Hawaiian and Pacific Islanders actually had the highest cancer mortality rates of any of the racial ethnic groups examined in that study. So that was completely hidden before.

>> Diego (interview): But it’s not just racial ethnic groups that experience disparities, right? You’ve also seen inequities when comparing groups of different incomes and even ages.

>> Dr. Shiels: Yeah, we’ve done some work on county-level measures of socioeconomic status. So what, for example, we did a study looking at median county-level income and how it related to cancer mortality over time. And what we found was again like similar to racial ethnic disparities, there has been progress across both high-income counties and low-income counties. The cancer mortality rates have declined. But they're declining faster in higher income counties. So if you think about what that does to a disparity, it makes it bigger over time, right? So, higher income counties already have lower cancer rates and they're improving at a faster clip. So, it's increasing the cancer health disparities over time.

We also looked in this study at 25 to 64-year-olds versus those aged 65 and older. And what we saw was that the disparity by county-level income is much greater in 25 to 64-year-olds than it is in those aged 65 and older. We can't actually directly say what the difference is due to but it could be due to the fact that in this country, everybody has health care once they reach Medicare age. So at Medicare age, there is still a disparity but it's much smaller than what we see for 25 to 64-year-olds.

>> Diego (interview): Right, right, right.

>> Dr. Shiels: It's sobering to see those statistics. And we actually estimated that among 25 to 64-year-olds, these are data from 2015, if everybody had the cancer mortality rate of the highest income counties, so the highest I believe it was quartile of counties, there would be 21% fewer cancer deaths in that age group. So that tells you how big of a difference it makes to live in a higher income county.

>> Diego (interview): Wow, yeah, It is sobering when you put in those terms. But you know, it might not be too surprising to hear that, you know, that higher income leads to lower cancer incidence. Still, I do think it is very powerful to have the data to back up what might be intuitive. I mean it makes sense to think that higher income might lead to higher access to health care and therefore, lower rates of cancer. But having the numbers really makes it a reality, you know.

>> Dr. Shiels: Yeah, as you said, it wasn't surprising to us to see this finding. But I think a lot -- with a lot of descriptive studies, having the numbers to show what we already maybe know to be true is very powerful. So I think of it in two ways. You know, existing disparities that we know about, we need to continue monitoring them and we need to provide the numbers to show how big the disparities are and where they exist. And then as we go along there will be new data hopefully to highlight and identify these disparities that we haven't been able to quantify before. For example, separating out racial groups into more refined categories.

>> Diego (interview): Right. All this just makes me think that epidemiology really is a cornerstone of public health. I mean, the numbers can actually shape policy and serve as a foundation for so much, like evidence-based practices in the field.

>> Dr. Shiels: Yes.

>> Diego (interview): And really, without the type of work that you do, we wouldn't be able to identify gaps in our healthcare system, to figure out the factors that put certain groups at risk, and come up with targets for preventative health care. There’s so much that comes from the types of studies you do.

>> Dr. Shiels: Yeah. I think of descriptive epidemiology or surveillance as like the beginning and the end of everything. So, it allows us to identify populations at risk or, you know, alarming patterns. Like if some cancer starts increasing, something is not -- something's going wrong, right? That's the beginning. And then at the end, the goal is to reduce disease and death. So, we need to know that whatever public health efforts we're putting into place, we need to actually see that those interventions—that public health impact needs to be also seen in the surveillance. So I see it as the foundation for, you know, where to start with a public health question. And then, also we need to check to see that at a population level, we are making an impact.

>> Diego (interview): Well I think that segues nicely into talking about a prime example of how you’ve been keeping an eye on impacts to public health. And that’s with HIV.

In one of your studies you found that the proportion of HIV-infected adults in the US 65 years and older was actually projected to increase from around 8 percent in 2010 to 21 percent in 2030. And I think at face value, that might seem a little bit confusing or maybe a little disheartening to think that cancer statistics are increasing. But that's not really what's really happening here, right? The bump actually represents something good?

>> Dr. Shiels: So, in 1996 in the United States, more effective treatment for HIV was introduced. And so, HIV, well AIDS, went from something that was a fatal illness to something that now is, you know, managed with medication. And the upside of that is that people are living much longer with HIV. Somebody who's on treatment and controls the virus would be expected to live into their 70s, for example. So, a similar, you know, projected life expectancy. So for the first time, you know, over the last decade or so, people with HIV who were maybe infected in their 30s and 40s are now living to older ages. And what that means is that those individuals are at risk for aging-related illnesses. So, whether it has something to do with their HIV or not, people with HIV are now, you know, reaching ages where colon cancer, for example, risk increases. Now, colon cancer does not occur at a higher rate in people with HIV. But because people with HIV are now living to older ages, there's going to be more colon cancer among people with HIV just because of the demographics of the population.

>> Diego (interview): -- and again not because HIV increases a person’s chances of getting colon cancer.

>> Dr. Shiels: Yeah. So, in one of our papers, we projected out what we expected the most common cancers to be and I mean it's from a few years ago so we projected in 2020 and 2030. And so in 2030, prostate cancer was the cancer that we estimated would be the most common cancer in people with HIV. Prostate cancer is not associated with HIV. In fact, it occurs less frequently in men with HIV than in the general population for unknown reasons. But because the HIV population in the US is disproportionately male and because those men are aging and prostate cancer is so common, you know, in terms of cancers at older ages, that that's the cancer we estimated would be the most frequent. So, for cancer in people with HIV, we have to think, you know, in terms of prevention both about those that occur at a higher rate and then also like what are the, you know, age-specific protocols that should be applied to people with HIV now that they're reaching older ages. If you think about cancer screening for example, people with HIV are living to age ranges where they should be receiving mammography or colonoscopy the same way that somebody without HIV would. And that's also an important part of cancer prevention in addition to, you know, focusing on those cancers that occur at a much higher rate.

>> Diego (interview): You mentioned that some cancer do occur at a higher rate in people with HIV. What are those cancers? And do we know why people living with HIV have an elevated risk of developing them?

>> Dr. Shiels: Sure, yeah. So, many of the cancers that occur at an elevated rate among people with HIV are cancers that are caused by other infections. So, whether it'd be HPV, human papillomavirus, or hepatitis C virus, or Kaposi's sarcoma herpes virus which causes Kaposi's sarcoma, a lot of those cancers occur at an elevated rate in people with HIV. And that's because people with HIV have suppressed immune systems. So the body's not able to fight the infection the way that it should. Some of these viruses also just the prevalence is higher in people with HIV because of shared routes of transmission. So if you think about something like hepatitis C virus, one of the main routes of transmission is injection drug use and that's of course also a risk factor for HIV. Or HPV, anal HPV, infection is higher particularly in men who have sex with men with HIV. And again, that's a shared route of transmission, through sexual contact. So, it's a combination of a higher prevalence of some viruses, not all. On top of that, suppressed immunity that prevents the body from fully controlling those viruses. And then there are some other factors, like there are other risk factors for cancer that are higher in people with HIV such as a higher smoking prevalence.

>> Diego (interview): Oh really? Why is that?

>> Dr. Shiels: I haven't done research into why smoking rates are higher so I can't really speak to that angle. But we did do some analyses where we tried to understand if that could be the sole explanation for the elevated risk. And we think at younger ages, the elevated rates of lung cancer in people with HIV, it's not just smoking. We think that could be something related to immunity maybe or people with HIV get more frequent like lung infections, for example, so that could be part of the explanation at younger ages. At the oldest ages, we don't really see much of an elevated risk at all in lung cancer. The complementary piece of that is that, this is just what we were talking about, as people age, if they are smokers in particular, the risk of lung cancer increases. So while at the older ages, somebody with HIV may not have a higher risk of lung cancer than somebody without HIV on average, the absolute risk of developing lung cancer increases with age.

Cameron Haas who's a postdoctoral fellow here did an analysis and showed that in people in their 60s, the risk of lung cancer is actually much higher than Kaposi's sarcoma and non-Hodgkin lymphoma which are like the hallmark HIV-related cancers. So it's important to keep in mind that as the population of people with HIV is aging and, you know, over time, then the cancers that we will see in this population will sort of shift with that aging. And that's why a lot of kind of my work is focused on really like updating the snapshot of which cancers are occurring in people with HIV and why. Is it because it's an elevated risk or is it because the population has now shifted towards older ages.

>> Diego (interview): So with that being the bulk of your work, what made you choose the intersection between HIV and cancer? This is probably a little ignorant, but why not like diabetes and cancer or something else.

>> Dr. Shiels: So I started on my work in HIV and cancer as part of my dissertation when I was getting my PhD at Johns Hopkins. And I mean, I think that the topic was offered to me basically as often happens with a PhD. But my projects in my dissertation were using cohorts of people with HIV, so not these huge databases but like carefully followed cohorts of people with HIV over time.

And while those studies have a richness of data that's been collected and they're really amazing, they don't have the numbers to look at all these perhaps less common cancers. So, I came to NCI to do a post-doc with Dr. Eric Engels who at the time was the only PI of the HIV/AIDS Cancer Match Study. So I came to do my post-doc with him and then I started working with these large datasets and I still -- now, we're co-principal investigators of that study. So, I mean he's right next door. We still work quite a lot together. But yeah, so I transitioned into the large database studies through my work on HIV. And so when I came to NCI, so much of my work was using these big surveillance systems and asking population-level questions that whenever I transitioned to being a PI, I still have much of my work focused on HIV but I took that sort of approach, the descriptive approach to expand to other areas of research.

>> Diego (interview): Great. Well I do want to highlight your path in the IRP. So you came to NIH as a postdoctoral fellow in 2009. And then you were promoted to research fellow in 2011 then became a tenure track investigator in 2016 and are now a senior investigator as of 2021.

I would venture to say that makes you a perfect example, and not necessarily rare, of the longevity you can have here. You’ve built a thriving, lifelong career at NIH. So, what's motivated you to stay for so long?

>> Dr. Shiels: So, I love the Intramural Program at NIH. And the type of research that I do is perfect for the Intramural Research Program. Because with surveillance research, if you're an investigator at a university, you have to get grants to both cover your salary and to fund your work. And the type of research that I do isn't necessarily something that is easily fundable through a grant mechanism. It also is generally tracking the health of the United States. So it's something that should be done by the federal government. And so in terms of research, it fits really well here. And I've always loved that aspect of it, being able to do surveillance work, descriptive work within the federal government. I've always just felt very supported here at NIH and felt like there are so many opportunities, you know, to expand this area of work.

>> Diego (interview): Are there any words of advice you would give to someone that's like, you know, just starting out in your field at NIH?

>> Dr. Shiels: I think I would say that it's really important to figure out the type of research that you love. And it's equally important to figure out the type of research that you don't love because I've done other things along the way and I deem the experiences that I didn't love to be just as important as the ones that I did love. Because if you’re going to build a career, you have to really like what you're doing. This is very challenging at times. It can be stressful at times. It's possible you will be told, you know, there's not a place for this here but that can also change over time. So like when I was a fellow early on, there was no PI whose mandate was to do surveillance work. And in fact, I was told that wasn't a thing that was going to be an investigator position. But over time, people evolve, priorities evolve and that position kind of opened up at exactly the right time for me. But I already knew that this is really what I loved to do. So I really think just use the experience as a fellow at the NIH to really try different things, get experience in different things, but figure out really what motivates you. Because in order to be successful, you're going to have to, you know, really feel a passion about what you're doing and also want to spend the time doing that.

>> Diego (interview): Yeah. And I will say just from this conversation, it's very apparent that you love what you do and I think it made for a very enlightening conversation. So thank you so much!

>> Dr. Shiels: Yeah, it was so nice to meet you! I’ve enjoyed our talk today.