Using Genetics To Discover New Treatments With Regeneron Genetics Center's Aris Baras, M.D.

On this week's episode of the Business of Biotech, Aris Baras, M.D., SVP, Head of Regeneron Genetics Center (RGC) and Co-Head of Regeneron Genetic Medicines, talks about using large-scale human genetics to find better drug targets based on "superhuman" variants. Aris explains how multi-omics, linked health records, and AI are helping developers move faster in the clinic and reduce the odds of failure, while also pushing biotech toward earlier disease prediction and more preventive care. 

Access this and hundreds of episodes of the Business of Biotech videocast under the Business of Biotech tab at lifescienceleader.com.  

Subscribe to our monthly Business of Biotech newsletter.

Get in touch with guest and topic suggestions: ben.comer@lifescienceleader.com

Find Ben Comer on LinkedIn: https://www.linkedin.com/in/bencomer/

Transcript

Welcome And Guest Overview

Ben Comer 0:03

Welcome back to the Business of Biotech. I'm your host, Ben Comer, Chief Editor at Life Science Leader, and today I'm speaking with Aris Baras, M.D., Senior Vice President, head of Regeneron Genetics Center, and co-head of Regeneron Genetic Medicines. Aris joined Regeneron in 2011 and helped found the Regeneron Genetics Center, or RGC, in 2013. The RGC database now contains over 3 million sequenced exomes through collaborations with more than 150 global partners. Research at the RGC has produced over 50 new genetic discoveries and has moved more than a dozen programs into the clinic. Last April, Regeneron received FDA approval for Otarmeni, excuse me, the first gene therapy approved to target the genetic cause of an ultra-rare type of hearing loss, a treatment capable of actually restoring natural hearing ability in patients. Qualifying patients in the U.S. can receive the treatment for free, which is amazing. I'll ask Aris for a little bit of background on that gene therapy and how it came together. We'll talk about the work RGC is doing in genetic sequencing, RGC's global partnerships, and how it all translates into new therapies. Aris, thank you so much for being here.

Aris Baras, M.D. 1:23

Thank you, Ben. Pleasure is all mine. Love your podcast, and I'm thrilled to be doing this with you. Thanks so much.

Aris Baras Career Detours

Ben Comer 1:30

Thank you so much. I wanted to start off with your background. Um, you got uh several um diploma, several degrees from uh Duke University, a BS, an MBA, and an MD. Uh, and also, I believe, consulted for Liquidia Technologies during that time. Is that is that right?

Aris Baras, M.D. 1:49

That's correct. You know, I was young. I didn't know what I wanted to do. I was trying to figure it out.

Ben Comer 1:53

Yeah, of course. And I I was curious, you know, about what kind of career you were envisioning while you were at Duke.

Aris Baras, M.D. 2:02

Well, look, I love science, medicine, and the business of it all, right? And so I really got into all of that. Uh my background, you know, I started learning about science and got exposed to some amazing mentors and labs back in the DC, Maryland area where I grew up. So I'm referencing the NIH. And so when I went down to a place like Duke with enormous opportunities to learn more and explore in these areas, I loved it. And then I started learning, hey, maybe it wasn't solely a basic science interest I had. Maybe it was something that was a little more translational or clinical in its application. And, you know, got into medicine. And then it was, well, maybe it's not solely that, right, in terms of the practice of medicine and making a difference for patients one by one, but it was also more of this innovation and entrepreneurial angle to it. And it was, what if could I have an impact on actually building new products, right? And the future of medicine and helping not one patient at a time, but who knows, maybe millions of patients, right, with these types of medicines. And Duke was really a wonderful place to experience and explore what an institution, right? How many success stories in terms of amazing science that's spun out, people that have involved in these types of companies and could give you confidence you could do something like that, you know, mentors everywhere. And I got involved in some, as you mentioned, consulting with some local, really prominent, um, seasoned drug developers, venture capitalists, a company Liquidia Technologies. You know, one thing gets to another. That's actually where I met Roy Vagilos through an amazing story, a senior year class, a Duke about leadership, where they absolutely push you not just to learn in the classroom, but to absolutely find a living, breathing leader that's really kind of top in the field in whatever you're doing, and to make that connection and learn from them. And then Roy obviously got me connected with Regeneron, and you know, here I am at Regeneron. Um, but it's amazing how life works and how those stories happen and how you have plans and the detours are more exciting and all those kind of things. So um, yes, a long time, well worth it in retrospect at Duke. Uh learned a lot. And as I mentioned, um had an amazing um amount of detours and and fun along the way.

Ben Comer 4:32

Okay, so that's how you got to Regeneron in 2011. Um I actually uh side note, but uh in 2013, I took the train up from the city to Terrytown and uh interviewed George and Lennon for a profile piece in pharmaceutical executive. And I guess you you were you must have been j the the Regeneron Genetics Center, we didn't talk about at the time, but it was probably something that was brand new and kind of happening behind the scenes while I was there.

Aris Baras, M.D. 5:03

It was super top secret. We couldn't talk about it just yet.

Ben Comer 5:07

Well, let's talk about it now.

Why Regeneron Built The RGC

Ben Comer 5:09

Uh let's talk about the database you're building. Um, what what were the uh initial reasons uh for creating the database? You know, what what was the kind of use case that you guys put forward?

Aris Baras, M.D. 5:21

Well, thank you for that question, because it is absolutely one of my favorite topics. So, what we're all about at Regeneron, and if you look at our history, it's a lot of science and technology. It's not just drug development. And it's a lot about having a perspective on what are some of the biggest bottlenecks, right? And how can we break those bottlenecks with innovative science and technology? And that's all about differentiation, right? We're really competitive field, a lot of hot science, a lot of investment in big drug development programs. How can we be different? How can we be ahead of the curve? If you look at, I'm a bit of a historian, if you will, in uh, you know, with all that time I spent in school about our pharma and biotech industry. And the short version of the history is you know, this all started with screening natural compounds, right? It was a lot of luck looking at plants and looking at marine life, and what could we find that might have some benefit on some ailment, you know, killing cancer cells or halting immunity or something like that. Then luckily, people like you know, uh, my hero Roy Thagelos brought some amazing chemistry to things and a rational, target-based kind of screening process to uh drug design, right? The whole revolution with Merck and all that success. And then the biotechnology revolution, right? And you get antibodies and protein therapeutics and insulin and recombinant DNA. That's where Regeneron had its birth. And now we're living in an amazing time of that history in terms of all these um genetic technologies, sRNAs, gene therapy, gene editing. So that's amazing, right? And the simple point, the punchline is we have such a war chest, so many weapons to fight disease with. The problems still exist. What are the targets? Where to point all those amazing technologies, all those modalities, that whole war chest that I talked about, what are the targets? We did a fun analysis. We looked at all the targets that have ever been pursued, all approved medicines, everything that had been tried in a trial. And it was something like maybe 10% of the genome. 10%, right? There's 20,000 genes, and we'd only been able to really go after about 10%. So clearly, we don't know what's causing all these diseases, let alone what are the targets we can pull on, what levers can we pull on to try to fight against them, right? Try to develop cures. And that's why our industry has a 90% failure rate, right? I mean, most people don't appreciate that. We wake up every day and, you know, most of what we do fails, unfortunately. You know, we're 10% success mode. That's tough living. And so we were thinking about how do we break that bottleneck? And one of the things that was most compelling to us is human genetics. We were inspired by a few examples that came before our effort here internally.

PCSK9 And The Superhuman Idea

Aris Baras, M.D. 8:27

Things like the famous PCSK9 example. Amazing scientific breakthrough. Helen Hobbes, Jonathan Cohen, absolutely love them. Kudos to them. And they came up with this discovery. Dallas Heart study, amazing story. People down there in that study, there was a prevalence of a loss of function mutation, which means uh genetic variants that broke that gene, PCSK9, and those people had 90% lower rates of heart attacks and very low levels of bad cholesterol. It's like, wow. And then we're general and amgen, and then later L. Nylum and all these other companies now. Actually, I think wasn't today the news that maybe uh I think Merck had their or That's right. Yeah.

Ben Comer 9:14

Merck's got a small molecule uh product now targeting PCSK9. And we might even be getting close to uh a gene-edited, you know, kind of one and done uh knockout that's entering late stages.

Aris Baras, M.D. 9:25

You're absolutely right. So that was the inspiration, right? I mean, to think you could make a medicine when you've got that type of starting point, right? The the clinical trial has almost already been run, right? You can look at thousands of people with their lifetime worth of data missing that gene, right? And look at the benefits that they have. Okay, well, now all of a sudden it's not looking like 90% failure rate. It's looking like that's that could be a sure thing, or it's a higher probability. So that was the inspiration. And then the Regeneral Genetic Center was all about the question is this a one-off, a one-time thing, or do these uh gems, these treasures exist everywhere in all diseases if we just search enough uh human genetics genomes. So that's what it is. So it's a very rare thing to find that Dallas Heart study, that PCS canine example. As you said, we've now sequenced three million people. We look for these superhumans, right? We call them superhumans, the ones that have these genetics that protect them from certain diseases. And they're very rare. They're very hard to find. It's one in a thousand, one in ten thousand. And finding one of them is great, but to be absolutely sure, you want to see hundreds of them, you want to see thousands of them, you want to be absolutely sure that it's gonna be a great thing, it's not gonna have some sort of safety issue, and that you know how big of an effect it is on that disease. Simple math, and that's how you get to we need to sequence millions of people to find these. And then the the final, you know, conclusion here is after 13 years of doing this, it was not a one-off. We have found over a hundred, over a hundred of these amazing, kind of mesmerizing, almost unbelievable stories of protective human genetic factors. We've taken about half of them or so, maybe 50, and turned them into new medicines over a decade now into regenerance pipeline. So the impact has been very real. Things that are in late stage development to the hottest new targets at our company. So that's that's why we started human genetics. And as we kept getting a taste of success and learning that this was not a one-off and that we were finding them in every disease area, not just heart disease, neurogeneration, one of the hardest areas, you know, battlefields in drug describing development, even cancer. We have found protective genetic factors. Imagine that. People born with genetic variants that protect them from ever getting a cancer. Imagine the possibilities. So that's that's why we did it. And kind of a little fast forward here to 13 years later, you know, that's what's happened. We're even more excited now than we were back then.

Genetic Diversity And Global Partners

Ben Comer 12:18

I read that the Regeneron Genetics Center database contains the largest sequence populations of African, Asian, and Latino and ancestry. Um, most of the major genetic databases and tissue banks globally skew heavily to European ancestry. I think most business of biotech listeners will probably know why this genetic diversity is important for research, but maybe you could put a finer point on it if you would.

Aris Baras, M.D. 12:46

Absolutely. It's a great point. And maybe just a couple points on this one. Um it's a bit surprising that genetics can vary um so differently between different um, you know, we call them ancestral groups or ethnicities, yeah. Different ethnicities, people from different parts of the world. Um actually, let's tie it back to that story I told you about the Dallas Hart study. Those superhumans are actually more enriched, more common, more prevalent in African Americans. So that that protective human genetics mutation in PCSK9, I forget the exact number, 10 times, 100 times more common in African Americans than in every other type of person.

Ben Comer 13:30

Interesting.

Aris Baras, M.D. 13:30

So the fact that Jonathan and Helen could study such a large African-American population of patients who volunteered and joined in Dallas, that was make or break. That made that discovery. And that exists everywhere. We find those stories. We have a huge study going on in Mexico. Uh so people of Hispanic or Latino uh ancestry have very unique um genetics. People of Asian ancestry, people of European ancestry, you know, you get the point. But there are these amazing stories that exist in every pocket, in every population. The other point I was going to reference is, you know, our genomes are so similar, but there's also this slight variation. And that slight variation is not just maybe the difference between discovering a new target or not, but it's also a big factor in terms of how you might respond to a certain medicine. So that's been another major discovery for us, which is we have looked in our clinical trials and we've seen that maybe there's a different response to a certain therapy we have, right? Or someone else's medicine we can learn from. So that's a major reason. I mean, we're developing medicines here for all, for everyone. So it's really critical that we study the genomes and the genetics of the entire world.

Ben Comer 14:48

Well, I mentioned that the database is supported by over 150 global collaborations. Is that is that part of your remet, Aris, to identify those partners and expand the database? And if so, how do you how do you do it?

Aris Baras, M.D. 15:04

It is. Um so as you said, it's over 150. Um, you know, it's a deep breath to think about, you know, the last decade or so uh and the amazing, you know, fun life stories that's been from you know, working here with amazing collaborators in the US and other kind of mainstream areas like the UK. I mean, the UK has been a leader in genetics forever. But this whole journey, this amazing scientific exploration, I call it, you know, one of the greatest scientific studies or or explorations of our time, has taken us everywhere, from Iceland to South Africa, you know, to countries in Asia, um, small isolated geographic populations and founder populations, for example, you know, villages in India. Um, it's amazing. It's remarkable. Uh, and the team here has done a phenomenal job in being so collaborative and working all over the globe. And a huge thanks, you know, we always say thank you, like immense thanks to the patients who make all this possible, but also an immense thanks to all these collaborators, these dedicated scientists and clinicians who put these studies together around the world and make this happen.

From 3 Million To Everyone Sequenced

Ben Comer 16:15

The uh the Regeneron Genetics Center has now sequenced over 3 million exomes. Is it possible to reach a point of diminishing returns? I mean, or will you continue to build this database to as many millions of individuals sequence that as you possibly can?

Aris Baras, M.D. 16:35

You know, our vision now, over a decade into this, is for everyone to be sequenced. So let me answer your question a little more, you know, directly though. Um, why? 3 million is great, right? We're really proud of that. We're excited about that, but we're also really realistic and pragmatic. And there are a lot of limitations to our database. You can tell I could not be more enthusiastic about the hundred discoveries, right? The 50 programs in our pipeline. But for every one of those successes, there's a lot of strikeouts. And we're having a lot of success in places that we have a lot of data, in short. I mean, that that's kind of the magic about it. And we've got a ton of data in cardiometabolic disease and other very prevalent diseases. But there are been hundreds of diseases where we don't have enough data, right? Some of the less common autoimmune diseases, you know, some of the less common uh neurological disorders, uh behavioral disorders, right? Um, you know, so psychiatric conditions. And we just need to get to hundreds of thousands of patients worth of data. And then I'm so confident that discoveries will just roll off the databases and the analyses just like they've done for all those other conditions. So we do have to keep going. We've got to go from 3 million, we're already on the path to 10 million. We've put those collaborations in place, we have the samples flowing in. We're actually aspiring to get to 50 or 100 million. And now we have to rethink how we do this. I'm not sure this can be done, you know, with our amazing RGC team that I just talked about over a decade going all over the globe, we've got to think about something more scalable. I'm happy to talk to you about a pretty kind of revolutionary concept there of how do you truly get to tens of millions to 50 million people worth of data on a on a platform like this. But then the other perspective we have, and you know, we'll certainly touch on this, is how outside of discovering drug targets and inventing new medicines, what have we learned about the value of these data? And in short, we've learned that the data is immensely predictive and powerful for understanding everyone's future health risks. And as someone who trained in medicine went to medical school, I have a little bit of background in terms of, and we all do, frankly. We, you know, how we think about our health, we go to our doctors every year, hopefully. And we understand what can be predicted and what can be prevented and screened for and what can't, and what you kind of have to be reactive to. Unfortunately, a lot of medicine is reactionary. And what these data are showing us, and we're learning, is we can predict a whole lot. And so that's going to change medicine. We're going to be able to predict for everyone, not just us, but kind of the field, what are the few things that you have a very high likelihood of having, and they are really serious health concerns.

Ben Comer 19:37

Based on your genetics.

Aris Baras, M.D. 19:39

Genetics, your proteomic. Let's talk about proteomics as well. And all of a sudden, you can finally practice preventative medicine, right? Get ahead of things and fix it when it's easy or prevent it altogether, right? And that's when I put in that line, I really see a future where we're going to be generating the genome and the proteome on everyone, right? So there is no diminishing returns. There is no finish line here. We're going all in, right? It's everyone. We see a future. We want to be a part of that. Well, we're bringing genomics and preventative medicine to everyone. That's the future.

Exomes Versus Genomes Plus Proteomics

Ben Comer 20:17

Uh, you just mentioned uh the proteome. There's obviously the transcript, transcriptome, uh, other ohms, they're multi-ohms. Uh, you, you know, talk to me about exomes, maybe as compared to whole genome sequencing. You know, why why focus on exomes right now?

Aris Baras, M.D. 20:36

No, there's too many ohms. I get your point. Um exomes and genomes. It's all genetics, right? And, you know, we grew up here in genetics at a time where the costs were really significant. Right. So to sequence a whole genome, so the whole genome is that. It's everything, right? It's it's your several billion base pairs in every one of our genomes, right? That's the whole thing. Probably people are also familiar with something that's maybe on the other extreme of like 100%, right? And that's let's just pick like 1%, because it's so expensive, and let's just do kind of a minimal amount of sampling. Like let's pick like a million markers uh across the genome. And that's where those like genotyping chips came into play, right? And if anyone has done 23andme or Ancestry.com or a number of those other earlier uh approaches, they were using those types of chips and arrays, right? And more recently was uh kind of this era of of full sequencing, but focusing on the genes. So those 20,000 genes that are in your genome. And I'm not sure many people know this, but your genes are actually only about you know one or two percent of your entire genome, right? Those you know, three billion base pairs. It's maybe 30 to 60 million of those base pairs. Base pairs, right? So you can see how you can do the experiment, you can sequence everyone's genes and make it much more cost efficient, right? To be able to sequence 3 million people or 10 million people and get all the benefit of understanding what happens when I break a gene or I rev it up, turn it on, and what are going to be the health consequences. It's becoming much less of a strategic advantage for us, for example, to have been kind of a leader early days in large-scale exome sequencing, or to even have to make that choice because luckily the cost has come down so much and the cost difference between a genome and an exome is not what it used to be, right? It's just not that big of a deal anymore. So we're doing a lot of genome sequencing now, a lot of exome sequencing. Um and the path in the future will be we'll be transitioning over, and everyone, you know, as well. We'll just be doing it'll be cheap enough just to do genomes on everyone in these types of studies. The the much different thing, and this came much later in the RGC story, is proteome. Like, what does that own, right? Um, we're hardcore genetics believers, right? You know, we've got a hammer. Show us the nail, right? We want to you know seek exome sequence everything. All right. So we got pretty intrigued by this. What is the proteome? So if if your genome is your blueprint, it's your GPS, it tells you where things are going, it tells you how things are built, you know, what could happen, right? For the most part, genetics is we inherit our genome. That's what we're born with. Now, there's edge cases, you know, things can change, you can get cancers, and then your tumor has its own genome and it's awry, and it's you know, that's a really important area. But for the most part, genomes is what I said. It's what's inherited. It predicts you know what health risks you might have, what diseases you might have. The proteome is very different, right? It's not your blueprint, it's not your you know, GPS. Instead, it's it's really more like your sensor. The best way to explain it is so what are we doing? We're we're taking a blood test. So that's sampling your peripheral blood, right? Just blood flowing through your entire body, through your heart, through all your organs. And on that journey around the body, it's able to sample and sense things going on in every cell, in every organ in your body, right? Things are constantly turning over. Things are coming out of your cells in your heart, in your liver, from your bones, right? From your brain. And we get that blood sample. And now we're looking at about today, the technology is about five to ten thousand proteins. Eventually it'll get to all of them, right? We've got 20,000 genes. Some of them can give you multiple versions of proteins, but whatever, it's 20,000 plus proteins, right? And it's this amazing sensor. And so we're getting a sense up right now, not what could happen in the future, what's happening in the body. Fascinating concept. All right, we'll try it. So in true RGC fashion, we do a massive, let's call it pilot, right? Kind of a little bit of an oxymoron there. And we work with the UK, you know, the amazing Sir Rory Collins and his team there. Same team that we did the 500,000 largest study ever in genetics, 500,000-person sequencing project. We go back, you know, to the large study well with them and we do a 50,000-person proteomics study. Now we're completing that a couple years later and doing everyone, all 500,000 in UK Biobank, because it was so amazing, the pilot. In these 50,000 people, we've got these thousands of proteins in everyone. And we start asking the question, well, what can this really do for us? And we're looking at, hey, do these proteins correlate with diseases that these people, that these volunteers in the UK Biobank are getting a year later, 10 years later, 20 years later, uh, after they joined this study? It was amazing, Ben. I mean, the genetics are predictive, but we couldn't believe what we were seeing with the proteome. The ability to predict new onset heart disease, new onset neurological disorders, cancers, liver disease, hundreds of diseases we looked at and we couldn't believe it. It was just so predictive. So we're all in now. So we're doing the UK Biobank project. On all of our large U.S. health system projects that are enrolling hundreds of thousands of people, we're doing the genetics and the proteomics. So this 3 million person plus genetics database that's growing to tens of millions, it's not just genetics. Moving forward, we're also adding the proteome.

Health Records Tokenization And Scale

Ben Comer 26:51

And not just the omics. You're you're you've also got what, something like 300 million electronic health records and uh a kind of constellation of health data through partnerships uh as well. How do you how do you think about those data? Um, and and I guess combining it or analyzing those data together with the sequencing data?

Aris Baras, M.D. 27:15

It's a really important part of our future. So thanks for bringing that um important point up. There's really important basic stuff that we do in our industry, right? When we develop a medicine, we need to understand um, you know, the relevance of it. Um, what patients with a certain disease are experiencing today, right? What are the treatment uh paradigms, the standards of care? What are the unmet needs? What are the costs of managing a patient with a certain condition? And how important could a new therapy be that we're bringing forward, right, uh in that field? And so having data on hundreds of millions of Americans or in other major regions teaches us all about that, right? I mean, you know, in today's world, data is everything, right? We're seeing that from the anthropics and the open AIs about how data and you know, not just searches, but now you know AI type models is changing the way we work. It's the same thing for biotech and pharmaceuticals, right? The more data we have, the better we understand the patients we're serving, the products we're developing and bringing, you know, two patients, two countries, two markets. So that's kind of basic table stakes, really important for our business. Maybe not like anything earth-shattering and innovative, but the data itself is so valuable for our core business purposes and function, right? So we've been working very hard to bring in-house the highest quality of that type of data. And it's exactly what you mentioned. But we like to be innovative here. We like to think about, hey, could we do something different and game-changing? And it was a wonderful opportunity. So not only do we need that for the core business, but at the same time, we've got a big problem. How are we going to scale this decade-old process that's sending us to 30 to 40 countries on six continents, as I mentioned, you know, villages in India and isolated, you know, areas like Iceland, but anywhere we can launch studies and get appropriately consented patients and volunteers and collaborators to join us in this effort. So the new idea, the new innovation is hey, we've got something scalable here. We can work with health systems, with patients, you know, with great stewards of large health record data, places like the NHS, right, with the UK government. And let's really aggregate in a very compliance, secure fashion, hundreds of millions of health records, right? Now so that's that's one piece of the equation. The other piece of the equation is right, how do we get patients who want to, or volunteers who want to, you know, donate their biospecimens, right? So you can pair that. And there are places where you can do that, right? You can work with, as I mentioned, those direct-to-consumer companies, right? Or you could launch, you know, a project.

Ben Comer 30:15

Are you guys partnered with the like 23andMe or Ancestry? Is that useful data for you guys?

Aris Baras, M.D. 30:21

Um It's it's definitely useful data. Those companies should absolutely be commended for amazing work they've done. We've got a uh a fun and interesting history in this space. We've got wonderful collaborations with these groups. Uh, for example, when everyone was kind of working together on huge challenges like COVID, right? Um, kudos to all these companies who kind of said, Hey, let's quickly ask our consumers, will they give us permission to study COVID? And we need to we need to learn stuff quick, right? And, you know, we put, you know, those people were amazing, right? Those, those, those members, those consumers, they gave permission to use their data for this purpose. We all collaborated, worked together, had massive amounts of data quickly on people who had gotten COVID, who've gotten vaccines. Um, you know, it was it was pretty remarkable. And we found important things. We found important genetic factors that could really put you at really high risk of getting severe COVID and hospitalizations. We also found the protective genetic factors, right, that could really protect you from having anything bad happen to you, right? And we we published those, we made all that information public. So, yes, we've worked with those groups. We continue to work with those types of groups in specific areas and say, how can we work together, get large amounts of data for other disease areas? But there's other groups too. There's, you know, large, you know, labs, uh lab companies that you're aware of, you know, Quest and Lab Corps. There's blood banks all around the country, amazing volunteers, right, who donate blood. You know, all the companies that we work for, I'm sure, have been involved, or schools that we've gone to with blood drives and all that kind of stuff. Um, so we work with them and the hundreds of thousands and millions of people involved in that. And the point then that I'm getting to is those are also large-scale, like very scalable ways to get millions of people who are already giving samples or giving blood to understand studies and the impact of something like this, consent to be involved in something like this if they want to, and then get samples quickly, right? To the tune of millions on a quick amount of time. And once they've done that, you can link it to their health record. There's really nice technology. Um, it's privacy preserving record linkages. And in short, people call it tokenization. Once you get people's consent to join these studies, you create a token, right? So Ben, you know, or Aris becomes just a uninterpretable hash or series.

Ben Comer 32:53

Because this is a way to de-identify the patient. Exactly.

Aris Baras, M.D. 32:56

De-identify and link, right? And so that's that's what we're doing. It's the core, but it's also the innovative way of the future for us to carry this vision, this dream, and go from 3 million to 50 million or 100 million.

Ben Comer 33:09

Yeah, I think those linkages are key. I I remember, you know, this was several years ago, but 23andMe, they hired you know, Richard Scheller to head up their drug discovery uh kind of function within 23andMe. Uh I don't think very much came of that, but you know, they they have this huge data set, but uh it seems like you know you need to connect these different data sets to really uh have the discoveries come forward, maybe. I mean, I I don't I I never uh I never got to speak to Richard about about that work that he did at 23andMe, but it was always interesting to me to have you know a real heavy hitter in there working on it. And and it didn't maybe I missed it, but it didn't seem like uh a whole lot came from it. However, you know, a lot is coming out of uh of Regeneron and the database that you're building. And I wanted to see, uh Aris, if you could maybe describe internally at Regeneron how the database functions, and maybe we could use the gene therapy that was approved last April uh for uh you know uh an ultra-rare hearing disorder as an example, or if you want to use uh a different example, that's fine too. But can you take me kind of from discovery through clinical development, you know, and and uh you know, eventually to FDA approval?

Aris Baras, M.D. 34:27

Absolutely. By the way, I love Richard Sheller too. He's he's a heavy hitter. Absolutely what you said. One of my idols, heroes. Um he's fantastic and doing great work now at uh Bridge Bio. So, how does this work? We got this huge database. I love the examples that um you provided, right? Um, so we like to simplify things. And you can think of the way we do human genetics in maybe two ways. One of them is disease-causing mutations, right? And that's exactly the example you highlighted with the odopherolin, is the gene where there's a deficiency that causes hearing loss at birth and for life. So we can find those types of mutations. And Regeneron has done plenty in terms of working on rare disease therapeutics, right? We've got a couple approvals, not just that one over our time. Actually, our first approval was also for a rare disease like that. And there's probably a handful of them in our time at Regeneron. There'll be more. So we can discover that gene, maybe. We don't have to discover it. Maybe someone else discovered it. But then we can also use our genetics to learn a lot more about it, right? That makes drug development much more successful. We can understand with much more precision exactly how many people might be affected by such a condition, right? It makes a big difference. For a rare disease, you might have 10 people, right, in America. You might have 100 people, you might have a thousand people, right? Too many people that might say, what's the difference? That's all really rare. It's a huge difference, right? It might be the difference between, hey, this is actually developable and doable, or it's not. It's just we can't find enough people. And you see a lot of these companies, you know, especially smaller companies, will give widely varying estimates on how many people actually have this condition they're developing a therapy for. So in that case, for example, we were able to really help. Um, so the Regeneral and Genetic Medicines group that we mentioned in the beginning, kudos to them. Remarkable job. I think it was only two and a half years from the first patient dosed to getting uh that approval. Um, for me, it's kind of a once-in-a-career uh type of experience to see that type of uh story, um, a really you know touching and pulling um indication, right, with hearing loss and to see the types of effects you know published in the New England Journal in terms of you know um the potential there for um you know kids to hear again.

Ben Comer 36:59

Yeah, amazing.

Aris Baras, M.D. 37:01

So, you know, we were able to, as I mentioned, have a much better understanding of how many kids uh are born every year with those types of mutations. All those collaborations that I mentioned, um, the networks we have, working with health systems, actively enrolling patients and sequencing genomes and feeding back information uh to health systems, um that also helped to rapidly connect with these groups and identify patients and get those trials, as I mentioned, to happen so quickly. The other type of genetics that we do is going back to what I mentioned, the protective genetic factors. And that's just the opposite, right? So instead of finding a mutation that causes a disease, we're now finding these genetic variants that somehow are protecting patients from ever getting diseases. And to me, honestly, those are some of the most exciting things. I'm incredibly passionate about it. I already told you I think it's the greatest scientific exploration of our time. And one example to make it concrete, you can look at a gene that's pronounced side B. And we discovered that thing here. We're very proud here. There was very little known about it before that discovery. And what it was was studying large databases of people that had liver disease. And we found these loss of function. So we know that these mutations actually break the gene, they turn it off. And people have very low amounts of fat in their liver. That's a good thing, right? Fat in the liver can be a very bad thing. It leads to inflammation, fibrosis, ultimately liver failure. And with the amazing job that our industry, very proud of biopharma, uh, what was done with kind of the infectious, you know, hep C, hep B, those kind of things, um, that has now been well, much better treated and is not the leading cause of liver transplants. But fatty liver disease, going along with kind of the obesity um problem, has become the leading cause of liver transplants and liver cancers and those kind of things. So it's a really hot field. It's been a graveyard of drug development, but finally breaking through recently with Madrigal and others getting approvals in this space. So, what we can do now at Regeneron to lead you through that story, we get a starting point, a target like that. It has a huge, huge effect size, right? The largest effect size we've seen in terms of reducing liver fat and reducing the risk of liver disease of any genetics, of any gene of any target. So we move like lightning. And the next question we have to ask is how do we make a therapeutic? What's the target hypothesis? What's the what's the molecule we're going to invent and create? And the genetics are really helpful there because it, you know, I just summarized that the gene is broken. It's turned off. That's what's the good benefit. So we immediately know we have to make an inhibitor. Then we can start to look to the biology, to the other data we have. Where is this target? Where is it expressed? Is it everywhere? Is it just in the liver? Is it inside the cell? Do I need some sort of medicine that gets inside the cell, or is it pumped out of the cell and I can use an antibody? The general loves antibodies, right? Fortunately, we're able to get quick answers to all those questions, right? We had other omics you mentioned, transcriptomics. We have huge amounts of liver transcriptomic data and around other tissues. We're able to, the quick answers are side B needs to be turned off. It's mostly expressed in the liver and it stays in the liver cells. So we got to get a medicine that gets to the liver and inside the liver cell. Can't do it with an antibody. So we leveraged our wonderful partnership with L Nylum. I can't imagine a better technology to do exactly what the mission kind of required there. And we've got an sRNA in late stage development that's you know designed to silence and turn off uh side B. And we're excited, we hope later this year, uh, to be able to, you know, update the world about, you know, when we get data and reporting out on how that's doing in patients with liver disease. So that's kind of how that works. And then obviously the final chapter would be if we've got things that are working really well, we would run kind of final registrational studies for approval and take those data packages uh to the regulatory authorities.

Ben Comer 41:19

That's great. Thank you for that, uh, Aris.

How AI Helps Find Signals

Ben Comer 41:22

Um I think anyone that's hearing this conversation about this massive amount of sequencing data in 2026 is probably going to wonder how you're using AI to speed and surface uh new discoveries. Or are you I'm sure you probably are. How are you using machine learning? Are you using agentic AI? Um, how do you deploy it on this massive data set that you have?

Aris Baras, M.D. 41:49

We do use AI. Um, everyone is using it now. We were early adopters of some of the leading technologies, certainly you know, clawed uh models. Uh our chief data officer, a shout out to Jeffrey Reed, um, who just retired here after you know really marvelous career at Regeneron, um, really pushed uh those technologies here. And there are some basic things that we do. Uh so we glossed over the fact that we're sequencing massive amounts of data, right? I said billions of data points per genome multiplied by millions of people. And you got to understand which ones of those are benign, they don't do anything, they're meaningless changes in the genome versus which ones are really important, right? They could they could be causing disease or they could be this protective factor, right? That's not an easy job, right? And so you've got to learn from a lot of historic data on seeing hundreds of millions of variants about whether it's an important change or not. So we use AI, for example, to train and learn and help us be more accurate when we say, hey, this variant is important. It's called variant calling. Or we look at a ton of imaging data, right? Um, no one wants to sit there when we've got hundreds of thousands of images and sift through each image and circle the liver or circle parts of the brain. You know, we've developed automated processes and AI to do that. So that's kind of the simple answer, the more obvious things to do. But what I think is really big about how we're using AI, it's my opinion. I'm actually excited to see some other people, you know, I think maybe they had these uh perspectives before mine. Um, but people like Larry Ellison and Oracle who are looking at how they're going to be involved in this. Um, it is true that it's phenomenal what these companies have done in developing these AI models. And they're only getting better, right? And more efficient and costly. But they do tend to start to look the same, right, when they're trained on public data. So Anthropic and OpenAI, they're largely training on a lot of the same data. So they can differentiate. On their models. But what about if you could pair those models, or even better, train them, right? On the type of data that we're generating here, right? I mean, it it's taken I mean it's well well documented every week you you read, you know, the the New York Times or Wall Street Journals, just how many billions those companies are putting into, right, to develop their AI data centers and train their models on public data. And that's hard enough now to think about the billions we've spent and the decade we've spent building these proprietary data sets that are more specific and unique to healthcare. So it's not just about building larger data sets here, right? But imagine building, you said it, hundreds of millions of health records for decades, you know, people's health history, right? And now all of a sudden you pair that with such depth of data around the genome, the proteome, and all that kind of stuff. And what's actually possible is you can finally start to train models to understand human biology and predict it for purposes of drug discovery and development, right? Or for purposes of our own health care, right? Hey, let's upload our health record and let's get some assistance, right, for the providers and docs or for ourselves as patients and consumers take a little more knowledge and understanding of our own health care and wellness moving forward. So um that's one of our dreams here as well, right? You know, respecting and appreciating the power of AI, but also that we may have a big part to play in that story as well by bringing incredibly valuable health data and information to train on and have our version of AI moving forward.

Free Gene Therapy And Rare Disease Math

Ben Comer 45:45

I want to ask you a business question, Aris, and you you know, you alluded earlier to this like important question of how many patients are, you know, if it's one, if it's five, it's 10, 100. Um, the the gene therapy that you guys got approved um in in April, it's an ultra-rare condition. You're giving it away to patients uh for free in the U.S. uh that qualify. Uh what how do your shareholders feel about that? Are they good with that? A little bit, right? Um That wasn't actually my question. I really wanted to ask you, you know, there's a lot of excitement, I think, of renewed excitement right now on the prospect of personalized therapies. There's the baby KJ story, FDA has this new plausible mechanism pathway for ultra-rare diseases. My question is, is that a space that Regeneron wants to work in, potentially designing therapies for extremely rare diseases or even individual patients? Is there a business model for that?

Aris Baras, M.D. 46:47

Yeah, you know, it's it's evolving, right? It's you nailed it, it's tough. Um, it certainly historically hasn't been the most successful business model. That's okay. You know, there's a lot of failure in biotech. There have been some uh gene therapies that have been very successful, right? Not only in terms of the benefit to patients, but also um, you know, from a business standpoint for those companies. Um we need to have successful business models, or else these things won't be viable, right? That's right. We won't be developing really important medicines for patients. So um I'm excited to see that there's a lot of creativity and interest and activity in this space. Regeneron is all about genetics, is all about patients. Patients are our North Star. And you know, we do a lot um for the good of patients, and we aren't obsessed with just is every single opportunity, you know, a viable uh business model. So something so rare, like 50 kids uh born a year in America, um, it's hard to imagine how that would be a huge business success, right? Um we're a big company with a big portfolio, and at the end of the day, we're looking to make an impact on a lot of patients, right? Including those 50 a year, but you know, I would hope tens of millions of patients, right? Um So the way we think about this is I think quite important and quite right. We will always, if we can, try to develop those types of like the gene therapy for hearing loss. And we've had some other examples of equally rare, but more importantly, impactful, right, in terms of that disease and those patients. Usually there's always an opportunity for that development to help even more patients. So in hearing loss, this could be just maybe the tip of the iceberg there. There's a lot of other genes that also cause hearing loss, right? And we've now worked this out once, right? So I hope we can work it out again. And there are also other causes of hearing loss that are way more prevalent. We all know that, right, in terms of the elderly, for example. Um, so that's the hope with a lot of these, right? That we can help a small group of patients right now. Maybe there is a viable business model or not. But when we pursue these things, there's always the possibility, right, that there's a much bigger patient impact, much larger patient populations that could benefit by the technology platforms or the next or the next medicine that we might be able to develop because we did that first one, right? I'll be very brief on this one, though. I wanted to thank you for calling out, you know, what the company has done here. A huge kudos to our leadership, to Len Schleifer, and George Ankoppoulos and our board and everyone for doing something, I think, pretty profound like that. And I also want to bring this back to a mentor that I um mentioned before, Roy Vagelos. Um, and let's not forget what he did in terms of the Ivermectin donation program, uh, well documented and in his book written about how scientists at Merck came up with this amazing cure for that parasitic disease. And there was a decision that Roy had to make. Do we invest in this thing when there's no business model? At that point, they knew there was no viable financial prospects for Merck. And they did because scientists were so passionate about it and they had something that really worked. We're tying back all our themes here, right? And most of what they work on doesn't work. So when they get something that has such a profound effect, they want to see it get to patients. And there were so many patients in Africa that benefited from this devastating river blindness that happens from this. And I think the story ends with Merck donated something like over $2 billion over many, many years to continue donating and supporting that program. So Roy's the man, um, huge kudos to him as well. And amazing that his legacy survives at Regeneron. And we try to do something similar like that. Different story, different context, but but you appreciate the similar sentiment and patient first um intent there.

The Two Year And Ten Year Vision

Ben Comer 51:01

Yeah, thank you for sharing that, Aris. Uh, your your excitement about this field and the work that you're doing is infectious. Uh, we're already out of time, I'm surprised to say, but I I did want to ask you one final question, which is what what excites you the most uh about, I guess, what is possible right now and then what might be possible in let's say I was gonna say 10 years, but things are moving so quickly that's kind of impossible. So maybe I'll say two years.

Aris Baras, M.D. 51:29

Yeah, I'm so excited about this stuff. You know, I'm very passionate about what we do. I love it. Um I love working on some of the biggest problems in medicine, in biotech, and I love working with great people who want to tackle this stuff. And it's really an extension, Ben, of the things we've been doing. I cannot wait for the next two years, you know, and the next 10 years. And I think that what we're going to be able to do with these protective genetic factors, you've only seen the first few of them, right? But I'm telling you, and we've told people, we have found them everywhere in neurodegeneration. We found them in cancer, we found them in things like osteoarthritis and our joints falling apart, right? So I'm excited about what that means for the future, where we have not been able to make meaningful change in terms of moving the trajectory of Alzheimer's and Parkinson's and things like that. I'm hopeful that we finally have the targets that we can change the paradigm of what those patients experience, right? Intervene early enough with medicines that mimic protective genetic factors and completely change the face of those diseases or even something like cancer, as I kind of dared to dream on that. And I'm also really excited. It's gonna take a longer time. It's not two years, so I'm gonna have to um ask to use your 10 years that you offered me. But it's that point I made about forget drug discovery and development. And let's realize how big healthcare data and deep healthcare data, all the omics that you're so proficient at pronouncing perfectly, from genetics to proteome to transcriptomics to whatever else, um, and NAI, and how you and I and everyone else, and our kids and generations are gonna be able to go in and experience medicine in the following way. They're gonna be able to go in and cheaply and quickly get their genome and proteome and their health record and be able to know what the few things are that might be major health risks for them in 10 or 20 years from now, and to get on it early, right? And to not make it a big deal, to catch it early and to fix it early, or dramatically bend its curve and not make it a devastating, you know, life-changing, if not life-ending type of condition. So that's what I'm really excited about at Regeneron and for the world of biotech.

Ben Comer 53:54

I think that's a great place to leave it. Uh, thank you so much for coming on the show, Aris. I really enjoyed speaking with you.

Aris Baras, M.D. 54:00

Yeah, pleasure's all mine. You have a wonderful podcast. It's always fun talking with you. Thanks so much.

Ben Comer 54:06

Thank you. We've been speaking with Aris Baras, MD, Senior Vice President, head of Regeneron Genetics Center, and co-head of Regeneron Genetics Medicine. I'm Ben Comer, and you've just listened to the Business of Biotech. Find us and subscribe anywhere you listen to podcasts, and be sure to check out our weekly video casts of these conversations every Monday under the Business of Biotech tab at life scienceleader.com. We'll see you next week, and thanks as always for listening.