EP 185: Newborn screening to lifelong data in evolving genomics landscape | Madhuri Hegde of Revvity

Thank you, Patrick.

Sure. So Revvity was formally announced in May of 2023. And Revvity really stands for Rev as in revolution and Vita as in life. So that's the double V you see in the middle of our name. And Revvity really was, as you mentioned, the rebranded version of our former company where we divested our food and applied markets business and then went on to become purely a diagnostics and life sciences company, very focused on healthcare. And it's been an exciting change for us because really it helped us focus on what we really have been doing for many years and then now take that further.

Sure. So you know, as you mentioned, I truly kind of came from academia to industry and I feel like I was at the right place at the right time because I got to see both sides of the world. Largely because the technology has changed so much. I mean since I sort of finished my PhD, went on to do my postdoc, and we were just in the world of Sanger sequencing for a very long time. Now we all talk about sequencing a lot, but all the ancillary methods around sequencing, I must say right at the start, are still very important. They are still helping us figure out a lot of basic questions in science. So it is not just sequencing, but sequencing really transformed right from going from Sanger sequencing to next-generation sequencing and what we were able to do in a clinical setting and even research. A lot of my research has been in neuromuscular conditions and gene discovery was really facilitated by next-generation sequencing. You see that steep rise from about 2006 to 2010 where a large number of gene discoveries started happening and largely facilitated by next-generation sequencing. So going on from there, we started with the short-read sequencing and we have all seen that evolve in the last 10 years both in research and in clinical, the different applications of it. What has happened in the last I think five or six years is now we are starting to see this sort of availability of short-read but some different chemistries now which are at our fingertips that we are not depending on one single chemistry anymore. There are different chemistries we can bring to our labs now. That's one thing. And then the evolution of long-read sequencing. Right now we are still not there for long-read sequencing making its way from research to clinical. And clinical is a little tough because we have to do a lot of validation even though they are laboratory-developed tests. I must say that there's a significant amount of analytical validation that is involved before you can bring a test to the market and offer it to individuals. And also you got to get paid for the work you do. So that is still evolving. But I think right now where we are is short-read and long-read kind of starting to run parallel with each other and let's see how this evolves. So that's with just the technology side of it. But again, as we all know, questions still remain. There are some certain conditions like let's take the FSHD or some of the other methylation disorders which really need some different methods to address it. And that's where optical genome mapping methods like that have started becoming very important. And that's why I said it's not just sequencing. We are starting to use a whole gamut of different methods in the clinical labs and even in research. And the last but not the least I want to mention is mass spec. You know, as sequencing has sort of evolved over the number of years, I think mass spec has also come a long way. I'm a molecular geneticist by training but it's just been an amazing experience for me in the last five or six years of my exposure to mass spec and how we can actually sort of start marrying these two data sets together. So that's where the field has gone in the last decade or so. Talking about just Revvity and what we are doing at Revvity. So at Emory I used to run the Emory genetics lab and at Emory genetics we used to get samples from all over the world. Coming here, I'm running labs in different countries now. So we have clinical labs in India, China, Sweden, UK and US. So now you can just turn it on its head, right? Instead of getting samples here, we now have local labs. So it's really an amazing experience. So Revvity probably is the only company which has its own unique clinical laboratory network under the umbrella of the parent body of Revvity, which is truly a manufacturer of diagnostics and life sciences reagents and tools. But Revvity has been the global leader for newborn screening. We manufacture IVD kits and sell in about 110 countries.

It's, you know, newborn heel prick blood spot type testing as well. Absolutely. So just taking the dried blood spot card, you know, every baby in the US, 4 million births in the US, they get the heel prick. If you go to the hospital, and my daughter recently had a baby, the nurses will say it's a PKU test, right? It's only starting to change now that they say newborn screening, but the PKU test or the heel prick has been done for a while. And there is this recommended universal screening panel which is done in the US states and is actually referred globally as well. Now every state does it depending on its budget. Some states might do 10 and California does more than 50. So it really depends on the state and their own budgets. But it is so critical, right, that you save lives at birth if you detect early. And the concept is very clear that early detection, early intervention. And I think congenital hypothyroidism is a great example for it, right? If that thyroid pill is given to the baby as soon as the baby is identified, you're talking of literally a complete change in scenario where the baby may not survive to actually having and leading a normal life. So it's a huge deal, right? But think of countries like India where there are like 26 million births compared to 4 million births in the US and newborn screening is still not mandatory there. So we have still a long way to go globally when it comes to newborn screening. But personally for me, I think coming here and sort of learning how newborn screening system actually works. And when I came from Emory to Revvity, this was my opportunity really because at around that time we were starting to move from gene panels, exomes were starting to become routine and people were just dabbling in whole genome sequencing. So I used that opportunity to kind of start whole genome sequencing here. But I wanted to do it a little bit different, right? Because here now I have that infrastructure, if you can say that, where Revvity has a channel called Via where we do cord blood and tissue storage. So in the second or the third trimester, the parents are introduced to cord blood and tissue. At that time if we talk to them about whole genome sequencing, then we are really talking of implementation science here. We see a lot of projects happening now for newborn babies. But if we really want to make it something which is more universal, introducing the parents to that concept very early on is important. And we published that paper in 2023 in JAMA of our early experience in launching this whole genome sequencing like that. And then we took it a little further which was clinical whole genome sequencing. So really for sick individuals, babies, pediatric population and adults. And also then doing the rapid NICU which has also been an amazing experience because if you can turn around those results that quickly, and we are doing it in 48 to 55 hours, you really can bring about change in how the baby is being treated. But that's with whole genome sequencing. I think what has equally excited me, and I said this earlier, is the mass spec aspect of it. We are starting to see that untargeted mass spec aspect coming to life now. A lot of newborn screening products are mass spec based. You're looking at one analyte at a time, right? This concept of untargeted mass spec, eventually I think as we go forward in the next decade or so, you're going to see sequencing evolve significantly but alongside I think mass spec is also going to come up at the same rate.

You're really looking, you know, think of it as, no, whole genome sequencing, I think the closest analogy I can look at is doing over whole genome sequencing. You're looking at almost a similar way of doing mass spec, looking at analytes and you're identifying new analytes or identifying those peaks which might give you a clue into the disorder which might be still undiagnosed. And there are some really good papers that are getting published now.

So I think just at a very fundamental level globally, I think what newborn screening, which has been around for now more than 60 years, has taught us is that that Guthrie card, you know, that we have really sort of got a sample type at our hand that we can actually use both for biochemical analysis, which is the current screening approach, and for DNA analysis or the molecular approach. So now we have, and also it is a sample type which is portable and by that I mean it doesn't need to be frozen. It can be literally kept forever. So at Revvity we're really excited about this because we do whole genome sequencing on babies from DBS cards and we've really honed this technique down. The sort of the expertise that comes within having labs and having the expertise that comes with these kits, who are the people who are putting together these kits, it's amazing collaboration, right? So now you take that to sort of the next level of the DBS card as your starting point. I think newborn sequencing definitely is kind of going to be the next and very important step but we all have to think through this carefully and I challenge this on probably every second, right? We have a lot of very interesting projects going on which are absolutely critical and important. I mean we just, our lab in UK will be doing sequencing for the Generation Study project for Genomics England. We are doing sequencing for the RTI Early Check project in North Carolina. There are some other interesting projects going on in the country. These are the whole genome sequencing projects which are absolutely critical to move this field forward. So two or three ways to look at it, right? One is the knowledge that we are going to gain from these whole genome sequencing projects but we are also going to understand which genes are critical to look at. And the fundamental problem of whole genome sequencing is how cheap can it become. Now we all have to think of this carefully. Today whole newborn screening, it is a public health program and it is not just the test, you know, there is a whole system around it. There is a quality component around it. There have to be nurses who are going to do the sample collection. There has to be a follow-up. I think what we are not doing well right now is to think of newborn sequencing in a public health setting what it would mean. And the reason I mention this is obviously cost is important but aside from just whole genome sequencing being really cheap, its data storage is not that cheap if you're storing the entire genome. So let's think of some practical solutions if we are going to deploy it globally. And I think a gene panel makes sense but then the question is how do you design a gene panel which is globally applicable. And we have been doing this exercise. We have been collecting all the gene panels. So even though all these groups are doing whole genome, they're actually looking at a subset of genes, they're not looking at the entire gene. So we've been collecting this data and we have actually released a gene panel last year. We call it the Neo-NGS project and that product is there. But we are now starting to dive deeper into really looking at the evidence, the incidence, should it be on a gene panel or not, and coming up with that baseline gene panel which is globally applicable, which will help you control the cost, deploy it globally and also manage it in some such a way that as discoveries happen, you know, let's say in India someone is doing it but the genetic condition is so common in India which is not common in the western world that you should be able to put it in there. So there's a lot of thought process that is going on. I think next 10 years we are going to see that settle. Again in that process we continue to do these whole genome sequencing projects, learn from it. There are going to be some countries who will just deploy straight to whole genome because they can afford to do it. But when you talk about making it equitable, I think a gene panel makes sense if you're going to look at a global approach.

Sure. So I think let's talk about data storage first, right? Because I've been thinking and thinking and thinking about this. And I think it is better to store the DBS card and the DNA than actually storing the data. Even though we are storing the data right now, we don't discard data, but the best thing to, because the technology is evolving so much that it's better to go back and resequence. I mean every year we have an improvement, I think every three months there is new and improved chemistries launched by these companies. So it's better to resequence than to store because that is quite an expensive exercise. Downloading, storing, archiving and downloading the genome again is not cheap, right? Yeah, I'm going in that direction. I think biobanking DNA is a better exercise than actually storing the data. But again, you have to comply with all the global, every country has their own data security policies and there's a lot with that privacy. So we have to comply with all of that and that will continue to evolve. But if you talk of sequencing price, I think probably we are going to hover around that $50 to $100 mark for a while. And I recently saw a presentation where they predicted by 2030 we will be at $10. But let's just think about it more practically that it is not just about generating the sequence, right? There is so much that gets sort of built on when you are talking about sequencing. And what I mean by that is a company might provide a $10 or $100 genome, but you still have to do the prepping the sample, the library, you know, there's labor that goes around it. A laboratory has to keep its lights on. So there are those overheads that go along with it and then interpretation. So I think we all have to get away from this mindset of $100 genome to what it means to do a complete genome. And by that I mean from sample prep to interpretation and getting it out. And that price is going to look different from just doing that what we talk about and what we hear about often is this $100 genome or $10 or whatever we want to call it. And I get very concerned with that because what happens is when we are interacting with individual parties who want to do sequencing, they're so fixated on that $100 genome that they forget that there's this whole ecosystem that is needed around it.

You know, I think interpretation of the genome is more like a science and an art to it, right? The science you can automate it as much as you want. And I think we are sort of on that track where the more and more variants that are classified are deposited into public databases such as the ClinVar where, you know, I find a variant I automatically plug in that it is pathogenic or benign and I don't have to worry about it. But in reality that's not how things happen, right? When a sample is actually sequenced and it comes to us, then a lab director is looking at a number of things beyond just looking at the known pathogenic, not pathogenic. We have to look at the data set in its entirety. And I think that's where things get forgotten because variant classification is still required and it still requires human intervention. You know, probably 10, 20 years down the road we would have sequenced so much but you are still going to have those unique variants come up. And the last but not the least and a very important piece is that many times what happens is we could detect a variant that could be disease-causing but that disease does not exist in that individual just yet or does not exist. And there are many cases like that where a previously reported pathogenic variant and we just don't see evidence of the clinical disease in the individual. The more healthy population we do, the more we are going to encounter these situations. So we still have a long way to go.

I mean I think I hope that at some point there is a policy that gets put in place where newborn sequencing programs do not need to store data, right? I think it's going to be very important that we sort of somehow get to that. And you're right, I mean right now as I mentioned to you in our whole genome healthy genome sequencing program where families are signing up in their second or third trimester, what happens is many times if they have older kids they are getting their genomes done too. And this is not direct-to-consumer at all. We are going through a counseling service. There is pre-test counseling. There is post-test counseling and the test is ordered by a physician. But many of these, if these are kids closer to 18 or 21, they can come back and ask for a reanalysis of the genome. And we have done that because where we have stored the data, we have downloaded the data and we have gone and done the reanalysis. But it is an expensive exercise for the lab to do that and I don't think we are really doing just justice to the amount of work we actually are having to put in right now with it. So that's where it stands today. But I really hope that as the technology is evolving so fast, there is some sort of a policy that comes that we actually don't have to store the data. You know, one small caveat to that which is a very important caveat to remember is that many times what happens is that people could challenge your results that possibly a lab missed something, right? And that becomes difficult because if you have dumped the data then you have no recourse to it to go and check anything. But that is a very rare occurrence but it does happen.

And you know, it is very hard to grasp this concept of penetrance and expressivity too. I attend many patient advocacy group meetings where a family of four siblings have got the same genetic pathogenic event but they are expressing the disease very differently and it's very hard for families to grasp it. Penetrance is even more difficult, right? When you're talking of, let's take the example of LRRK2, the most common Ashkenazi Jewish mutation, the penetrance is from 15 to 80%.

It's a very hard concept to explain to families because there's a lot that comes along with that, the emotional side of it. And that's why the counseling piece is so important.

I think chatbots definitely have helped where you ask a question and I had tried a couple of chatbots. They really kind of direct you in the sort of right direction telling you as you go on asking questions. But it really comes down to education. I don't think you can actually physically scale so much of genetic counseling because we all tend to forget that we are not machines and we're not robots. We need that human interaction. All of us need that. It's like having your annual checkup, you do want to see your doctor to ask questions. So I don't think you can sort of completely make it a robotic type of event. But there's a lot of education piece of it that can be done. Like what is an autosomal dominant condition, what is a recessive condition. When you are doing a prenatal, you're going to find everything. So for examples, when we do prenatal sometimes we find non-paternity. But all that education is very important right from the get-go so that when you actually get into the session, you're coming to a little bit more of a level ground where you are able to explain the results more easily. Imagine doing this on, so one another thing to remember is that let's imagine a situation where four million babies are going to get newborn screening, right? A fraction of that are actually going into these sessions. Most of them, they're not going to have any significant finding. Carrier status is at age 21 and on when they reach their reproductive age and they can actually query it again. But very little is actually moving. So we have to think about this carefully that what percent is actually moving in that direction. And last but not the least, I think we all have to somehow figure out how we are going to marry the new clinical trials and therapeutic strategies that are getting developed back into this newborn sequencing system. Remember newborn screening has conditions right now, many of them are inborn errors of metabolism which are addressed right at the bedside and physicians can take action on it. But let's take the example of Duchenne muscular dystrophy which has been, that application has gone in two or three times now, not approved yet. There is an FDA-approved kit for Duchenne, it's a CKMM kit but you have to do molecular confirmation. The reason I mention that is because there are CRISPR-Cas9 strategies and a variety of different strategies that are being developed. It's also about figuring out how do you connect these individuals and these families to the right clinical trials or the right treatment strategies. Because you cannot expect the physician at the other end or the neonatologist to have it all figured out without pre-giving them that information. And there are some amazing work done by many pharma companies right now in these type of conditions.

Yes. And you know, I'll tell you like in many cases, families are very proactive. They will push harder. I think the Lennox-Gastaut Foundation is one example. They're doing some amazing work, right? Where these families are pushing that envelope a little bit further so that these treatments are getting developed at the other end of the discovery work. And we interact a lot with the pharma industry right now in it's just very early discovery work, the drug discovery work I'm talking about, where biomarkers are getting identified. These biomarkers eventually can become companion diagnostic type assays when these strategies enter the clinical pipeline because you need to monitor these individuals over probably their entire lifetime. So again we're going back to that basic fundamental data of sequencing or a biochemical assay which is a mass spec based assay and taking that down to the entire life cycle of this product development.

Yeah, I think OLink and a couple of other companies which are doing similar work. SomaLogic is also there. There are a couple of others now. I think it's really exciting. There was one paper that was published a couple of years ago doing the pQTL assays showing that protein-protein interactions are not typically the way we think about it. Meaning in a particular condition a couple of interactions were identified and then we kind of just lock that down that that protein interacts with that protein and this is how it is. This particular publication actually showed that that same protein could be interacting in completely different conditions which we have not even thought about and they had some really amazing data. I think this is where these new technologies are going to be very important. I think the question really is how do we translate the data that we are generating from this kind of technologies into a clinical perspective of how we are going to use that. Right now we are using it in understanding just basic discovery work or basic understanding of protein-protein interactions or number of proteins that are there in a particular sample. But how do we take that a little bit further and bring it into clinical implementation? The way I look at it, I think right now where we are is in this implementation science and how do we actually enact on it is going to be very interesting to see.

So, I'm really excited that Early Check is doing that because they are the only program which has gone a little bit beyond than looking at standard monogenic conditions. And they have spent a lot of time curating that list of SNPs they are going to look at to derive those scores. You know, type 1 diabetes is interesting, right? So when you talk of type 1 diabetes you're also talking of two types. The type 1A and type 1B. Type 1B has got more of an inheritance component in it which is a MODY type. So you're really talking of type 1A. And there is a drug on the market now which is FDA approved which can delay the onset of type 1A diabetes in kids. Which is very important. Those autoantibodies can be detected as early as 6 months and the drug is given at about age eight and it will delay the onset. It is like a 14-day infusion and it can be repeated every two years or some similar algorithm like that. But now imagine taking it a little bit step further and marrying the genetic polygenic risk scores with that autoantibody data. I think this will be the first example where we would have shown that marrying these two data sets is actually going to help at least delay the onset as much as possible. But that is just with type 1 diabetes. I think in breast cancer also the polygenic risk scores have been shown to be effective. There are some very good publications on that but it's just very early days for polygenic risk scores because we are all so spread globally and if you look at different populations and the pressures that come by being in a certain environment that change our gene expression and the exposure with that. So just having a polygenic risk score is probably not going to be enough because that risk score could change depending on where you live as well. Whether you live in Asia, Europe, the Americas, it could change. And how do you actually bring that closer together to create a clinical algorithm from it is going to be interesting. There are a lot of groups that are working on this and really some very cool work getting published right now but it is still early days.

Yeah, thank you for asking me that question. I've been here for eight years now. So it's not a couple of years and yeah, it's just been amazing. I mean, I feel I've learned so much. You know, in academia, obviously, I wrote a lot of grants. I got funded for my work and I was fortunate that I was running a clinical lab. So closer to the industry in an academic setting. So I was able to split my time coming here to see the early stages of product development and how much work is put into R&D before a product comes to the market. You know, I probably when I was a researcher or running the clinical lab only really did not have the real insight into the amount of work these companies put in. The QA aspect of it, talking to people, fine-tuning that product, that requires enormous amount of effort and investment before bringing that product to the market. I think I've just been really fortunate that I have had this opportunity to be here, right? And of course them adjusting to me too that I came from a purely academic setting, a little bit clinical which you can talk about it as a commercial laboratory. But this has been a very exciting journey for me.

Yeah, I mean I'm more than willing to talk to people and people reaching out to me to discuss some novel ideas. And nowadays innovation and how you think about innovation is, innovation is not just a new product discovery but there are a lot of different areas we can innovate in and that's how I look at it. So really willing to talk to people reaching out. It'll be really exciting to have that conversation continue.

Thank you. Thank you for having me.