Protecting Participant Privacy and Making Predictions Using Alzheimer’s Data

Transcript

Intro: I’m Dr. Nathaniel Chin, and you’re listening to Dementia Matters, a podcast about Alzheimer's disease. Dementia Matters is a production of the Wisconsin Alzheimer's Disease Research Center. Our goal is to educate listeners on the latest news in Alzheimer's disease research and caregiver strategies. Thanks for joining us.

Dr. Nathaniel Chin: Welcome back to Dementia Matters. Continuing our series on the National Strategy for Alzheimer’s Disease Data and Research, today I’m joined by Dr. Sean Mooney, associate director of technology for the National Alzheimer’s Coordinating Center. He is also a professor of biomedical informatics and medical education at the University of Washington in Seattle as well as chief research information officer for UW Medicine. As part of his work with NACC, Dr. Mooney analyzes data to generate new hypotheses about how Alzheimer’s disease progresses in order to come up with new interventions. Dr. Mooney joined me live at the Spring 2022 Alzheimer’s Disease Research Center Meeting, hosted by NACC, to discuss his work with data on neurodegenerative diseases, the ways NACC protects research participants’ privacy, and how this data can be used to better predict Alzheimer’s disease risk and allow earlier clinical interventions to be made. To hear more from Dr. Sean Mooney and learn more about NACC’s work in Alzheimer’s disease research, register to join their Fall ADRC Meeting hosted virtually and in-person in Chicago, IL from October 20th to the 21st. Registration is free and open to the public. The link to register can be found in the episode description. Welcome to Dementia Matters, Dr. Sean Mooney.

(Music Interlude)

Dr. Nathaniel Chin: Welcome, Dr. Sean Mooney, to Dementia Matters.

Dr. Sean Mooney: Hi.

Chin: So, why don’t you tell us a little bit about your background.

Mooney: Great. I’m super happy to be working with the National Alzheimer’s Coordinating Center. I’m currently a professor at the University of Washington in biomedical informatics, so basically I’m a computer guy. I’ve been at the University of Washington for about seven years. Before then, I was at a nonprofit private research institute called the Buck Institute for Research on Aging, where I also led informatics there. We worked on lots of aging-related diseases, which included Alzheimer’s disease, and trying to understand essentially molecular mechanisms of how diseases – particularly aging-related diseases – are caused. We would do a lot of data analysis and things like that. I’ve really spent my whole career doing data, trying to interpret data, and trying to use data to either predict things or to try to better understand how patients will respond to therapies, how patients will – over time, how their disease or condition will progress, who’s at risk for different conditions. Our thought, in my world, is that if we have more and more data about people, who are also patients or research participants, if we have more data about them earlier we’ll be able to predict their future better and, therefore, also give them better clinical care.

Chin: Ok. Now that’s starting to make a lot more sense as to why you also work with NACC – big data systems, collecting of data, harmonizing data as Dr. Biber just explained. So what is your role with this new expansion of NACC and the things that Dr. Biber has mentioned?

Mooney: Sure. You know, neurodegenerative diseases – it’s an organ based disease. It’s extremely complex. It’s hard to study because it’s our brains. Unlike, say, cancer where often you could have, maybe, a solid tumor that might be removed or it may be in a tissue where you can get access to it, in neurodegeneration we don’t have that luxury. We can’t easily actually interrogate the disease tissue in a way that we would like when we do clinical research. That means that we have lots of innovative and really clever ways of getting access to data that characterizes patients or, if they’re participating in research studies, research participants. That could be imaging, like MRI images of the brain or PET scans of the brain. It could be genetic information. It could be proteins or other molecules that are in the blood of the patient. You know, things like that. It could be the behaviors of the patients, you know. Perhaps we can detect, based on early behavior like activities and things like that that we might be able to monitor in a patient, that we may be able to detect a patient who’s going to decline cognitively and eventually have dementia in the future earlier, because then we can design clinical studies for new drugs that may be able to delay the onset of dementia. It could also potentially give us insight into who is at risk and be able to make other clinical decisions based on that.

Chin: So because we cannot get brain biopsies, we need other forms of data to understand people and then be able to use that in these predictions that you’re talking about. So, you being a data guy, what data are you most excited about when you think of the field of Alzheimer’s and aging?

Mooney: Oh, that’s a super good question. What data am I interested about? Hmm… so I spend a lot of my time doing genetics. I think maybe that’s where we can connect it to this talk that I’m about to give about digital biomarkers. A lot of us now have smartphones. I have an iPhone in my pocket. You can’t see it, but I have an Apple Watch on my wrist. They are collecting data about me. They are collecting the number of steps I take. The Apple Watch is measuring my heart rate while I’m sitting here. It measures when I sleep. It measures my body temperature. It measures all sorts of things about me. I think that there may be things that we can measure, as we normally do kind of now in our everyday lives when we put phones in our pockets, that will be indicators or predictive of risks of getting dementia earlier before a doctor actually makes a clinical diagnosis. There’s some evidence in literature that this can happen 10-15 years even before a diagnosis is made, where you can see risk factors, that are essentially what I call, subclinical, that may be able to give us insight into a patient that’s going to get dementia or get Alzheimer’s disease. If we can make that prediction 10-15 years before, we can intervene because at that point the disease has already began. It just hasn’t gotten to the level that a doctor would diagnose. By the time a diagnosis has happened, there’s been damage that’s been done and the process is not likely to be easily reversible, so we want to catch it as early as possible. There's growing amounts of evidence that there are keys that are predictive of these changes and that those could be in behaviors that a patient might be doing. It could be present in their voice. It could be simple memory issues that may be coming up. There's a lot of things that we might want to measure that we can use to be predictive. Again, this is all hypothesis at this point. What we want to do is show that we can take this data, that this data is predictive, and then show that that prediction can be used clinically. And I just described the next ten years of my research right there.

Chin: And not all data is good data though, right?

Mooney: That’s true.

Chin: How do you know what is going to be meaningful and what is not going to be meaningful? What if you don’t have an iPhone and you have a different company or a different type of smartwatch? How do you compare apples and oranges?

Mooney: Right. This would be easier if we knew what data was going to be interesting, and in science we often don’t. We’re doing research. Nothing that I’m describing is something that would be used clinically. It’s purely in the research space, and what we want to do is show that some of what we’re looking at is interesting and translate that to be used by your doctor when you go for a clinic appointment. That’s the grand challenge is how do we identify what’s interesting. So how do we get over that? We collect a lot of data. We have a lot of hypotheses. That’s how science works, right? In science, we form a hypothesis. We design an experiment to test that hypothesis. We, then, observe things and we see if that hypothesis works. With the data world, we are generating lots of hypotheses because we are collecting lots of data. We can test them all simultaneously using statistics or other forms of data science, and those will tell us what those key indicators are that we think are predictive. Then we can put those together to be able to think about how we might be able to translate that clinically.

Chin: When you talk about big data, one of the key questions to me is privacy and security. And so, what do big data collectors and people who study this, what do you put into place to make sure that your participants are safe and that their information is protected?

Mooney: Sure. So every research participant that participates in the studies of data that we collect go through a process of informed consent. They have a consent form that describes how we will use our data. When researchers actually get access to our data – after it’s been collected, we collect the data together. We integrate it with perhaps other data sets with other centers that are across the United States. We create a single database, and then we can make parts of that data available to other – to investigators who want to do research on that. When investigators want to do research get access to the data, that data has been deidentified. So, we’ve removed the identifiers of individual participants and we’ve made it very difficult for them to be able to take that data that they have and map it back to an individual. Every investigator when they get access to the data, even though it’s been deidentified, they also sign what we call a data-use agreement, which is a legal agreement that says they will treat that data very securely. They will treat the participants in there and respect their privacy, and they will not try to reidentify anybody in that dataset using any means that is possible. We hold those investigators, with that legal agreement, to basically securely and, you know, treat that data secure and privately. We also manage all of the data centrally within our systems and the centers across the country, we manage the data in a computer system that is secure. It’s similarly secure to like, say, how your doctor might manage your clinical medical record. There are both federal standards by how we do that, and then each center and NACC, for example – the National Alzheimer’s Coordinating Center –  also has their own standards about how we can securely manage data, protect the privacy of our individuals, and minimize the risk of having a breach or something like that where some bad actor would get access to that data. But in general, most of the data is deidentified.

Chin: And so you mentioned earlier that some of your research has been in genetics. Is that your – what is your current investigation? What are you currently working on?

Mooney: That’s a great question. So when I was a graduate student I was a chemist. I was doing – building computational models, molecular models of human diseases at the molecular level so that we could try to understand and design drugs that might intervene in whatever condition that I was studying at the time. As I’ve advanced in my career, I’ve realized that we had these what you just called ‘big data approaches’, these approaches where we’re collecting lots of data that we really don’t understand how that data might be useful clinically. When I moved to the University of Washington, I took the role of chief research information officer of the University of Washington’s Health system. We’re four hospitals, almost 100 primary care clinics, and maybe 300 or 400 specialty clinics and a cancer center. We have tons of data that’s being collected from patients as they undergo their clinical care. I believe that that data is going to be very, very important to advance both genetic and other molecular studies of Alzheimer’s disease, neurodegenerative diseases, and frankly any condition. Twenty years ago, when I was an assistant professor – I was at Indiana University in the department of medical and molecular genetics. We would do genetic studies on say Parkinson’s or Alzheimer’s disease or something. I would get access to data that had been collected. There was almost always an Excel spreadsheet that had a column in it that was, “Alzheimer’s: Yes or no?” It was basically a yes or a no, whether the patient had Alzheimer’s based on certain criteria. I reasoned at that time, and I think other investigators have too, that in the future we’re going to want to know a lot more about those patients than just a yes or a no. And in fact, if you’ve ever met a patient with a neurodegenerative condition like in Alzheimer’s or Parkinson’s disease, they are all different. Distilling them down into just a yes or no is probably going to be insufficient for research. Now that we’ve gotten to a point where we’ve discovered a lot of genes and genetic elements across the human genome that are associated with neurodegenerative diseases like Alzheimer’s, we need to take that further. We need to understand the phenotypes and understand what’s interesting from those phenotypes to be able to do something or cluster those patients into groups that we can then use for different treatments to improve their care.

Chin: What are you taking away from this spring ADC meeting? What are you excited to be hearing about?

Mooney: First of all, I think one of the most important things that I’ve seen here that I’m very excited about is that the federal government - the National Institutes of Health, which funds the majority of, frankly, biomedical research that happens globally – has been investing in the National Institutes of Aging to the point that it has become one of the top five funded institutes for doing research. Aging-related diseases such as cancer and neurodegenerative disease, and all of the aging-related arthritis, conditions that we get, are all historically very understudied. We need to understand both aging and the molecular mechanisms of aging. There are tons of hypotheses about how humans age that are out there. There’s a ton of data and scientific literature both in humans and in model organisms, like rodents but also invertebrates like nematodes or drosophila fruit flies, about aging that we now understand a lot about the mechanisms of aging. But we really, in my view – and maybe there are some scientists that are listening to this podcast and disagree with me, I’d be happy to hear their comments by email or something like that – but I really think that we haven’t really put that together to really understand aging. I think that’s critical for us to understand something like dementia or Parkinson’s or Alzheimer’s, whatever, because let’s say a patient comes into a clinic visit and we sign them up for a research study. We do a blood draw. Then we take their blood and we look for protein biomarkers that we think are associated with Alzheimer’s disease. Well, the protein biomarkers in our blood change as we age and we age at different rates. We know that. So, we need to understand aging to be able to understand biomarkers better. We need to understand aging to understand what are the underlying causes of neurodegenerative disease better. I think that’s a super – I can’t hit that point hard enough. It’s really great to see that the NIH is investing that level of research into these conditions and into the basic biology of aging because I think that’s really key for us to understand aging-related disorders. I think, first of all, that’s the main message that I see at this meeting that I think is really exciting. I think the second main message that I’m hearing is actually transformative changes at the National Alzheimer’s Coordinating Center, which I’m conflicted because I work there, about how we’re modernizing our systems to be able to give broader and better access to data – big data – to enable investigators to make whatever those next set of discoveries are going to be on how to treat patients with dementia better.

Chin: Well thank you, Dr. Sean Mooney, for being on Dementia Matters.

Mooney: Happy to be here. Thank you very much.

Outro: Thank you for listening to Dementia Matters. Follow us on Apple Podcasts, Spotify, Google Podcasts, or wherever you listen or tell your smart speaker to play the Dementia Matters podcast. Please rate us on your favorite podcast app -- it helps other people find our show and lets us know how we are doing. Dementia Matters is brought to you by the Wisconsin Alzheimer's Disease Research Center at the University of Wisconsin--Madison. It receives funding from private, university, state, and national sources, including a grant from the National Institutes of Health for Alzheimer's Disease Centers. This episode of Dementia Matters was produced by Amy Lambright Murphy and edited by Caoilfhinn Rauwerdink. Our musical jingle is "Cases to Rest" by Blue Dot Sessions. To learn more about the Wisconsin Alzheimer's Disease Research Center and Dementia Matters, check out our website at adrc.wisc.edu, and follow us on Facebook and Twitter. If you have any questions or comments, email us at dementiamatters@medicine.wisc.edu. Thanks for listening.