 So, at the May Council meeting, there was a request to have a report on the newborn sequencing in genomic medicine and public health or N-site program. Did you find you're doing that? The goal of N-site is to explore the opportunities and challenges of using genomic medicine in the newborn period of life. And Anastasia Wise, program director at Division of Genomic Medicine, is going to give that presentation. Thanks, Rudy. Okay, so I'm just going to give everyone here a brief update on the N-site program today. We're not quite as far along in the program as some of the other updates that you've heard today in part because we got a little bit of a later start. But I did want to let everyone know where we stand today. So I'll start off with a little bit of information about the background behind N-site, provide you with more details on the individual site's protocols, present you with a couple of the early findings that we've gotten to date, and then tell you a little bit about an upcoming workshop that we're planning with NICHD. So to start with the program background, N-site stands for newborn sequencing in genomic medicine and public health. So we totally cheated on the letters, but there's a reason why. And this is because this is a joint initiative between NHGRI and NICHD. And we really wanted this to represent the fact that we were looking at newborn sequencing as a technology and then trying to see how this could be applied in different settings where NICHD was particularly interested in the context of newborn screening because that's somewhere where they do a lot of research. And then in more of the genomic medicine setting as well where we were looking at how the sequencing could be utilized and if this could potentially be one of the early cases where genomics could potentially be used for the care of newborns. We have four sites in N-site. One in Boston with Allen Beggs and Robert Green as the PIs. One at Rady Children's Mercy Hospital in Kansas City, which is with Stephen Kingsmore as the PI. A group at UNC Chapel Hill with Cynthia Powell and Jonathan Berg as the PIs. And then a group at UCSF with Jennifer Puck, Phueg Kwok, and Barbara Koenig as the PIs. And so they're a diverse group of sites and they also have a diverse group of settings. And within N-Site there were three main questions that we wanted to cover in these projects and these were all covered in the FOA. And each of the projects are addressing at least one of these questions and many of them all three in different ways. So the first of these questions was looking at the currently screened for newborn screening conditions. How could the use of sequencing data be used to either replicate or augment no newborn screening results? So can you get the same information that you do today from biochemical testing using sequencing for newborn screening? Then our second question was really building upon this to look at what knowledge about conditions that aren't currently screened for on today's newborn screening panel, could we learn from having sequence data? And then our third question was to look at what additional clinical information could potentially be used by having the sequenced information that could be relevant just broader to the clinical care of newborns. And so I mentioned that each of these projects has a different setting and the way that we've broken these up for thinking about it with N-Site is that the first of our settings is really looking at prevention and this is what we consider to be the newborn screening setting where you have a condition that you are screening for and you know that if you get a certain result that that could potentially be used for taking some action that could help with preventing a disease or condition from developing. The second is diagnostic and this is a setting where we're looking at the NICU or sick infants that could potentially benefit from having sequencing information in order to be able to help provide a diagnostic. And then our third setting is more of a predictive setting and this is where we have some projects that are looking at healthy infants and moving beyond just what you could do in more of a newborn screening setting to what could you potentially do with this information if you had it available, looking more towards the idea of lifelong genomic medicine. And then the last general piece of information to know about these projects is that for each of the four N-Site awards they were required to have three components. So each of these projects had to perform some level of genomic sequencing so they had to do something more than a targeted panel. They also all had to have a clinical research component and then each of the projects had to have an ELSI component looking at the ethical, legal, and social implications of the research because we felt that that was really important in these pilot studies for everyone to be looking not just at can we do this, but the implications of having those results. So now I'll move on to a little bit of information about the individual protocols at the sites. More details about the studies can be found in a marker paper that came out in pediatrics in February. So if you have interest in learning more about the study that's somewhere that you can find more details. So the first of the studies, if we start once again on that platform moving from traditional newborn screening towards genomic medicine is the study that's being done at UCSF. And in this study they start with dried blood spots from the newborn screening program in California so that there is less of a population bias because these are really done from the newborn screening program which is mandatory collection. They are looking at newborn screening and that's their larger cohort but then they also do have a very small cohort that's part of their clinical study that's looking at patients with immune disorders. Overall they're doing exome sequencing at a part of their study and what they're trying to do is to evaluate sequencing for newborn screening. And so they have both focus groups and some groups that are looking at the legal implications of using sequencing in newborn screening and their two projects are also both very focused on evaluating newborn screening. So they have this large cohort of the dried blood spots from the newborn screening program in California where they are looking to see can sequencing replicate or augment the known results from those tests and then they have this small cohort of immune disorders separate from the population cohort from California that they're looking particularly at the clinical implications then of having more of this sequence information available. For the UNC Chapel Hill study design they have two different cohorts. They have a healthy and a sick infant cohort, individuals that have been diagnosed with newborn screening conditions that are known conditions. They are collecting saliva samples for their source of DNA and the large focus of their project is on electronic decision aid. And this electronic decision aid is then allowing for parents to be able to make decisions about what additional information they would like to receive if they are randomized to be able to have the decision aid as part of their, part of the study. So if they are randomized to this decision aid then beyond receiving back just the information that's related to the newborn screening actionable conditions in childhood then they also can receive back three other types of information. That's childhood onset and non-medically actionable information that is adult onset but medically actionable and information on carrier status for recessive disorders. And what they're really trying to look at in their study is parental decision making and trying to learn more about the decision making process and if this type of decision aid could potentially be useful as you start looking at scaling up using genomics in medicine. The third of the projects at Rady Children's Hospital is focusing on a NICU cohort so these are all sick infants. They are all suggestive of having a genetic disease that was one of the requirements for enrollment in the study and they are then randomized to receive rapid genomic sequencing and in this case the sequencing can be done as quickly as 26 hours and typically it's done in a matter of days. They did allow crossover between the arms that was just receiving standard testing and the group that was receiving the rapid genomes and then followed this up with parent and clinician surveys and I have a little bit more data about some of the results from this study but because of the results of the study which showed that there was some benefit there seemed to be from having the sequencing done there started to be an issue with the clinicians themselves not necessarily wanting to have their patients be randomized. So they actually have a new study design that they just started recently since they have moved from Children's Mercy Hospital in Kansas City out to Rady and using this new study design it has many of the same elements of the first study but now the randomization is between doing a rapid exome and a rapid genome in order to be able to help maintain that equipoise in the randomized clinical trial design. And then the last of the studies to present from Insight is also looking at a healthy and sick newborn cohort out of Boston. This group is randomizing to be able to receive genomic sequencing in this case it's typically going to be exomes that they receive and then they have the option of being able to receive an indication based analysis if there is genomic sequencing. So within the sick cohort they would receive if they got the genomic sequencing an indication based report based on whatever the condition was that that infant was presenting with at the NICU but also within the healthy cohort if that infant presented with some sort of condition within their first couple of years of life while they were being seen at the hospital and they had the sequencing results on file they could order an optional indication based test based on the information that they already had in the database. And they are also conducting parent and physician follow up surveys in order to be able to learn more about whether or not both the physicians and the parents found this information useful to have returned if it made any difference in their care management as well as looking at some of the economic outcomes. So then moving on a little bit to the findings to date starting with the question of replicating or augmenting current newborn screening results. The UCSF study has found that exome sequencing currently cannot replicate all of the known newborn screening results out of their first 182 exomes they got about 12 percent that were false negative so right now they're not able to completely duplicate what you found from the biochemical results and find all of the individuals who have been confirmed to be positive. However, what they have seen is that exome sequencing can be useful to be able to augment newborn screening results. So if you think about this more as potentially a second tier of testing or an additional test that could be run clinically that being able to determine exactly what gene is involved in one of these newborn screening conditions or what variant can potentially be useful in being able to make decisions about treatment and it can also be useful in distinguishing a condition. Some of these conditions that are tested for on the current newborn screening panel have biochemical test results that have overlapping thresholds so sometimes it can be difficult to be able to distinguish exactly which condition an infant is positive for and having the genomic information can help with making that call. Just looking at the question of what knowledge could we gain about conditions that are not currently screened, the UNC Chapel Hill study has been really working on this electronic decision aid for the parents to be able to inform their decision making and evaluating the outcomes from this testing. And this allows them to be able to look more in a general context at what would happen if you actually could get back all of this, these different types of genomic information from an infant's test. So beyond the fact that everyone receives back these childhood onset medically actionable conditions, the individuals who are in the randomized section have these options to be able to receive additional information and they're hoping that by being able to study this decision aid and look more at the process of how parents make decisions can learn more about how useful this information is on conditions that are not currently on the panel and whether or not parents find value in having this type of information. And then towards the clinical care of newborns, from the Rady study looking at the NICU population it appears that sequencing can lead to changes in diagnosis and clinical management. And from their preliminary study they were able to diagnose 20 out of 35 of the infants and in 13 out of 20 of the individuals that were diagnosed, the diagnosis impacted clinical management. So it was changes in medication, some of the infants were put on palliative care or there were changes in the reproductive genetic counseling that was provided to the parents. And so this is part of what drove that change in their model when they had their move. And then finally there's been some other additional findings that have come out of insight that were unanticipated. One of the interesting ones that we have a writing group working on right now is related to enrollment and understanding reasons for decline. So initially the Boston group had done a survey based on asking parents if they would be interested in receiving sequencing results back on their newborn. And when asked about this in a very hypothetical setting they found that 46% of these parents reported that they would be very or extremely interested in doing newborn genomic testing. But then when they actually went to the clinic to do this they found that at first only 7% of individuals would actually enroll in the study when presented with this in an actual setting. And so this really showed that it was similar between both the NICU and the healthy baby cohorts but they found that there were actually some slightly different reasons given where at first if you presented this to the parents the ultimate reason that they would turn this down just when approached by a nurse at the hospital after birth was that it was a logistical concern. We don't want to come back to deal with this study. We don't want to be involved in a research study right now. We just had a new baby. But that if individuals were willing to then consider being enrolled in a research study and actually talk to a genetic counselor about the consent form then you started getting individuals who were concerned about privacy and whether or not the results were uncertain and potential insurance discrimination. So they're finding some interesting results there about enrollment and due to this finding we've actually started looking across all of the different insight projects because they are doing their enrollment in different ways to try and see if we can learn more about the reasons that these parents are making a decision to either participate in research or not. And that leads into the fact that we do have some network-wide working and writing groups and insight beyond these four projects. We don't have a set coordinating center for insight but NBSTRN which is the newborn screening translational research network funded by NICHD does serve to provide some of the coordinating center services to insight they help us with scheduling some of our calls and helping out with some of the meetings and have helped a lot with the common data elements working group, collecting information and helping with getting that data into their longitudinal pediatric data resource which is a resource for researchers in the newborn screening community to be able to learn more about research projects that are going on in that space and get data for doing analyses. That is one of our two working groups, this common data elements working group. We also have a working group related to the ethical, legal and social implications and looking at the different studies that are being done across insight. And then we have three different writing groups that are going on right now. One related to enrollment as I just described. One related to the investigators' experiences going through the FDA investigational device exemption process. And then a paper looking at variant interpretation across the studies and how the different context can potentially have an impact on what kind of information you might want to return. And with that I'd like to tell you just a little bit about an upcoming workshop that we're planning. With the successful collaboration of insight between NHGRI and NICHD we're looking now to explore further opportunities for collaboration and looking at hosting a workshop in spring of 2018. And the idea would be to be able to look and evaluate the gaps and any new emerging opportunities within this space. We had a similar meeting focused on sequencing for newborns back in 2010 which kind of is what led to the insight program. And at this point we'd like to really look broader than just the sequencing and newborn context to be able to evaluate whether there may be new opportunities for our two institutes to work together going forward. So we hope to be able to provide you more information about that going forward. And with that I would like to thank all of the insight participants and their families. All of the investigators that are part of the insight program and then all of the program staff here and at NICHD who helped to make the program possible. And with that I'll take any questions. Thanks. Yeah I mean there are going to be additional presentations for example at ASHD and I think there's some additional papers coming out about just the clinical use of exomes in the NICU. And I can just tell you that you don't have 90% of parents decline. So I think there's also something about the Boston study where if it was likely they had a Mendelian condition they were exempted out and offered the testing as a clinical test. So I think there may be some additional reasons it was so low. But I do think this is an area where clinical practice is pretty rapidly moving both from the results of insight and other centers just doing it. And so I wonder how much presenting it to the families as a research question with a control group may have also influenced people as opposed to just a new diagnostic test that we're using in other children and we're recommending it for your baby. How much that influence you think might have influenced the low uptake and was how much did the uptake differ in the other centers? No that's a very good point and even just comparing the uptake from the enrollment that was done with the Boston group in their NICU population that was where these individuals were approached shortly after giving birth within like 48 hours before they left the hospital that's a very rough time for people to be approaching someone to be able to do a research study versus the way that it was handled with the Rady Children's Hospital group which is also recruiting from NICUs and had a much higher recruitment rate. I do think there is something that's about the difference in the way the studies were being conducted. But that's definitely one of the things that they're trying to look at in the enrollment paper is just trying to see even in some of these places where they were looking at different settings did the way that the projects were set up influenced the way that people were making their enrollment decisions. Sharon what were the rates that you... I actually I can't give you a number but it's routinely recommended for very ill, I shouldn't say routinely, it is clearly now recommended by Genetics Consult Service for very ill babies in the NICU and the decline rate is... we're talking different when you may want to comment. I mean I'm sure there are people that decline. I mean similarly in our pediatric cancer study we had 30% of parents decline and I think in a setting where the parents think it may be clinically useful the uptake is generally been fairly high. I would just say clinically you know if you exclude secondary findings and just focus on the diagnosis I'd say 90% upwards in terms of agreement. But it's not being presented as do you want to participate in this research trial and I wonder how much that influences. It's a doctor going I think this might be a useful test your baby's really ill and I think that may be influencing these rates. And that is something that we saw even within the other NICU population at the Kansas City group was that once people could actually do this clinically then they were less interested in doing this as part of a research study where you might get randomized. So once again the Boston Group is also a randomization. There were other parts of the study that you had to be willing to participate in beyond just the sequencing. So part of this was just the burden of if people can do this clinically they might not be as interested in being able to do it as part of a research study. And the field has been advancing very rapidly towards that over the course of these projects. Yes. This is such a technical question but I think I saw that not everyone's using the actual newborn screen dried blood spots. Just what percentage of sites are using that and have there been any technical issues in terms of doing that? I don't know CNVs or anything else in terms of what you can or can't get from the dried blood spots? Yes so there's only one group that's using dried blood spots which is the UCSF group and they're only using it for part of their study where they're doing the comparison between the existing newborn screening results and the results from doing exome sequencing. So they've been able to get good results back as far as the quality of the exome sequencing but they haven't looked deeply at things like structural variants or things like that at this point in time. Anastasia I have a question from Jeff Botkin who's watching this but he doesn't have a phone line to come in so he's asking is the goal of the enrollment assessment to evaluate recruitment or to enhance recruitment? Are there challenges to obtaining meaningful results from the studies due to low enrollment? Okay so the information that I presented here was some of the early data regarding the enrollment which really led to why we've started a group that is writing a paper about enrollment. There have been increases in the enrollment at the Boston site they made some changes to their consent protocol and have been able to improve recruitment there and are gonna have no problems with meeting their overall recruitment needs for the project overall. So it's not gonna influence whether or not we can meet recruitment milestones to be able to complete the study but we are interested in looking at the fact that there are such diverse settings across insight. Are there lessons to be learned about enrollment for this type of genomic medicine study that we could learn by looking across these different studies in the way that they were set up? Yeah, please. You in one of your slides you referenced frequency or percentage of diagnoses. How is the word diagnosis defined here? Is it specifically a case where the genetic or genomic information is part of the diagnosis or contributing to it? So it is a diagnosis that was made from the sequencing. May have a related question to the last one that was asked. So you gave some data on the clinical care of newborns where there were I think 20, 35 cases they came up with a diagnosis. I assume that was with whole exome sequencing? Yeah. This is actually from genome sequencing. That's whole genome and my question was whether they had a pre-test pre-test differential diagnosis or they just looked at the genome in general and follow up on that is how do we know that these diagnoses are correct? Okay, so the information about how is this done? Let me go back to the project piece because I think that'll actually be helpful. So this is the project design that those results come from. So these were individuals who were selected for enrollment based on the fact that they were suggestive of having a genetic disease. So they had clinical features that made them feel that this is likely to be a genetic condition. Beyond that the information that was used for being able to do the analysis because this is part of this rapid genome sequencing was phenotypic information was collected about the child that's fed into part of their algorithm for being able to try and make the rapid diagnoses and that's how they then come back with the information that's potentially used then to provide back to the physician who can then make the diagnosis. So it's in some cases returned to the physician before they have an official clinical report if they feel like this could potentially have a really dramatic impact on the care management of that infant where time is of the essence. But in most cases this would be the patient who has been returned back to them as a clinical report saying here are the findings from the sequencing and then the physician makes the determination of a diagnosis. Did they also sequence parents to look at de novo changes for example? They did look to see whether or not changes were de novo and look for segregation and things like that. Yeah. And the new study is particularly looking at doing trios as a potential option where they were looking at provisional diagnosis. Can you get something just from doing a pro band and then trying to see if adding the trios added additional information? I'm curious about what types of actions are taken to improve that end up improving outcomes. So in that case of the NICU you're referring to the these changes in care or management that okay. So they basically fall into three main categories for the most frequently cited changes. Sometimes it is a change in either medication or procedure that is going to be conducted. I know that in at least one case they had an infant that was being crept for surgery because they felt that there was one diagnosis. They got the sequencing results back determined that it was actually a different potential diagnosis where there was going to be a high risk of complications from the surgery. So they were able to stop the surgery. So there are those kinds of more dramatic kinds of cases. Then I would say the next large group is a change to palliative care management where they're able to determine that at this point in time they have a diagnosis and that there is not necessarily a need to continue testing and that they can switch to more of a palliative care regime. And then there's the final category that they were classifying for changes in care and management that came up frequently is related to being able to provide information for reproductive counseling being able to provide the parents with information about what the potential genetic cause was of this was this a de novo change is this something that's likely to recur in future pregnancies. What are your options. And those are the large categories. I think it is also important. I'm trying to find the abstract because I thought it was one of the one of these projects. But maybe you reviewed it and I missed it. It's not that sequencing is always better. Right. There was definitely an abstract at ACMG that metabolic profiling was far superior than exome sequencing. Yes. So that's a different study. Yes. Okay. I'm sorry. Yes. No. So there's there's I just somehow I didn't think that I just want to be sure the group was aware of that that. Yes. So that's the UCSF study. And so that's that's the group that's really looking at exomes sequencing compared to the current newborn screening. So there's a difference between the findings so far in the NICU population where it seems that you can get diagnoses that potentially could actually change management and this more public health setting of can you actually replicate what we currently do with biochemical testing for newborn screening. And what they found is that so far that does not seem to directly replicate. Right. I just want to be able to I mean that the talk was quite persuasive. Like they just miss them. You know and they really couldn't find a number of these mutations even after they broke the code. They still couldn't find the genetic cause. So it was quite striking in the setting of newborn screening disorders how poorly it and I assume they did go back and confirm that. No they couldn't find them. So they confirmed the biochemical test. They confirmed the biochemical test. These were newborn screen biochemically confirmed diagnoses. Yeah they have the information from the California public health labs on both the original testing that was done on the blood spots as well as you know what whether this was you know confirmed. Did they go on and do whole genome sequencing. They have not done genome. They have not. They would like to. Yeah you look. I still struggle with what the purpose of that arm of this project was I mean you're never going to do better. It's markedly more expensive. You already have the gold standard. So I still struggle with with that because it's a question that comes up. There are a lot of believers probably including in this room that but what's the belief that the sequencing might be better and more efficient than all of all of this mass spec. I mean I assume that was the rationale for doing it. On the other hand you don't you test for but it's a set J you put your testing for a finite set of conditions. So sure for expanding outside but within the within the conditions the actionable conditions that are screened for right. And what would at least maybe this exists but I'm not aware of it but what's you know the list of what else you would screen for you know you only screen for the things that are actionable. What's the what's the broader list of things that you would screen for that aren't captured by the metabolic panel. I mean the actionable list is larger than what's screened for. Right. But is there an actionable list. Is there a list of things that there will be. I mean with time this will get that list will get larger in the cancer space. I have colleagues that have suggested we should be doing newborn screening for retinoblastoma which could never be done because of the very early onset and the attempt to save vision which could never be done by any method other than sequencing. So there are disorders for which if you decided you were going to do exomes for the metabolic disorders that you might then add other disorders which right now are not practical. Just a comment on J's thing. So the biochemical tests are are not always unambiguous. So often you get a result and you don't know what it means. It's abnormal. You don't know what it means. So perhaps this could contribute. But I'm sort of with you there. I don't think this is that's sorry the solution. The the other thing I wanted to point out since this is an open session I think in that one line on the comparing the well when you got when they claim to have gotten 20 answers out of 35 and they had the one line that said suggestive of a genetic disease or something like that. That is I don't I don't think that statement should be underestimated of the work that goes into that the clinical work and you would so in reality what they're looking at is there they're using clinical expertise to help guide who to do whole genome sequencing on and that is extremely important and different from just sequencing everybody and that should be noted going forward because you don't want to sequence just every sick infant in the NICU. Many or most of those will be infectious diseases. That's a very good point and this is a very selective cohort especially this first study of them looking like they were genetic and having a neonatologist who was willing to present the study to them and all of those things these this is a higher percentage and as they have then worked to make their enrollment criteria a little bit looser and enroll more infants from the NICU they have seen that number start to come down a little bit so I do believe you have a very good point that this is definitely including a lot of clinical information in there as well but they are definitely seeing some result from having that genomic information available as well even once they start broadening it. The role has been imposed thank you very much Anastasia.