CPHR Seminar Series - Nina Gold

welcome everybody uh on behalf of the center for precision health research it's my great pleasure to introduce today Dr Nina Gold who is the director of prenatal medical genetics and the director of metabolism at Massachusetts General Hospital she's an assistant professor at Harvard Medical School she received her baa suami from kby college she got her MD at Harvard medical school and she's getting a master's of biomedical informatics currently that would be my dream to be able to do that she's trained in both Pediatrics medical genetics and medical biochemical genetics um but today we've invited her because of her research interests she's using populationbased genetic screening to identify treatable genetic disease across the lifespan her most recent work focuses on identifying genetic disorders that are treatable in fetuses and that's why our section which is the section on prenatal genomics and Fetal therapy thought that Dr gold would be the absolutely perfect person to to come and speak with us um she's also examining the penetrance of genomic variant associated with childhood diseases in undiagnosed adults just a couple of housekeeping things this um Talk is supposed to be recorded although I don't see the recording sign on here Gerald told me that it is being recorded and then it will be posted one week later so that you can review it or other people will be able to see it if they couldn't come today importantly there is a Q&A box at the bottom of the screen with a question mark feel free to write your question questions um during the talk I will be reviewing them if it's something that I think would um be good to clarify for everyone then I will respectfully um uh interrupt Dr gold if not we'll wait to the end and we're going to leave around 10 minutes for questions at the end and I've just received confirmation that we are recording um the talk so with that I am not going to take up any more of your time Dr gold and I'm going to hand the virtual microphone over to you and thank you so much for coming thank you so much for inviting me Dr bian I'm so excited to be here and really honored to speak with the group today um so as Dr bian said I'll be talking today about the identification of treatable genetic disorders in fetuses and newborns and as a disclosure I was recently on an Advisory board for a company called myome so today we'll be discussing the landscape of genomic newborn screening research first then exploring the Continuum and link between newborn and Fetal genomic diagnosis and then finally considering future directions for investigating and improving upon genomic screening so first the landscape of genomic newborn screening research which has really expanded quite a bit over the last 5 to 10 years so as many people in this audience are already familiar with um newborn screening as a Public Health Initiative was instituted in the 1960s initially with PKU and over the ensuing decades many other types of disorders have been added including other inherited metabolic disorders like lact toia and maple serup urine disease hemoglobinopathies like CLE cell disease and multi-stem diseases like cystic fibrosis and then we've added other types of screening too other modalities in addition to filter paper cards like the hearing screening and screen for congenital heart diseases as well and increasingly genomic information is being integrated into newborn screening as a first tier test for disorders like SMA spinal muscular atrophy and sometimes as a second tier test in certain states to distinguish infantile from late onset phenotypes or to identify common variants in disorders like Pompei disease so the health resources and service Administration currently recommend commends that there are 37 core conditions that each state should be screening for and this is just a recommendation so it's actually implemented varies quite a bit state to state and then there are 26 secondary conditions which share biomarkers with the core conditions um but are not necessarily the specific targets of newborn screening but can be ascertained by the seam test and since the Advent of newborn screening um the conditions that have been added to the panel have really been Guided by the 10 foundational principles that were published by Wilson and youngner in 1968 and those have emphasized the inclusion of early onset and highly treatable disorders but over time we now have begun to recognize that there are actually over 800 genes that are associated with treatable conditions and so I think you know many investigators people in public health and um involved in clinical genetics have begun to ask if it's time for us to rethink the disorders that are being screened in newborns and other individuals particularly by adding genomic sequencing as another tool to expand screening so what happens when a disorder isn't screened so you know many of us who are in medical genetics or in the you know medical fields are familiar with this kind of paradigm right a baby might appear healthy at Birth and then maybe they have Miss Milestones or other symptoms they can get referred to a geneticist eventually or another specialist and then months later if they can get access to someone in their area who has that expertise they may see someone undergo testing and weeks or months after testing is sent if it's approved by insurance they may get a diagnosis and so um you know what this means is that there are sometimes people who may not present initially with symptoms until it's too late and so these are a couple stories that are sort of ripped from the headlines um for example um a young teenager in London who didn't know that he had an inherited metabolic disorder called OTC deficiency which predisposes people to high ammonia levels after protein intake so he unfortunately drank a protein Shake um developed vomiting and incopy and it was not known that a pneumonia level should have been checked and so you can imagine that had he been screened for this disorder at Birth or in early childhood it could have been potentially lifesaving um the same goes for things like cardiac arrhythmias um which may not you know necessarily be symptomatic or may have vague symptoms um but can ultimately be fatal in childhood but when we know more when we screen for things we can offer therapy before irreversible progression of a disease occurs so in these cases gene therapy um was offered to babies with imuno deficiencies um who otherwise would have had much more severe symptoms and and much worse quality of life um so of course there are many stakeholders um in the expansion of newborn screening this isn't something that can just happen over night um there's the newborn screening Labs themselves and the resources um from the government that support them pediatricians medical geneticists and other Specialists the pharmaceutical companies that make these therapies obviously rare disease communities people who have these conditions and their families parents of newborns and insurance companies those are just some of the key informants that um need to demonstrate support and resources in order to take this resource Paradigm from an idea to a public health tool that could be implemented so one thing that my group has studied are the perspectives of medical geneticists and other Specialists on this topic and what we did was we actually created a survey to try to figure out which genes should be prioritized for implementation so we're probably not going to start with the whole genome but maybe there are some key targets that fit that Wilson and Young criteria that we can all agree upon would be high priority to be added to newborn screening and so we designed this survey we um aggregated information from five different databases of treatable disorders we actually depleted the list of genes associated with conditions that are already on the rusp um that recommended uniform screening panels since we thought you know those are already being screened um and then we actually created a list of 651 genes broken down my clinical category and asked our respondents whether or not they would recommend um inclusion of each one of those genes so here's what it looked like on their end they didn't have to go through all 651 genes they could just select the clinical area with which they were most familiar um and then they were simply asked for each gene you know yes would you recommend that this be included no or are you unsure um since there were so many genes for them to fill out we didn't ask for their reason for each gene but there were free text comments that they could add at the bottom of each section and many people took advantage of that to explain their rationale for certain disorders um so we primarily surveyed medical geneticists and other Pediatric Specialists who care for children with rare disorders um and uh we invited 386 experts to participate and um with lots of cajoling and incentives we got 238 respondents um with a a participation rate of 61.7% and so this was what we found um there was a wide range um there were some genes which it seemed that everyone who responded to those genes agreed should be included in newborn screening um and those were often things that fit the newborn screening Paradigm that we use but may not have a reliable biomarker that can be tested by our traditional biochemical methods like Mass spectrometry so some of these disorders had even been piloted and it was shown that those biomarkers really weren't particularly accurate but the disease itself was still a good Target so um we had some things everyone agreed upon some things everybody said you know really this treatment isn't ready for prime time or it's not really a childhood onset disorder that kind of thing and then many genes that fell into this wide area of discordance but overall over 400 genes actually had um a majority of respondents that supported their inclusion so there were eight genes in particular that had 90% concordance and to us what this list began to represent was um a map or a pathway that perhaps um might lead us toward initial public health implementation of genomic sequencing in newborn screening so as I said the whole genome itself is overwhelming as a public health initiative but these were genes that were all highly agreed upon by Specialists who care for these children um that they would be very high yields that it would be important to test for these things and that we could help to um surveil or treat children who might be affected with variant in these genes so of one of the other projects that we've undertaken is of course understanding the perspectives of parents and we're not the first to do this there are many groups that have interviewed parents and had focus groups of different kinds but one really interesting takeaway from this study which was led by a colleague of mine Dr Jackie Omarian um was that parents had a different view of what was considered treatable than we did as Specialists so in this study we interviewed 20 mothers from diverse backgrounds diverse Geographic backgrounds um diverse educational and ethnic backgrounds and what we found was that what we sometimes consider not a good Target for newborn screening for example childhood onset non-treatable conditions that might cause things like neuro regression where a child would appear healthy at Birth and then develop neurodevelopmental symptoms later and we would think there's nothing we can do we have only supportive care to offer in fact in this case most of the mothers who we interviewed felt quite differently they felt that there was great personal utility to this information and they were very supportive of actually a wider range of conditions than our Specialists had been so I thought this was a really interesting quotation from one of um the interviews that Dr Omarian and I did um a mother when asked about one of these conditions something like rat syndrome she said I wouldn't like to think why doesn't she do her child why doesn't she do those things other kids are age do why instead of advancing is she falling behind so I would like to know not to expect things that will not happen not to compare her with other children and not to doubt myself wondering what I'm doing wrong that makes her fall behind um and so in general you know we found I think reassuringly that um you know parents in theory were very supportive of this I this idea of screening additional conditions as well um and they felt in many cases that information was power so um so to scope out a little bit um there are many programs worldwide that are examining the feasibility of adding genomic sequencing to newborn screening or Pediatric Health Care in some way and in fact there are now over 30 programs that are addressing this question um many of them are working together under the opes of What's called the international Consortium of newborn sequencing or icon which was started by my mentor Dr Robert Green and Dr David Bick in the UK um and this is just a map demonstrating how widespread these Studies have become and down here at the bottom you see their intended enrollment targets um so U many of you may be familiar um with Dr Wendy chun's paper on the guardian study their pilot data and they actually intend to recruit over a 100,000 infants into their study what's really interesting based on my interest of sort of which genes are the highest priority for screening and for implementation is that although many of these programs are all built upon the principles of using the Wilson and young ner criteria and have the intention of including early onset treatable disorders in many cases some with a kind of a broader scope and others with a more narrow one in fact the genes that are being tested by these programs differ enormously um so this was a project that was led by one of my terrific mentees Dr Thomas Minon um who is um an economist and a medical student in Belgium and what this heat map illustrates it's um a jacard index which basically shows the overlap between two lists and so what we see here is that um on on each axis we have the various studies we looked at 26 studies that we were able to get the gene lists for and examine what were these studies you know actually um assessing newborns for which genes were they querying and you can see that in most cases this heat map shows a lot of dark purple dark blue and a little bit of green and what that's showing is that the jard index was quite low there's um quite limited concordance among the genes that are being used in these studies and the only place where we see more yellow in the upper right hand corner is really among commercial studies um which mainly used panels of genes that corresponded to disorders that are already on the recommended uniform screening panel but when we tried to go beyond that when these studies tried to explore an expanded number of genes they actually didn't really agree upon what to screen part of the reason is that a lot of the characteristics that we take into into account when we think about screening are very subjective so um you know for example some disorders are completely curable you know like SMA if the treatment is given early a child can become asymptomatic or we may only have supportive care to offer we may just be able to enroll a child in early intervention um in some cases there's an early onset of disease or there's a much later onset or a wide spectrum and we can't always anticipate when someone will develop symptoms based on the variant that they have um an important tenant of screening is that there's not a diagnostic test for Every Disorder so for some things a lot of inherited metabolic disorders for example you can send what we call an orthogonal or non-molecular confirmatory test um in order to demonstrate that the child actually has the phenotype associated with the variant but that's not available for everything for example you know um an rb1 variant associated with retinol blastoma you would simply need to surveil the child over time to see if they developed retinol blastoma but there's no secondary test that can be done at Birth to ensure that they're really at risk of disease some things have low penetrance some have high and most we don't know truly what their penetrance is and then definitions of severity really vary especially among what a clinician or researcher might consider severe and then what parents consider severe or mild as well so I have two little kids I have a 5-year-old and a 2-year-old and so in looking at this data and in seeing that heat map and thinking everybody is screening for different conditions I thought of a book that they love um which is called We're different we're the same um and it made me think okay there's a lot of discordance among what people think we should be screening for but where can we find the commonalities and so um Everybody agrees like the rest conditions should continue to be screened but outside of those there were 34 conditions that are not currently on the recommended uniform screening panel that had 75% concordance across these lists and the one with stars are the ones that were also among that 90% or more concordance on our survey study as well so there's beginning to emerge some clear genes that would be the highest priority targets for public health implementation what we then did was we tried to create a master list um so we used a machine learning Model A boosted trees model that we trained on 80% of the data tested on 20% and then actually reapplied back to the whole data set to to basically rank the expected inclusion of these genes across lists and we created uh a list of genes that was um ranked on their um expected and observed inclusion across lists and these again are depleted of the resp conditions which everyone is sort of familiar with and expecting um but again disorders like OTC deficiency and glycogen storage disease typee 1 a um have kind of led the pack in every type of study that we've done um and what's nice about this ranked list that we created is that we ranked over 4,000 genes and so for um a study that's just getting off the ground or a newborn screening lab that's thinking about piloting this they can basically draw the threshold wherever they'd like they could say we would like to implement you know 10 genes in our genomic screening Pilot We would like to implement 100 Etc and they can use this list to guide that um so you might be wondering especially since the title of my talk was about fetuses in newborns how this relates to fetuses and um I have kind of a funny clinical role where I see a lot of patients with metabolic disorders both children and adults but I also help to direct our prenal medical genetics program and I really have begun to see um a Continuum between fetal life and newborn life and so that's what has led me to a new interest in fetal screening because sometimes newborn screening simply arrives to late and so um I did my metabolism training at the Children's Hospital of Philadelphia which is where I've actually met many of my colleagues and research collaborators and before I came there um one of my colleagues um Dr Rebecca erens Nicholas had published a really interesting study in which they demonstrated that as more hospitals in Philadelphia became Baby Friendly for example emphasized the importance of breastfeeding and all mothers um there were actually more children who had U medium Tre and asil COA dehydrogenase deficiency or mcad who were presenting severely symptomatic or died even before the return of their newborn screening results and so the reason for this is that the treatment for mcad which has been on the newborn screen for a long time is frequent feeding and so children who are receiving you know robust breastfeeding from you know a mother who is maybe an experienced breastfeeder or whose milk came in really well very early and the baby's got a great latch they may be essentially treated even before they've been diagnosed because they're just you know cluster feeding they're frequently getting breast milk but in some cases where the breastfeeding relationship is more difficult they're not getting this and historically they may have been supplemented with some formula at the hospital or at home but due to this initiative that was becoming less recommended and so what um this team had recognized is that um as the percentage of children who were exclusively breastfed increased so did those who had severe symptoms and severe biochemical markers of disease at the time the newborn screening came back so you know could genome sequencing fix that maybe it's getting faster um you know incredible teams like Dr Stephen Kings Moore's team at R children's has demonstrated you can sequence a genome in 13 hours but that's probably not going to be a scalable approach or at least it isn't right now for um you know huge populations of new boards so what does that mean that means that we need to go backwards in the timeline we need to probably begin to sequence some infants or identify these treatable conditions even before birth to optimize care because identifying a prenatal genomic diagnosis can facilitate a lot of important things like appropriate labor and delivery planning mobilization of relevant Medical Teams if they're not necessarily present at the birth hospital and acquisition of specialized medical formulas medication or other Therapeutics that are needed for immediate intervention so I was a part of two cases that really um you brought this point home for me one was a child who was diagnosed with an inherited metabolic disorder called calaman SE deficiency um in utero because his parents had had carrier screening in that case and they were both found to be carriers of this disease um he had an amniocentesis he was diagnosed with the condition and because of this case report that was published in 2016 um what we knew was that this child in picture B she was um a child who had been diagnosed after birth with calaman C deficiency and her sister in picture a um was diagnosed prenatally because of the known family history and subsequently treated with high dose B12 to the mother which can cross the placenta and so the younger sister actually had typical development she had some opthalmologic issues um but but um her older sister had much more severe symptoms of disease and so in my patient we tried it we tried hyos B12 in the mother and I'm happy to say that at a year of age my patient is doing extremely well developmentally on track and has right now minimal symptoms of disease another case that I was a part of was um a baby who is diagnosed um with something called pyruvate dehydrogen complex deficiency um she had an anomaly that was seen on ultrasound underwent um fetal aom sequencing and what this facilitated was this disorder is not highly treatable but many patients who are um treated with the ketogenic diet can have much improved outcomes um and so what this allowed us to do was make sure that on day of life zero the day she was born that there was a keto enic dietician waiting to start her on this diet and what you can see is that um in red as her Ketone levels went up um in blue her lactate levels went down um and so she never had a severe lactic acidosis crisis or development of seizures which often characterizes this disorder because we simply knew that we needed to start this dietary therapy on day one and she's also doing quite well though so um this led us to think a little bit more about other disorders that fit that mold what types of disorders you know if identified early could be more treatable and what types of things um you know could actually be treated in the first week of life before newborn screening results even return so along with a wonderful colleague of mine um Dr Jennifer Cohen who's at Duke we decided to create what we are calling a treatable fetal findings list so um we created a list of genes that we said um either needed to be treatable in utero or in the first week of life and that the treatment should really improve morbidity mortality or organ damage it needed to be a safe therapy and for those that are in utero theapy is it needed to be safe for both mother and fetus um it needed to be something that was associated with a critical illness or chronic disease and it needed to be a gene that had really been substantiated to have um reliable Gene disease validity so um this is essentially what we have called an integrative review we use two different strategies for these two different lists there are the genes with in utero therapies and the genes with treatment in the first week of life for the genes with in utero therapies um Dr Cohen did a literature review with many many search terms and tried to figure out um what met the criteria of having an available inudo theapy we identified 11 Gans for which there are in utero therapies that are currently in clinical trials and then there were 42 successful case reports or case series sort of like the um calaman c um case that I described um that uh also met the criteria for this study for the genes with treatment in the first week of life we already had from our survey study a preconstructed list of treatable genes or genes associated with treatable disorders we had clinical experts in each clinical area review those and identify the ones that they thought would be most valuable to identify in the first week of life um and we came up with a list of 267 of those so there's some overlap between these two lists um but in total it's about 200 and 50 genes so for In Utero therapies there are a number of mechanisms by which these can be delivered it can be an oral medication that's given to the pregnant person umbilical cord vein injection amnio reduction in disorders like barter syndrome an intraamniotic injection um or fetal surgery and then Dr Cohen also identified many animal models of in uog gene therapy um so Dr Cohen identified 53 gen that are associated with inudo thies as I said and um this is a patient of hers this is the first child who was a participant in what's called now the Pearl trial um who had older siblings who were known to have infantile onset Pompei disease and she was treated with enzyme replacement therapy for that disorder in utero and here she is I think she's actually not even one-year-old in this video and you can see she's walking um which is an amazing outcome because children with Pompei disease the infantile onset form typically have such severe muscle weakness both skeletal muscle weakness um cardiomyopathy and eventually respiratory muscle weakness um that this outcome would never have been possible um even among some children who are found on newborn screening and treated um quickly after identification um so some of the other treatments in week one I've sort of put in different buckets of different themes um for example first Do no harm so the first example here is a variant that is associated with aminoglycoside induced deafness a variant in the empty DNA and aminoglycosides are commonly used to prevent neonatal sepsis and so early identification of this variant in the prenatal setting would allow for a niku team or a nursery team to Simply select a different anti biotic and not put the child at risk for hearing loss which would be acquired um there are a group of genes or two genes associated with arthr poetic protop poras and in those disorders the children are um at risk for jaundice which is obviously very common in the newborn period however unlike typical children who develop jaundice um these children are extremely photosensitive and the standard therapy for jaundice phototherapy or Billy lights can actually cause disfiguring Burns in children with this disorder so knowledge of this in advance would allow people to understand the underlying cause the rare cause of jaundice in these cases and instead of using phototherapy apply a blood transfusion or um a prophylactic splenectomy um an inherited metabolic disorder like hereditary fructose intolerance um is something that is exacerbated by the intake of fructose and so because many infant formulas contain fructooligosaccharides and um sweedies which is uh just sugar water often is used as an analgesic or a way to Tom babies during exams things like that um since that also contains sucrose those things can trigger symptoms like hypoglycemia vomiting in these children and could simply be avoided if the diagnosis were known in some cases um critical illness can be prevented entirely if we know about these disorders before children become symptomatic um in biotin thiamine responsive basil ganglia disease for example um applying oral Biotin and thiamine both of which are very benign treatments um could prevent basil ganglia stroke and oftentimes these children instead present with symptoms rather than being treated prophylactically um or in this form of epilepsy complete seizure control can be achieved but it's achieved with um an anti-seizure medication that's not the first line therapy typically and so this would allow for these children to have a much more rapid treatment of their seizures because the right medication could be applied early um in some cases knowing a diagnosis before birth could even improve lifelong health so for genes associated with Long QT syndrome these are not fully penetrant but there could be shared decisionmaking around um what next steps a parent might want to take um because these children are often asymptomatic but Long QT can first present as um a stillborn or a um unexplained infant death and can be treated with beta blockers or an implantable defibrillator um things like congenital hyperinsulinism as well um you know if untreated if unknown the hypoglycemia that results can lead to permanent neurologic injury and what I think is interesting is there are different treatments based on the gene or the variant that a child may carry and so knowing that genetic diagnosis right away would allow for the earliest application of therapy so I think the creation of the gene list was an interesting idea but one lingering question that um remained was would we actually find fetuses with these variants um because they're really rare and maybe we would have found them anyway some of them are associated with anomalies some of them would have become symptomatic so as a pilot study I just got back this data today um one thing that we've done Dr Michael Dyson whom I work with at Mass General Hospital queried um 714 genome sequence trios that included probands with anomalies and we didn't find any of those fetuses that had uh two pathogenic or likely pathogenic variants in genes associated with AAL recessive inheritance but interestingly over 3% of this sample did have pathogenic or like pathogenic variants in genes associated with autosomal dominant or EXL inheritance and so I've listed them here um there's more than one fetus that had um several that had variance in several of these genes but you can see that we did find fetuses who would have been eligible for either potential in utero therapies or treatment in the first week of life um and so um this yield is pretty similar to the secondary findings list that the acmg recommends for children and adults which has historically um you know had variance in many large studies that affect about two to 3% of people so I think the next question for us is what will we you know want to do with such a list um and so I think the most logical step is that um right now the um American College of medical genetics has a secondary findings list that's re recommended for children and adults it's specifically not recommended for use in fetal exom sequencing however it's some diagnostic Labs do use it that way um and now the international Society of prenatal diagnosis is recommending that fetal exome be offered to any pregnant patient whose fetus has any kind of anomaly which constitutes about 1 to 2% of pregnancies so as fetal exom sequencing Tak hold in a larger number of clinical patients we are suggesting that maybe it's time to make a secondary findings list that's specific to fetal sequencing and that these are the conditions that are really most actionable in the fetal period or immediately in the first week of life now these doesn't that doesn't mean these are the only important conditions these are the only conditions that might affect reproductive decisionmaking but in the spirit of the secondary findings list of identifying things that are actionable or treatable this may be one place to start um additionally um at Mass General um Dr Michael talowsky group had demonstrated in 2023 that it is feasible to sequence a whole exom from self-re fetal DNA from a pregnant mother and So eventually maybe this would be um you know the basis of some sort of self-re fetal DNA test that could be offered early in pregnancy as well so as we think about all these directions for genomic newborn screening and for fetal screening or a fetal secondary findings panel I think it's important to really take a step back think about future directions for genomic screening and think carefully about whether or not this is truly a good screening test you know for other types of screening tests we need to understand their parameters their sensitivity specificity positive and negative predictive value and the same should be asked of genomic sequencing as well so in a few different pilot projects from U my research group we've examined some of these questions and hope to dive into them more um Dr Sarah Bick who was a fellow in our program um she asked the question of how sensitive is genomic testing for in this case treatable in Ed metabolic disorders so what we did was we looked at over 8,000 cases of kids who had been seen in the Boston Children's genetics and Metabolism program we identified um 635 who had metabolic disease that was diagnosed clearly from their biochemical testing and clinically labeled by a metabolism doctor as being the definitive diagnosis for that child and then we asked the question of so how many of them would have been picked up by genomic sequencing how many of them actually had a genetic diagnosis and in fact what we found was that actually only about 75% of them did so there was this possibility of missing 20 to 25% of patients um and I won't go so deeply into Sarah's data but what she found was that there were many patients with variance of uncertain significance or who had just one pathogenic or likely pathogenic variant in a disease associated with aisal recessive inheritance and that likely there may have been an antronic variant or something like that so some of these problems can be solved with genome sequencing as opposed to exom sequencing or um by um curating more variants in um people of non-european ancestry over time um because we did find that children who were um who whose medical records did not identify them as white were less likely to have a genetic diagnosis for example um but right now you know we would miss people and this is um you know a larger proportion than is missed for example by many of the tests that are on biochemical newborn screening now we've also began to ask the question of how specific is genomic testing for treatable disorders so a few years ago um I worked with another colleague of mine to look at the Nomad database and what we did here was that we investigated um among Nomad which should contain um you know hundreds of thousands of samples from ostensibly healthy adults um we wanted to know how many of them had variants that were pathogenic or likely pathogenic associated with um rare treatable diseases so in particular in this study we were looking at disorders that were treatable with early bone marrow transplants um because disorders like um uh like some of the um primary imuno deficiencies that have been added to the newborn screen are treated that way and so um stepping back we said well there are other disorders that could potentially be treated with that same therapy but that's a very costly and risky therapy to apply if you're not sure if someone has disease and so although we found small numbers it was just um you know about one in um 2500 people had one of these variants that's a lot if you're thinking about applying bone marrow transplant to them and so um I became very interested in this idea was this a sample mixup was this somebody who had mild symptoms were they misdiagnosed or maybe these were less penetrant than we expected and so that led me to um one of my current projects which has really captured a lot of my attention which is asking the question of what is the positive predictive value of genomic screening and so to approach this question I've actually sort of looked backwards um as I said there's you know many programs that are investigating the feasibility of genomic newborn screening but what I decided to do was instead to look at adults and look backwards and identify adults who have um variants associated with treatable genetic diseases and ask do they have symptoms so what we're doing in this study is we're curating variants in we haven't quite gotten to 300 genes yet but we're getting there we we've started with 54 Gans um and 53,000 participants in our Hospital bio Bank who have undergone exom sequencing and then we've done an electronic medical review to assess their symptoms but beyond that we've also reconed them which is allowed through the consent of our biobank and invited them to participate in a clinical visit at which more deep phenotyping and lab work can occur and we're hoping to expand this to other disorders that other investigators in the hospital might study as well so these are early days but what we've found so far in this study is we've queried the 54 Gans that were at the top of that survey study so they all had 75% or more concordance among our respondents so we would consider them very high priority targets for genomic newborn screening and across 53,000 people we found that participants had reportable variants these are just adults in our biobank who um you know who get their clinical Care at our Hospital in 11 genes so there were some genes for which you know no people had reportable variance and it's either um one pathogenic or likely pathogenic variant in a gene associated with autosomal dominant or exlink disease um or two such variants and one associated with autosomal recessive inheritance but across those 11 genes we actually have found 82 people which is about one in 650 who had these variants and so this begins to give us an idea of the scope of what we might find potentially this is not a population wise sample it is selected from a hospital but an idea of how um prevalent these might be and we found that the vast majority of these participants according to their electronic medical records were undiagnosed they did not know that they had these conditions even sometimes when there was a family history of disease um of the 58 who were undiagnosed there were actually 25 of them who had symptoms 10 of whom had what we would have called severe symptoms based on the review of a medical geneticist of their medical records so we've Rec contacted all of these people and we're beginning to see them and some really fascinating stories are emerging that I think do illustrate the anecdotal power of genomic screening um so for example um a 65-year-old man with a variant in f8 um who had a history of a dermal hematoma from a minor fall had been diagnosed throughout his life with idiopathic thrombocytopenia he had chronically low platelets and no one knew why um on exam with targeted questioning about symptoms of this disease he revealed he had daily nose bleeds he had gingival bleeding that would occur when food repped against his gums while he was eating that was difficult to stop he had a history of GI bleeding and his factor8 activity was found to to be deficient it wasn't extremely low he didn't have the classic inversion in f8 but it was 33% of normal which is diagnostic for this condition um we've seen a 30-year-old woman who had an EX-L variant in a gene called phka1 which which is associated with a glycogen storage disease that causes muscle symptoms this is historically thought to affect males more but um of course due to skewed X in activation can also affect females and she had a long history of chronic muscle pain following exercise and exercise has been really challenging for her whole life due to days of muscle pain afterward but you can imagine she's been told to push through it that it's due to deconditioning all those things but her lab her lab tests actually revealed that she had an elevated creatine cyas or ck at rest which is uh you know biomarker that's very reliable for this disorder there was clearly um you know injury to that muscle which was occurring even when she was not exercising so this pilot data has raised a lot of new questions a lot of new directions that we'd like to explore one is I think we want to Pilot the use of this treatable fetal findings list in pregnant patients we want to understand does it help them to access um care better does it help them to have improved autonomy and have an option of enrolling in clinical trial um or does it allow for earlier treatment to occur and compar with the natural history of that disease did their child seem to have an improved outcome we want to continue to aggregate all this data from genomic newborn screening studies worldwide through the icons Consortium so that we can best understand the um you know the clinical utility and personal utility of these screening results and eventually with this return of results program progr to adults we would like to expand that to additional genes and to larger biobanks and we're working on that now so um in conclusion um you know newborn screening has been a highly successful public health program since the 1960s but we can find infants at risk for more treatable disorders using genomic sequencing that fetal screening can facilitate in utero and immediate postnatal care and is a new frontier for to explore potentially first using a new secondary findings panel for fetal sequencing and third we can begin to understand the characteristics of genomic screening like its sensitivity specificity and positive predictive value by using biobank data and deeply phenotyping participants who seem to be at risk for disease based on their genetic findings um I work with many many amazing collaborators um without whom this work would not be possible and just want to recognize some of them here um these are my own little kids about a year and a half ago um and thank you so much um for inviting me today thank you Dr gold that was a terrific overview and really um the perspective of going across the um lifespan was really terrific something that that really resonates with um some of the work that we're doing and that is also being done at nhgri so um there are some questions I I wanted to just start with a question then I'll get to the Q&A and I want to encourage anybody who has a question to put it in the Q&A um so with the pregnant patients the focus has largely been on those with sonographically detectable anomalies can you think of a way um to get access to bios specimens from fetus that look structurally normal because many of the metabolic conditions are not going to present with anomaly so what is your thinking about that how are you going to identify those cases yeah so there are teams that are looking at non-anomalous fetuses like Ron wapner's group at Colombia um and I know in some other countries in Israel they've also begun to do more sequencing of non-anomalous fetuses and have published that data um but you're right I think that's exactly what we're missing and so um I think um that our first thought is to at least start with the ones with anomalies and find variants that may not have been considered reportable because they weren't diagnostic of the phenotype that was observed on ultrasound um but eventually we will need to certainly expand um either by partnering with another research study or um by inviting pregnant patients with non-anomalous fetuses to participate in this kind of a study with a gene panel I also wonder what kind of Outreach you could do to families that have a positive family history and then you could test your um technology on those subsequent pregnancies but also I realize that not many people are having amniocentesis these days but that might be another source using amniotic fluid for people of advanced maternal age who are not having self-re DNA sequencing for some reason or another yeah that's a great idea yeah okay so there are now three questions in the Q&A so the first one uh is from an anonymous attendee there are muscular skeletal disorders on the system evaluation list that was one of your earlier slides is it because this type of dis these type of disorders are usually not treatable I'm not sure that I know which um list I apologize they're referring to yeah I think it was in the relatively early in the talk when you were looking at different systems that might present prenatal and you were trying to come up with a list oh okay um yeah so I think we we were really only focused on things that are treatable so I'm not sure I maybe maybe they're suggesting their disorders that we missed which are that's definitely quite possible but I'm not sure which specific musculos skeletal disorders there are muscular skeletal disorders that are treatable so I'm not sure why that person would have said that okay um Ben Solomon who's the Clinical Director at NHD said in the analyses you've shown you've rightly I think excluded the rusp conditions since they're already usually included the panels however I'm curious how they might be prioritized these days if we were to approach things without the historical precedent do you feel like the newborn screening list would be entirely different or is there literature that's already looked at this if so pardon my ignorance um that is a great question and one that I've thought about a lot I don't think that there's literature that's looked at it um to my knowledge there is a paper that sort of like revisiting the Wilson young nerve criteria in the genomics age and then actually icons um one of the subcommittees is creating some consensus criteria that like operationalizes the criteria for genomics um but yeah there are things that are on the RP that speaking as a metabolism doctor give us all you know a headache um things like short chain asil COA dehydrogen deficiency have um been very controversial as a disease entity um there are some studies that show that there are symptoms associated with it but for the vast majority of children who are picked up on newborn screen and thought to have that condition um they never May manifest symptoms um and so that would be one example where um that or three MCC syndrome um we sometimes um across we actually mentioned this in the um in the discussion of the Thomas Minton study that those were included widely across the genomic sequencing studies even though they probably don't really fit criteria anymore that we kind of have anchored on these disorders even if they're low penetrants or seem over time to be more of a biochemical variant than a true disease great um tresa L who's one of our post Balor fellows she says thank you for your presentation Dr gold I fully support prenatal screening for treatable diseases however there may be ethical concerns in cases of accidentally finding untreatable diseases in your opinion do you think these untreatable diseases should be disclosed as well yeah I think it's a terrific question um in a complex one um because as I said you know we're calling this the the treatable fetal findings panel we actually changed the name from what we were initially calling it which was the actionable fetal findings panel um because um you know a pregnant patient you know could could make a reproductive decision on any type of genomic information that's offered not just a variant in the genes that we included um so I I guess I always sort of think about shared decision making and that you know um maybe different options could be offered to um you know to to a pregnant patient so that they can choose what type of information they might want to receive that kind of thing um that can get complicated but I like to think that with a good explanation that would allow more patient autonomy right um the last one that's on the no there two there two more now um so from Larry Brody one of our faculty members how do you avoid confirmation bias in the biobank studies is the person doing the EHR told about phenotypes to expect the same question would apply to the recalled patient do you recall a sample of the gene negative patients and have them undergo additional testing yeah it's a great question um so there um is the person who is doing the electronic medical record review does know the variant um but they're not the person who is assigning the clinical score of severity um the reason why they have to know what the variant is is because there are different um data capture forms for each disorder because we're not going to like try to capture the same information for somebody at risk for um you know hemophilia a as we would for somebody who has a cancer predisposition so um for that reason that person knows but they're using a structured intake form that has fields that they have to fill out um that's taken directly from the chart and not paraphrased in any way and then someone else reviews it um we do plan to do a matched case um control cohort at the end we've done this before in another one of our studies that used the biobank where we used the um icd10 codes of the people who were affected with the variant versus those who did not have the variant and they were matched on age sex their number of encounters to the hospital and other variables um to ensure that as you said that there's not just confirmation back so that's one way that we can check that okay and the last one that's in the list is from an anonymous attendee which of these disorders if any have potential candidates for induced plop poent stem cell modeling and treatment um a great question um we really stuck to disorders that had either in uo therapies that are currently in clinical trials or that um had um case reports or case series that were described using a therapeutic agent that is um already like standard of care for something else um and then in the treatments for the first week of Life those are all standard of care therapies that are just being potentially offered earlier um so I'm sure there is a smarter person in a slightly different field than I am who could answer this but we were really sticking to um established therapies and not yet um things that uh are new candidates for things like stem cell treatment great actually another couple of questions have come in uh so the first one is from arini manoli who you may know has written on cabalan C deficiency uh wonderful presentation thank you Nina can you comment on the ven diagram overlap between your priority treatment panel AOG acmg recommendations and the large commercial prenatal expanded carrier screening panels great example of the prenatal pickup of the Cal cabalan C case and successful treatment um well first of all thank you Dr manoli because as you know you were the person I emailed when we diagnosed it and who directed me to the right case reports and to the right therapy and I'm really glad to share that that that patient is doing well um yeah I don't have I haven't created that bend diagram um but it's a it's a really great next step and something that we should definitely do um because you're right um there are probably substantial areas of overlap and then some places where we differ too so um that's a that's a really good idea for a future presentation okay another anonymous question are the diagnoses added to the study participants electronic health record and or are the patients responsible for finding follow-up care um so they are recontacted they are invited for a clinical visit and that's where um the clinical genetic testing is offered and once that's confirmed once the diagnosis is clinically confirmed if it's if it's a genetic condition like a metabolic condition they become my patient um if it is somebody who has a cancer predisposition syndrome they're referred to our Cancer Care Clinic um or our hemophilia Center Etc so um we also feel like this has been a Triumph of translational Medicine of really linking research and clinical care um because in many cases what we've done is found people who have been misdiagnosed or symptomatic but undiagnosed and we're now connecting them with either preventive care or treatment for their disease so nobody is left out in the cold um they all have um whatever follow-up care they choose to pursue but it is offered to them right another anonymous attendee how would one be referred to the clinical trial um so I imagine that they're asking about in utero clinical trials um so I'm not personally involved with them um Dr Jennifer Cohen who's my collaborator is involved with the Pearl trial um along with Dr Tippy McKenzie in California so I imagine that they would be referred the same way as somebody who has a known family history or For Whom the diagnosis is currently picked up on Carrier screen something like that um once these diagnoses are confirmed on fetal sequencing I think you know they're referred to check for eligibility in those trials um Molly maner who you may know from your nicd funded uh Grant who's now at NHD um do you think our society is ready to add genomic sequencing into the newborn screening public health system if not what can we do to help get Society ready for this um yes that is the the billion dooll question I in in many ways it feels that we've overcome many of the barriers it feels like um you know there would be societal acceptance there have been lots of studies of parents of the public of people with rare diseases that have shown support um I'm sure one of the major barriers is is resources and money um and you know getting this into the labs um so I'm hopeful that you know many of these studies as they demonstrate that they have found kids with treatable conditions it's certainly compelling um but um you know the powers that be I guess will have to decide when and how it gets implemented that was a little bit of a trick question Molly um but everybody stay tuned in that space that's all I can say right now um less you you tuned yourself out I didn't know whether you wanted to ask a question or any concluding remarks but maybe you had to um go somewhere else it looks like the questions have slowed down and we're five minutes over time I always feel like if you get a lot of questions it means people were interested and engaged and um it's really a sign of a great talk so you can see the little Applause and hearts coming up the Emojis on the side so we want to thank you very much for joining us today thank take care good night everyone bye everybody