Pilot program seeks to connect more infants and families with screening for rare conditions
Wendy Chung, MD, PhD, a pediatric geneticist and Chair of Pediatrics at Boston Children’s Hospital, talks about the impact of the GUARDIAN study and the importance of expanding newborn screening for rare conditions.
Question:
Can you discuss the GUARDIAN Study?
Wendy K. Chung, MD, PhD:
GUARDIAN actually stands for something; it stands for Genomic Uniform Screening Against Rare Diseases in All Newborns. We came up with that title with the hope that it really encapsulated what GUARDIAN is. I started, and maybe I’m embarrassed to say this, but I started over 35 years ago thinking about newborn screening when I started working on phenylketonuria. That planted the seed a long time ago as technology evolved and as we had the ability to really understand the human genome.
Then, for me, newborn screening, just to make sure everyone, to level-set, newborn screening is a public health program; I would argue one of the most successful public health programs, where every baby after birth gets their heel stuck to take a few drops of blood to screen for conditions like phenylketonuria that are treatable conditions. To me, it’s a wonderful system and I will emphasize it’s a system that’s in place where no child gets left behind. Just every baby automatically goes through that process. Within that, there’s no difference by what your insurance is, by whether you’re male, female, by the color of your skin. It just happens the same way for everyone. As a practicing geneticist, I’ve come to realize that that’s not true of anything else in healthcare that I know of. There’s something that’s really beautifully simple about that.
Fast-forward to 2016, I had been working with many others but on a condition called spinal muscular atrophy or SMA. I’m very proud to say that that used to be, past tense, the most common genetic cause of death for children less than 2 years of age. But in 2016, it actually started in 2015, I could see that there were going to be clinical trials coming down the pipe that might be able to change the course of that disease which otherwise for the most severely affected children, the unfortunately trajectory was dying by their first birthday. It was a very rapidly progressive and fatal disease.
With that, I understood enough about the biology of the condition to know that you would have to ideally get in and start treatment before symptoms occurred; that if you waited until symptoms occurred, it was too late, you were missing the window of opportunity. I knew enough about the genetics and I’d been in New York state at the time and had friends across the state who had taught me about the public health system. We decided to do a pilot study of newborn screening for SMA to be able to identify those babies who would likely have the disease and then offer, not require, but offer a clinical trial opportunity for them, which could be life-saving.
That for me was an extremely successful and very intentionally timed time to do the pilot study, because it was yoked literally with that opportunity for those families identified to enter their trial into we now know what was a life-saving clinical trial. There was a very rapid exchange then, where that medication got FDA-approved. Then, our pilot study from New York state was used for what’s called the Recommended Universal Screening Panel Nomination. Very quickly that was added nationally to newborn screening around the United States. Now it’s actually in many places around the world. Now due to the fact that we could identify babies, it propelled not just 1 treatment, but there are now 3 FDA-approved treatments for SMA, and I can now use the past tense in terms of that use to be the most common cause of death.
Within that, that was I think a really remarkable thing for the next generation of children not to have to live through that in the same way. But I realized that we had some success and went from that to do then another study of newborn screening for Duchenne muscular dystrophy. By the time I finished that one, which sort of spanned the COVID era and I think got me tired and everyone in society tired, but with that, I started getting impatient as well. I started getting more gray hair and started getting more impatient and said, “Doing these one by one takes a lot of effort, a lot of energy, a lot of money, a lot of resources. Isn’t there some way that we could do this in a more sort of expeditious way for many rare diseases and not have to do them one by one?”
GUARDIAN was actually planned by families who were part of originally our SMA newborn screen that had normal results. They were families though that just had been through the process and were thoughtful and willing to tolerate me asking them a lot of questions. Within this, they designed the GUARDIAN study, which is the following. You take the newborn screening dried blood spot, we use that to be able to extract DNA, but we don’t take any additional samples from the baby. We use the same dried blood spot with the parent’s permission, it’s consent for research study, and we do whole-genome sequencing. We have all the genomic information.
But we don’t look at it all. What we do is we constrict ourselves to what I call news you can use, so conditions that are treatable, conditions that we can diagnose by DNA testing that are treatable conditions. We read out that information and we let those parents know for those if they are screen positive. This is a screening test, so just like a mammogram is a screening test, it’s not a definitive diagnosis, but there’s a higher probability. We need to bring those babies back to know for sure do they have the condition or not.
GUARDIAN is not just the screen, but it’s also the confirmation of the diagnosis and it’s the onboarding to care. It is really trying to take families and babies through that whole process to be able to get them set up for health going forward, but with some of them knowing that there may be something that they’re at risk for that they need to be aware of to avoid those risks ahead of time.
In doing that, we didn’t know if it was going to work. We didn’t know if we would be able to get DNA, if we’d be able to do sequencing, how quickly we’d be able to do it. But we started out doing it for about 250 conditions. Just remember what I said before, not just 1 condition at a time, but the first version of this, we did 250 conditions. We’ve actually subsequently, once we knew we could handle that, the beauty of doing whole-genome sequencing is that you can then unmask or be able to read out information from additional genes literally with the flip of a switch.
On what I call version 2.0 of the test, we added about 200 more conditions, almost doubling what we could do in terms of capacity of information, but still focusing on news you can use. Within this, just to be clear, these were conditions affecting children less than 5 years of age. This isn’t like Alzheimer’s disease when you’re in your 50s or 60s. This is like what a young child needs to know. By the guidance of our families with a very high probability, which we operationalized to 90% probability that those children would have symptoms. We didn’t do this to try and do a 23andMe test, that’s not what this is at all. It’s very focused on rare genetic conditions, but very high probability with something we can do about it if we only know; but otherwise, without doing these tests, very likely that we wouldn’t know or we wouldn’t know until the disease was already symptomatic and perhaps missing a window of maximal benefit for those children.
That’s basically it in a nutshell. In doing it, I was in New York City at the time, so I did this in New York City and had the wonderful opportunity of having a very diverse community because New York City is a melting pot and diverse in every dimension you can think about. We were really able to test whether this was something you could do in a sort of very diverse community and be able to understand who would adopt, who would accept the invitation versus who might not. That was a very good experiment to be able to do in that community.
Question:
Did you notice any trends in those who were more or less likely to consent to the expanded screening?
Wendy K. Chung, MD, PhD:
That’s a great question. We wanted to know that from the beginning. We were very intentional to not only understand who joined but actually understand and ask very humbly, but ask why they wouldn’t join or why they didn’t join if they didn’t. One of the things was this was such a new experience that we used the opportunity to really talk to individuals that we approached. We approached them in the hospital after birth, and we had literally research staff that were from our communities and understood what it was like to be in those communities. As a result of that, it was very comfortable for them to talk with the new parents. They actually had very insightful conversations both for parents who joined as well as parents who joined.
To answer your question, overall, depending on exactly when it was that we measured this in the JAMA paper that you cited, 72% of parents that we offered to do this ultimately decided that they wanted to do the study. We’ve since gone on to go beyond the 4,000 babies that were in the JAMA paper, and we now have screened over 14,000 babies. The uptake rate continues to be about 72% to 74%, so it’s still very consistent within that range.
Within this, we asked individuals how they self-identify in terms of their ancestry, so where in the world they came from. Again, because it’s New York City, it was majority minority, so about 31% self-identified as of European ancestry. But when you look, that means the majority of individuals self-identifying as Hispanic, Black, Asian-American, Middle Eastern, and many people that were actually more than one, and so a very wonderful sort of slice of the entire world, sampling of the entire world.
Interestingly, and we did this in terms of languages, we had study staff speaking English, Spanish, as well as Chinese. With that, admittedly we didn’t get every single language from around the world, but we did get languages that were where our communities, where those hospitals, those birth hospitals served in New York City. We very carefully looked at who signed up to do the study and who did not. At least by the self-identified ancestry, it was exactly the same.
It didn’t appear as if there was a preference for just white people on the Upper East Side to do this, or individuals just from Brooklyn who self-identified as being Black. We actually had very reflective, and when we asked people why they didn’t join, it was interesting that it wasn’t anything particular about genetics for the vast majority. For most people, the reasons were, “I don’t want to do research.” I will have to say, I respect that because there are many people who don’t want to do research. If the uptake rate had been 100% or 98%, I would’ve been suspicious that we were being coercive in some way. I actually think the number that we got sort of makes sense to me based on, like I said, who wants to do research.
Many people didn’t want to do research. Many people felt that the current of care of newborn screening was a very good thing and that was sufficient. That’s their decision. That’s perfectly fine.
There were a few people that just felt like, “You know what? I just had a new baby. I’m a little bit overwhelmed.” I get that as well, right? Within that, I think we started explaining this to our communities more passively in the third trimester of the pregnancy so that they had some time hopefully to think about it. But I get it. I’ve had babies and it’s a thing, it’s a big thing. We do give people up to 30 days after birth so that if they want a little more time to think about it, but your life turns upside down once you’ve got a little one.
All of those reason made sense. A small minority of people said they were concerned about genetic-specific issues. I’m not surprised that some people were. I would’ve been surprised if we didn’t get some of those responses. But it was, well less than 1% of people who declined that had those types of specific concerns. I felt like we did a good job because we are very careful about protecting people’s privacy. We were very much in line with what the community asked to do in terms of how to design the study and what to report out that people had a lot of faith and trust in us and that we did a good job in terms of listening to them.
Question:
Do you think this pilot study might help eventually expand the standardized newborn screening that occurs?
Wendy K. Chung, MD, PhD:
Certainly, we designed the study with the intention that this would provide the pilot data necessary to understand would this be, I’ll call it, an addition to the standard newborn screening. Clearly right now, I don’t think it should be a replacement for standard newborn screening, but I do think that one of the things we learned is that the standard newborn screening is not perfect. I’ll say that’s almost true of any test, any medical tests.
We did identify children in GUARDIAN, I should have said. I specifically included conditions that are in the standard newborn screen. We blinded ourselves to those results when we were reading out the DNA sequence, but it added a very good check and balance to see whether or not we were doing our job correctly. Again, I didn’t know the results, but before I called the family with the results, I would double check and see whether or not they were concordant with the standard newborn screening results so as not to ever confuse a family.
What was interesting is that within the sample that we reported in JAMA, we actually had not many, but a few cases where we detected someone by the sequencing-based method that was missed by the standard newborn screening that did not include the DNA sequencing. The standard test is not perfect. In fact, one of the results was so important that standard newborn screening, and again, I’m not blaming anyone, nothing is perfect in this, but a baby boy was missed who had severe combined immunodeficiency. He had a defect in his immune system. In that particular condition they attempt to pick up by newborn screening, but he was missed. For technical reasons I could go into, it’s not that anyone did their job wrong, the test is just not a perfect screening test.
When we picked him up, we realized that there were 2 other patients that had been reported previously with his exact mutation. One of them had died and one of them had lived, but only because they had a bone marrow transplant. We actually arranged to, number 1, check his immune system, and his immune system in fact was not working correctly. We literally kept him very safe from other infections until the time where we were able to do a bone marrow transplant within the first 100 days of life. That’s kind of a magic number, because children do much better if they get transplanted early.
He had a successful bone marrow transplant. He’s been disease-free. He has not had an infection that he could have died from because he didn’t have an immune system that was working correctly. He now, with a bone marrow transplant, has someone else’s immune system that is working to protect him. My hope is that he is now on a very different trajectory to health, but because he got picked up by this sort of additional test for regular newborn screening.
I won’t say it was easy for him or his family, even though it was news you could use and a treatable condition. It’s not like it was boom, just like a very easy treatment, but it was a life-saving treatment for him and otherwise something that I don’t think we would’ve known about otherwise.
There were other conditions where in many cases the intervention, the treatment I think is probably going to be as impactful, but the interventions thankfully are much easier. Another condition that we frequently saw is a heart condition that can cause a fatal arrhythmia, something called long QT syndrome. Based on the data that we’re seeing, about 1 in 1,000 babies look like they’re born with this condition. It’s not picked up at all by regular newborn screening now, so this was a complete addition for something that we couldn’t screen before.
The DNA sequencing proved to be extremely accurate. That is that when we did the double check of doing an EKG on the baby to look at the electrical system of the heart, when we saw it on DNA, we saw it on the heart. It was a very good concordance that we weren’t making mistakes in terms of doing that. In this case, the intervention is very simple, at least to start with, which is that you avoid certain medications that trigger the problem. There’s certain medications, there’s a list that we know of, that bring it out and cause the arrhythmia. It’s very simple. You simply avoid those medications for the child and that can potentially be lifesaving. I do know personally even of babies who’ve died of sudden infant death syndrome that were because of this long QT syndrome. It’s the type of thing that hopefully is life-saving for the child.
But interestingly, this condition also runs in families. By identifying the baby, in several cases we’ve now also identified parents or even other family members who were benefiting from this information as well, some of whom had symptoms but didn’t even know why they had the symptoms. This potentially is something that’s helpful for several families in addition to the newborn who is obviously our primary focus, but we’re happy to help the rest of the family as well.
Question:
Can you talk about the technology used in this study?
Wendy K. Chung, MD, PhD:
GeneDx was our sequencing partner or our laboratory that did the actual sequencing. The process, as I alluded to, is we took some residual or leftover dried blood spots from the newborn screening. We extracted the DNA, and then we did what’s called genome sequencing. That incorporates information, DNA readout, on about 20,000 genes. Again, we didn’t look at all 20,000, but that’s reading out the A’s, T’s, C’s, and G’s from all of those genes. Then, we use essentially machines to help us, so computers to help us filter through tons of information to get to the small amount of information that we need to double check. In every baby there were on average 6, 7, or 8 very specific spots that we then had a human go in and double check and look out and make a call about whether that was just a normal variation that we didn’t need to worry about, or whether in fact it was something that could be disease associated.
In the first JAMA paper that you mentioned, we ended up with about 3.5% of babies who ended up with a screen positive, that is there was something that we called out to mom and dad to let them know that we saw something. Then of those that were screen positive, the majority of them were true positives. That is, at the end of the day, once we did the double check, in fact, we found that those were things that baby needed to know for the parents and the doctors to be able to take the best care of them. With that, it was very good to prove to ourselves that, number 1, we could do this, we could do it accurately, and that this was information that families ultimately wanted.
We did also learn, I’ll be fully transparent, we also learned we don’t know everything. I didn’t expect we would. But we learned from our community very quickly that there were certain genetic variants that were predicted to be disease causing, but in those communities because we hadn’t seen those variants before, because in large part, those communities, I have to say, I think have been left out of a lot of genetic research. We hadn’t seen those variants before, and so we interpreted them incorrectly. But now we know, and we very quickly learned, and we very quickly quit reporting those out as being disease associated. Now those communities are benefiting from that because they won’t be getting false positives, not just from our test, but in fact across all of genetics they won’t get false information from genetic testing going forward. I think this is going to prove very good actually across communities, not just for newborns, but in fact for all genetic diagnostic testing to help make it better.
Question:
How might this study impact the future of rare disease medicine and management?
Wendy K. Chung, MD, PhD:
A couple things. I think number 1, we’re learning that when we screen a population at a population level without ascertainment of who’s coming to us, who finds their way and navigates through the healthcare system, when we just do every baby, I think we’re going to find that many of those rare diseases are not as rare as we thought they were. My prediction is that many of them will be two to 10 times more frequent than we thought. I think that’s going to be very important, number 1, to make sure that we don’t leave anyone behind, especially for these conditions that have treatments. I also think that we’re going to have, I hope this was my original intention, to be able to have a very flexible, agile system that can be responsive so that as new treatments become available, we can quickly add them to the screen by simply unmasking the new information.
With this, my goal was really to be able to catalyze rare disease diagnostics and therapeutics synergistically so that we don’t lose a generation. I was and I am concerned that that gap in terms of treatment and diagnosis, we have to be able to reciprocally go back very, very rapidly, not in the matter of decades, but really in the matter of days or months at the most as we think about this.
I hope this is paradigm shifting. That really was my intention in doing this, that we’re not perfect as we do this, but that we will get to be better and that as we’re doing this, we’ll be more equitable. I do think we’re going to have to keep doing a lot more of this. Even though the JAMA paper describes 4,000, even though we’re at 14,000 now, I originally thought that we would need to do about 100,000 babies, because to be recommended for the pilot data for the recommended universal screening panel, at least right now, you need to identify at least 1 baby with the condition to be nominated for that disease to be nominated.
If this, in fact, were to be something that fits into the public health newborn screening program for conditions that might be 1 in 50,000, we need to make sure that we go sort of screen more than 50,000 to be able to identify that 1 baby for a successful nomination. My hope is that this is a platform, an enabler that actually helps all rare genetic diseases. I hope that as we’re doing this, this is something that, although we started with 250, now we’re at 450; my fantasy is that sometime we’re in the thousands, and that as we’re doing it, we’re doing it because we’ve got treatments available for those. Then, my secret fantasy is that someday I’d like to retire. It’s not going to be for a long time, but I want to make sure the rare disease community is safe and secure so that I can put myself out of business.
