Bobbi Pritt, M.D. (00:05):

Hello, I'm Dr. Bobbi Pritt, a clinical microbiologist and laboratory leader at Mayo Clinic, and your host for today's episode. I'm excited to be here today with Mayo Clinic Laboratories CEO and president, Dr. Bill Morice, to get the latest news. And then we will have Dr. Chris Garcia join me for a conversation about how to make sure the data you're using for research and development drives meaningful insights. But first, Bill, welcome.

William Morice II, M.D., Ph.D. (00:31):

Yeah, it's good to be here, as always.

Bobbi Pritt, M.D. (00:33):

As always. I have a couple of topics that I thought would be really interesting to talk to you about. One is a topic that we've touched on a little bit in the past about how Mayo Clinic Laboratories collaborates with innovative organizations to bring new diagnostic solutions to patients and providers. And one of those collaborators is C2N Diagnostics. And, great news, I recently saw that their Alzheimer's disease test received a new FDA clearance. Can you tell us a little bit about that?

William Morice II, M.D., Ph.D. (01:05):

Yeah, happy to. It's really interesting, right? Because our conversations with C2N have been taking place over a number of years. It really goes back to, when you think, probably like five, six years ago when the Biogen drug, which didn't really take off, but we had therapies for Alzheimer's disease, which of course begged the question of the need for accurate diagnostics for Alzheimer's disease. And at that time, the tests were either CSF-based, or, the gold standard is PET scanning to look for the amyloid plaques and tangles. But what C2N has done early on, they were spun out of Washington University, they actually use mass spec to look for proteins in the blood and ratios of protein in the blood that are associated with amyloid and deposition in the brain, which is of course are key pathologic features of Alzheimer's.

So, they've been working on it a long time. We were drawn to them as an early collaborator. Number one, because their technology was such that they have patented it, so you had to work through them to make it available. But also because Dr. Joel Braunstein is the CEO. It's very patient-focused, science-focused, really trying to create a diagnostic that would be helpful to patients. And we knew that this was on their timeline, that they wanted to study this to really prove its effectiveness in the diagnosis of Alzheimer's disease. I was happy for them to see them get FDA approval. It's great for patients because the data showed even in patients down to the age of 40 with symptoms of cognitive decline, the test actually has greater than 90% positive predictive value and negative predictive value.

So, it's quite a useful test. And the other thing is actually Dr. Ron Petersen here at Mayo Clinic has been a collaborator as has Dr. Algeciras, and it's actually grown our ability to offer testing even from our own laboratories because, you know, providers, patients, are looking for different diagnostic solutions. So, it's actually been kind of a real win-win.

Bobbi Pritt, M.D. (02:52):

You know, I really love that about this type of collaboration because you have a specialty test that you have to collaborate with that group to be able to offer it. It's an important test. We want to be able to offer it to our patients, but we also need the full spectrum very often, in the workup of these patients, of different tests. And so, Mayo Clinic Laboratories, through the Department of Laboratory Medicine and Pathology, can offer the full spectrum of tests, but incorporating this other test from C2N.

William Morice II, M.D., Ph.D. (03:21):

Exactly. And most importantly, kind of the knowledge about how to use these different tests, which people look to Mayo Clinic for a lot. For you, to me, to others within the Department of Lab Medicine Pathology and in the practice that have this expertise. And most importantly, it's how patients can really benefit. It's quite gratifying.

Bobbi Pritt, M.D. (03:37):

Yeah, that's great. Well, another topic that I wanted to talk about is a little different, but again, a topic that comes up quite a bit, and that's about the use of AI. And there was an interesting Wall Street Journal article talking about how AI is helping patients and providers solve medical mysteries. There was a nice summary of the article from Becker's, but basically AI is helping patients take a little bit more control of their own healthcare and help their own physicians solve these diagnostic questions. And also, it's helping physicians shorten the diagnosis time for rare diseases.

William Morice II, M.D., Ph.D. (04:15):

Yeah. It's kind of funny because it does dovetail a little bit into even the prior topic in terms of information and groups of information diagnostically that patients and providers can really trust. When I read that article, it really brought me back to thinking about one of the major challenges that we have in healthcare and as a Mayo Clinic is how do we make the knowledge of a Bobbi Pritt or someone else scalable? And it's interesting that the example from Mayo Clinic that they show is cardiac amyloidosis. Every year at our field staff meeting for Mayo Clinic Laboratories, we have a patient story. And one of those patients was Matt Millen, who's a well-known NFL football player and general manager for the Lions, who actually developed cardiac amyloidosis and it was missed. But, he was getting short of breath. And so, it is a difficult diagnosis.

AI is great because it can help identify patterns that might not be obvious to someone who doesn't have years of experience and actually get that patient an answer more quickly. But it doesn't mean it's the be-all and end-all either. I think that's the flip side of it.

Bobbi Pritt, M.D. (05:16):

Yeah. I've been reading a lot about AI and of course it's so important keeping the human in the loop, but also understanding that, for the most part, AI is not replacing jobs left and right in healthcare, despite some of the earlier fears and some predictions, but it's changing the way we do our jobs. I mean, I look at how we're doing our job in the parasitology lab, and we're using AI to shorten the time to diagnosis and improve our ability to detect pathogens, but it's still our highly trained, highly skilled technologists and the laboratory director, me, making decisions. So, I think in healthcare, it's really important that you can't just turn it over to all of AI, you still need the human touch, the human interaction, and the human decision-making.

William Morice II, M.D., Ph.D. (06:02):

Totally agree. And on the one hand, wait times for patients to get time in front of a physician or provider are longer than ever. I can speak from my own personal family's experience. So, these things really make that healthcare more accessible. It's a new set of tools that as educators we have to teach healthcare providers to use, because I do believe the adage that I've heard that AI won't replace physicians or physician scientists or healthcare providers, but those who use AI will probably ultimately end up replacing those who do not. So, it's going to be a long road, but the most important thing is we're finding use cases that actually can improve people's lives.

Bobbi Pritt, M.D. (06:38):

Yeah. Now, I'm interested, Bill. One of the things in the article talked about how AI could improve test stewardship because perhaps a family is going to use it to narrow down their differential of what a family member might have, like a parent trying to figure out what their child might have. I think the flip side could be that it actually might be more challenging for test stewardship if people are coming in to their provider with a list of things that AI told them their child might have. So I don't know, what are your thoughts on that?

William Morice II, M.D., Ph.D. (07:08):

Yeah, that's really an interesting take. Because on the one hand, that's really AI in the consumer hand, right? Does that improve efficiency of care, or does it actually obfuscate some things or make things more confusing for a provider and a patient both? Particularly if a patient, they're very believable, AI is designed to make you really believe what it's telling you. And so, someone could come in with a conviction they have something or other. Now, for test stewardship, I do think AI will change what we think of as clinical decision support. But that operating in an environment where Dr. Bill Morice is seeing a patient and ordering a test based on a differential and AI flagging to say, "Hey, maybe you really want to get this, or maybe you should really do something like that." To me, that's a totally different use case. I think those internal use cases around stewardship, the radiology, other diagnostic modalities, makes a lot of sense. In the hands of the consumer, I think you have to look at it in that way.

Bobbi Pritt, M.D. (07:59):

Yeah. I agree. Two totally different situations. Well, Bill, as always, it's great talking to you about all these interesting things going on in our field. Thanks again for joining us today.

William Morice II, M.D., Ph.D. (08:08):

Oh, I will always relish the opportunity. So thank you, Bobbi.

Bobbi Pritt, M.D. (08:16):

Welcome to the deep dive. We're going beyond the headlines today with Dr. Chris Garcia, Mayo Clinic Laboratories chief digital innovation officer. Together, we'll explore how advances in data capabilities are fueling research and development in laboratory medicine. Dr. Garcia, welcome back.

Chris Garcia, M.D. (08:34):

Thank you so much, Bobbi. It's a real pleasure to be here with you.

Bobbi Pritt, M.D. (08:36):

Yeah, it's always great to have you with us. So, I have a number of questions for you today. I want to talk about research and development and data and laboratory medicine. I'm going to start with a relatively basic question first. How has the role of data in research and development evolved over the past decade?

Chris Garcia, M.D. (08:57):

I love that question. It is significant. Data has, let's say this, 10 years ago, a lot of the questions that came out were, "Can we get the data?" And it was a bit like rolling a stone uphill, and you kept on trying to get the data and beg, borrow, steal kind of, especially when you're trying to do something really significantly different in laboratory medicine. Today, the question is less “Can we get the data?” and now it's more "Can we make the data mean something for patient care?" And that's an exciting change. I mean, just think about our lives and how data-driven they are with your own phone. Think about your work life now and how many alerts, how many reports you get. Volume and variety of data have exploded. And in the medical world, the benefit is that we have such a huge variety of modalities.

We have results, we have imaging that didn't exist in the lab before, but we have now. We have genomics on different levels that we didn't have 10 years ago. We have notes and outcomes, and people have invested in data systems. So, it's exciting, even though the lab is a crucial part of medicine, I don't think people realize how much data exists and is created and supports what we do in the lab. And now, we have this huge objective to make that data mean something as much as possible. One of the other things I've seen, it's a little adjacent, but it really fits us well, is usually we would generate data with controlled trials in the past. And now, we're trying to take advantage of that real-world data that comes from those years of what we've been doing to take care of patients.

The data's more messy than it was, and it takes a skilled team to be able to find the signal through the noise, but it is a real benefit. And we're seeing it in our partners throughout the world who are using that data to create new safety precautions, new levels of quality, and new capabilities for our patients in lab medicine. I think that's really exciting. And we went from data being more scarce and just rare to being more abundant. The scarce thing now is having the right context.

Bobbi Pritt, M.D. (11:18):

Yeah. You know, that's so true in my own experience as well. We have all this data. And so my follow-up question is then, with having access to more data than ever before, how do you turn that data into meaningful laboratory insights?

Chris Garcia, M.D. (11:34):

Yeah. I think we're all struggling to figure out how to make that happen. Because while we have more data, having more data does not automatically mean we have better answers. And sometimes it means that we have more ways to fool ourselves, which is a little concerning. I actually had this great panel this year that I was a part of, and we had someone from a microbiology testing company, data scientist. We had one of the leads of science at Thermo Fisher over the big Olink proteomics offering. We had a Mayo physician, who had developed a precision diagnostic, and we had a DARPA program lead. And so, big mix. You've got devices and labs and scientists and clinicians and defense. And we asked, "If you could put money, would you put it on compute, would you put it on better data scientists, or would you put it on better data?" And all of them, without hesitation, said, "Put my money on the data."

I think the thing is creating the quality of your data, curating it, taking care of it, is the best thing that we can do to take our data and turn it into insights. Usually, I believe, this tells us kind of where the bottleneck is in research and development, in making sure that the data is clean, ready, and fit for use. And, this isn't too different than what we've had in the past. People who've worked in the data science realm, there's been this 80/20 rule that they've used a long time. 80% of the time is just cleaning and finding the data and getting it ready for use, and 20% is actually doing the data science work, trying to answer specific questions. And most people quote that even today, although the tools that are out there really do take the lift off the people.

So, it may be 60% of the time is now focused on using the tools, and they have some agents that help them fix that out. But what I would say in medicine, the biggest area of being able to take this data and unlock insights is tying it to the patient journey. I think we experienced this in our industry during COVID, when we were not only having the testing results, but marrying it up to the demographics and the patients, and then trying to marry it up to, "OK, what did this mean? Where was this in the patient's journey? Were they asymptomatic? Were they not?" And we were able to answer some really great questions, whether it was together with our colleagues for government insights or for scientific development. And it was really empowering because we moved fast, right? And it felt like we were making a big difference.

And I think we're feeling that same crunch when it comes to data today. How do we take that data, make sure that it's ready for use, match it to the patient journey, not only have better answers, but have better questions, and get to better answers that way. And the only way to do that is with clinical and scientific precision working together, I think. I've seen people be really successful in marrying those two up.

Bobbi Pritt, M.D. (14:43):

Yeah, that makes sense to me. What do you think about partnering to overcome some of these challenges as well, bringing people together, researchers, industry leaders?

Chris Garcia, M.D. (14:51):

It's necessary. I think it's a perfect segue, right? Because you need the scientific, you need the clinical, you need the market. Everybody has to work together. Whether it's an academic or an industry, you should start with a research question as opposed to a dataset. People can really fall in love with a dataset and then go look for a question. I haven't seen it work as well as when people have a very targeted question that they're going for. Another thing that's really key, I haven't been in a group where the idea of data sharing doesn't gather excitement. People want to collaborate together. One of the biggest things that kind of hinders that right now is the ability to share it itself. We all have great data, and we want to share it, but how can you share it? How long does it take? So, if you have an 18-month ramp-up to be able to get through all the hurdles and policies to be able to share it, the enthusiasm lags and a lot of the opportunities die.

So being able to work on being able to share data, knowing what you can share, what you can't share, how you can receive it, that tends to be really important for those collaborations. And groups who figure it out once, they usually like to continue to work together because it's that much of a difficulty. Looking for people who want to answer the same question together is step one. Step two would be, make sure that you can share that data. And I think the last one is making sure that you're working together. If one partner takes all data and says, "I'll do all of the work and I'll share it back," it's not the collaboration. And those don't tend to be very long-lived.

Bobbi Pritt, M.D. (16:30):

Yeah, it's not really the definition of a partnership.

Chris Garcia, M.D. (16:33):

That's fair. Absolutely.

Bobbi Pritt, M.D. (16:35):

Well, looking ahead, what is the most exciting to you with the way data is shaping the future of medical research and development?

Chris Garcia, M.D. (16:42):

There's a lot that's going on that's exciting. One that I'm feeling every day, and I feel like a lot of colleagues, we’re all mentioning it, is having artificial intelligence tools working alongside us is key. We talked about the quality of it and the kind of the explosion, but that diversity is huge. And as a human being, we can only handle so many of those different variables in our head at the same time, even on paper. And these tools allow us to get the data, harmonize the data. Goodness, people build APIs so that you can self-serve the data. And being able to answer more questions faster because you have these new capabilities is key. We have to be responsible. We have to understand what's going on there, but I think that AI alongside our researchers is huge. Another one would be, I think in the past, and I actually don't think this is bad, we had been focusing on smaller datasets, answering very specific questions, probably more focused on our own silos. And now, we're able to not only answer the question right in front of us in our own lab, but work together to answer outside the lab and maybe more predictive. I think the predictive capabilities are a key part of what's exciting and what will this mean for care. How does a clinician, how does a patient, think differently when they've got a predictive score versus something that, you know, we're saying is in their body right now or is going on?

And then the last one that's really exciting, and maybe it ties back to our experience in COVID, but it's closing that therapy to diagnostic gap. When you look at when new therapies come onto the market, there's a lot of data, just depends on the area, whether it's cancer or immunology or whatever it may be. But there's usually a fairly long gap between when a new therapy is out on the market and when the laboratory diagnostics are out there able to inform what this means for patients, whether it's monitoring or side effects, something outside of the companion diagnostic. I see a lot of really exciting research where folks are able to generate the data real-time and answer these questions. And I think that those are exciting for me, the prediction, the AI support, and then really making sure that the diagnostics are there to support all these novel therapeutics that are changing our lives and our patients' lives.

Bobbi Pritt, M.D. (19:17):

It is really an exciting time, isn't it?

Chris Garcia, M.D. (19:19):

It is. I love what I'm able to do. I love working with my colleagues because there's so much energy, but also there's such a great level of responsibility. We're at this point, we do what we do as laboratorians, as pathologists, as physicians, as scientists, and how do we make sure that we provide the value that really makes the biggest difference? And I think data is continuing to be an important part of all of our jobs, but R&D is kind of the place that you see it first.

Bobbi Pritt, M.D. (19:49):

Yeah, that makes sense. Well, Chris, as always, it's so great to have you on, and this was just an inspiring conversation. So thank you again for joining us today.

Chris Garcia, M.D. (19:59):

Oh, it was my pleasure. Thank you, Bobbi.

Bobbi Pritt, M.D. (20:06):

Let's wrap up with the top takeaways and how to learn even more on the topics we discussed. During our news segment, Dr. Morice discussed FDA clearance for an Alzheimer's blood test offered through a Mayo Clinic Laboratories collaborator, C2N, and then we discussed how AI is helping solve complex medical mysteries and what that could mean for laboratories. Learn more about both topics through the links in the show notes. Then in today's deep dive, Dr. Garcia shared how the growing availability of data is changing how data is used to drive meaningful research insights. We've included links to a couple of past episodes with Dr. Garcia if you'd like to learn more about how growing data capabilities are impacting laboratory medicine. Thank you for joining us today. If you haven't already, make sure to subscribe so you never miss an episode. Next time, I'll be joined by Dr. Leslie Donato to explore the latest on testing for bile acid malabsorption. I hope you'll join us.