Genetic Testing and Epilepsy-Genetic Testing in Epilepsy: Testing Considerations

Genetic Testing in Epilepsy: Testing Considerations

Video Transcription

All right. So, my name is Heather Mefford. I am a clinical geneticist and physician scientist. I study the genetics of pediatric epilepsy, specifically at the most severe end of the  spectrum, and have been doing genetic testing across many disorders, but specifically in epilepsy for many years now.  So, I'm going to talk to you today about the available tests that we have, how to think about which tests to use and when, what the diagnostic yield is, et cetera.  My slides will be available, so scan your QR code. You don't need to take photos if you don't want to. You're welcome to, but you can also just sit back and listen.  These are my disclosures. I work with a lot of specific genetic epilepsy support groups. So, today what I really want to do is, like I said, outline genetic testing options that  are available for clinicians to use in the clinic for patients with epilepsy. And for those of you who are with patient advocacy groups, this will be a nice reminder  of the types of tests that are available that you can advocate for when your physician maybe doesn't know about them.  And then, really, to think about what is the diagnostic yield of a test and which test do you use when and for which patient.  So, as a geneticist, I always like to start with this, and I think we've already heard some very nice talks or comments about the importance of a genetic diagnosis.  And from the geneticist's standpoint, and I think from the neurologist's, as you've already heard, getting a genetic diagnosis not only ends the diagnostic odyssey, which  can be quite long and frustrating for both the patient and the family, but once you have that diagnosis, it allows you to provide better counseling about prognosis.  What will my child's life look like? We know that there's a phenotypic spectrum to most genetic disorders, but once you have  a diagnosis, you can at least hone in a little bit on what that might look like. We can also talk to families about recurrence risk.  What's the chance that my next child, if I choose to have another child, will have the same condition?  More and more, genetic diagnosis in epilepsy is affecting your choice of medications, treatments, or other types of therapies.  And as you've already heard, a genetic diagnosis often connects families with each other who have the same genetic diagnosis.  And I will say that aside from the impact of advances in genetic and genomic technologies over my career or the past 10 years or so, the other most impactful thing that I've seen  in the largest change is the ability of families to connect and the emergence of advocacy and support groups, and it's been remarkable.  And then as those families come together and connect not only with each other but with physicians and researchers, it provides research opportunities.  If you have a specific genetic diagnosis, you'll be able to participate in specific research programs and hopefully take advantage of precision therapies as they emerge.  So many of you in the room do genetic testing or maybe families who have gone through genetic testing, and it can be really long and arduous to get to a diagnosis.  So there are many different paths to go, many different orders in which tests are done, and often it just feels like you're going around and around in a circle.  What we really want to do is change that to go from diagnosis of that first seizure to a genetic diagnosis in a straight line, ideally with a single test.  I think maybe someday we'll get to that one-size-fits-all test. Not quite there yet, but we're a lot closer. So there are a lot of tests available to send for a genetic diagnosis.  The most probably familiar to most of you in the room would be the chromosome microarray, which has been available since the early 2000s.  And then more recently, and the biggest and most impactful advance, has been sequencing tests, and specifically those that use next-generation sequencing technology.  These sequencing tests come in various forms, and we'll talk about advantages and disadvantages,  but while we used to do single-gene testing, as Dr. Milchop alluded to in his patients, one gene at a time, that is very rarely done today in the clinic, and most tests are done  either as gene panels with a few to dozens to hundreds of genes, or by sending a whole exome, which tests all genes at once.  There are several emerging tests on the landscape that I'll talk to you a little bit about,  including things like RNA sequencing, so sequencing the transcripts, methylation profiling,  I'll talk a little bit about that, and then polygenic risk scores for more complex genetic disorders.  So I think it's important to remind ourselves, what do each of these tests actually look for, and what can you find?  Because it's important to know what you've tested for and what you haven't tested for as you go through this diagnostic odyssey.  So the chromosome microarray, which has been around for a long time, really scans the entire  genome for copy number variants, or deletions and duplications, missing pieces of DNA, extra pieces of DNA, and can detect some chromosome rearrangements as well.  Sequencing panels, on the other hand, or sequencing tests, look for changes in the spelling of the DNA, so changes in the spelling of the gene.  Panel will test, as I said, dozens to thousands of genes at once. The exome looks at all 20,000 genes, so it's important to know that the interpretation  does not include all 20,000 of those genes, because not all 20,000 are known disease genes. Whole genome sequencing, as Dr. Patel talked about, is not used as frequently in genetic  testing, but rather than just testing all the gene sequences, which is about 1 percent of our DNA, it tests 100 percent of our DNA, or the entire genome.  Again, the interpretation of some of those changes is a challenge. Importantly, these tests do not detect some types of genetic changes.  So when you send a chromosome microarray, you're looking for big chunks of DNA that are missing or in extra pieces or rearranged.  It does not pick up sequence changes in genes, so if you're looking for a genetic change in SCN1A, you will not find it with an array.  It also doesn't pick up things like repeat expansions that we see causing Fragile X. CSTB, a form of PME, often has a repeat expansion that is the most common mutation and is not  picked up by chromosome microarray. Sequencing tests are also actually not very good yet for repeat expansions like Fragile X.  More and more, these tests actually will pick up copy number changes, so there are ways to analyze the data that shows you where the missing and extra pieces of DNA are, but you  kind of need to look at that ahead of time or ask the lab if they were able to detect that. And remember that panels test a select panel of genes, so 10, 20, 200, or 1,000 genes,  but they don't test them all. And so if you're looking for a change in a specific gene or you suspect a specific diagnosis,  you need to know whether that gene was on your gene panel. The other thing to consider is diagnostic yield of each of these tests, and they're  very different, and it also depends on the presentation of your patient. And although we're talking about epilepsy today, I've put in some numbers here for neurodevelopmental  disorders and autism, because you can see kind of what the variability is here. And sometimes a patient might present with developmental delay and intellectual disability,  but in fact has an epilepsy disorder that just hasn't presented itself. So I leave Fragile X up here, I know it's not necessarily the most common thing you  think of with epilepsy, but certainly presents, and it's something that in genetics we think about a lot, because it's an inherited disorder that then can recur in families.  But it actually accounts for a very small percentage of patients. Chromosome microarrays are actually quite effective in patients with neurodevelopmental  disorders with a 20 to 30 percent diagnostic yield, so this was a huge impact on genetic testing when chromosome arrays first hit the scene in this class of patients.  In the epilepsies, especially in the developmental and epileptic encephalopathies, it's also  an important test, but the yield is about 5 to 10 percent of cases who will have a deletion or duplication involving an epilepsy gene and causing epilepsy.  In the more common epilepsies, generalized and focal, chromosome microarrays can be important as there are some important risk factors that are deletions of DNA.  Gene panels in epilepsy are better right now than chromosome microarrays, about 20 to 30 percent yield, depending on which test you do and which panel of genes.  And then exome sequencing, of course, has made the biggest impact and has a diagnostic yield of somewhere 20 to 40, 30 to 50 percent, depending on the exact presentation of the  patient. And I just think it's interesting to compare, if you look at autism and ID, where we do  a lot of genetic testing as well, actually exome sequencing has a lower yield perhaps than chromosome microarrays.  So, again, thinking about the presentation of your patient and using a little bit of clinical judgment when you choose your test is important as well, and you shouldn't just  consider the numbers.  So, panel versus exome sequencing, I think we often have to, well, I'm not going to talk a lot about insurance, but I know that's the elephant in the room, but thinking about which  tests we send, of course, sometimes you might not have the option to send an exome and you're going to send a gene panel.  But how do you choose which panel, and, again, what are the advantages and disadvantages compared to exomes?  So, one of the advantages of gene panels that people often consider is that you're only testing genes that are known to be associated with the presentation of your patient.  So, you're only testing for things that you know are associated with disease, not for things that might cause dilated cardiomyopathy or cancer predisposition.  But we can talk about why that's important, and we've already heard a nice example. These gene panels also often detect CNVs, again, ask your test provider or your friendly  genetic counselor or geneticist who might help you figure this out. I think it's also important to know that many panels that are done today are actually sequenced  as an exome, so all the data's there, but the panel just means you're only looking at the data from the genes that are listed on the panel.  This is important because it's a technical aspect, but some gene panels, which are true panels, actually go in and look very, very carefully at the genes on that panel.  When you do an exome, sometimes you – if there's less coverage, you miss a few things, and it's not, perhaps, quite as deep of a dive into those genes.  On the converse side, if you do a panel that's actually sequenced as an exome, sometimes you can go back, then, to the company and say, okay, we didn't find anything.  Can we go back now and do an exome? And all they have to do is reanalyze the data rather than send – have you send another sample and redo the test.  And as I've already said several times, the thing to remember about gene panels is you don't find what you're not looking for.  So if there's a gene you suspect and it's not on your panel, it was not tested, you don't have an answer one way or the other, and you need to make sure that you find the  right test to test that specific gene. The other thing is that you don't test for things, as I said, like for repeat expansions.  And so just – even if a gene has a repeat expansion – has a repeat expansion as one of the mutations, you're not – but sequence mutations as another variety, you're only  going to be testing for one of those when you send a gene panel. Exomes, on the other hand, or genomes, test all genes, so all approximately 20,000 genes.  But it's important to know that most exome analysis includes only the about 4,000, 5,000 genes that we know to be associated with any human disease.  Most of them now include copy number variant analysis, so you will pick up deletions and duplications as long as they're big enough to see from the exome data.  And I think it – importantly, it's important to remember that even if your patient has a negative exome today, or especially if they had a negative exome in, say, 2014, the  data can be reanalyzed, and many companies will do this at least once for free. Sometimes you have to pay a small amount for it, because it is a lot of work to analyze this data.  But the data can be reanalyzed, and, you know, 10 years ago, we had fewer genes that we knew were associated with epilepsy than we do today.  And so if you reanalyze the data, what was negative two, five, or 10 years ago, you may now find a mutation in a gene you didn't know was a disease gene at the time you did  the test. When you do an exome sequencing, this often worries people that you may find something that you're not looking for. So there's a list of genes.  I think it's now up to 75 or so that the American Academy of Medical Genetics recommends that we analyze, because there is an intervention, so cancer predisposition, cardiomyopathy,  for example, genes that you can find mutations in that might not be related to epilepsy or ID or whatever else you're doing, the reason you're doing the test.  There's often an option for families to say, yes, I want to know about any changes of those genes, or no, I do not.  So this is important to involve, often important to involve a genetic counselor or geneticist in the pretest counseling for these tests.  But it can also be very important to test for these things, and as we heard from Leah this morning already, although they were not anticipating or looking for a change in a  gene that causes a cardiac disease, Ben's test result may have saved his father's life. At the same time, you might say, I don't want to find a breast cancer predisposition mutation  in a two-year-old girl. But if you do that, if you find that, there's a 50% chance it was inherited from her mother, who may not know about it, and may save her life.  So it's important to have these discussions with your patients before you send the test and know what the repercussions are of the results that you might get.  And the other important thing to remember is that the most efficient way to do exome sequencing to find a diagnosis is sequencing the child and both of their parents in a trio.  But remember that what you get from that data is information about family relatedness. Sometimes the people you sequence aren't actually related to the program, and you have to be  prepared to deal with that information and know how you're going to convey that or not to the family if you get that.  And non-paternity is one of the most common things, of course, that we come across. It's also important to remember if a patient's adopted, sequencing their adoptive parents  is going to give you zero information. And sometimes people just don't think about that, you know, when you're ready to send the test.  So having that information can be important up front. So what about the results of these tests? How do you interpret them?  So I think there are two ways to think about this, and one of them is getting the variant back and interpreting the variant.  These are the pretty standard now interpretations that you can get on a genetic test. A variant is benign or likely benign.  A variant is pathogenic or likely pathogenic, so disease causing. And then, of course, the dreaded variant of uncertain significance, right? We don't know yet what this means.  What that means for how you interpret the test result is that if you have a likely pathogenic or pathogenic variant identified on the test, this is a positive or diagnostic result.  If you have a benign or likely benign variant, this is a negative or non-diagnostic test result.  I think what's most important to remember is that if you have a variant of uncertain significance, this is also not a diagnostic result.  And in fact, the vast majority of variants of uncertain significance over time will be reinterpreted as benign rather than pathogenic.  That's not to say that some of them will be interpreted as pathogenic, but it is not a diagnosis and often it requires further follow-up and sometimes just time to determine what  the actual interpretation is.  So, genetic testing in the clinic, we've already heard from Dr. Millichap, and I'm just going to reinforce a lot of the things that he's already said. Early diagnosis is important.  And one of the reasons it's important is just to avoid that diagnostic odyssey, not knowing what the cause, why is this happening, is there anything that we can do about it or  not, will it happen again in a future pregnancy. Not only does it avoid the diagnostic odyssey, there may be a treatment option related to the genetic diagnosis.  In intellectual disability, early onset epilepsy, including the developmental and epileptic encephalopathies, it's quite clear from the literature now that the highest diagnostic  yield, the most efficient test to do, is an exome. If that's not available for whatever reason, gene panels are the next best test, and chromosome  microarrays being a good option when those other tests are negative. When you do an exome, as I've already said, trio analysis is the most efficient, and that's  because many of the severe epilepsies are due to de novo or new mutations. So, when you look at the sequence in the child that you detect and you don't find it in either  parent, you know it's de novo, and that helps you very quickly to give you one piece of evidence that this might be the pathogenic variant in your child.  I think earlier the onset and the more severe or syndromic the phenotype in your patient, the more likely it is that you're going to find a genetic diagnosis.  So, this can be frustrating. We try to test every patient, but I think the least severe they are, so if they have  normal intellect in an epilepsy or a genetic generalized epilepsy with normal intellect, it's much less likely that you're going to find a genetic diagnosis than if they present  in that first week of life with neonatal onset seizures. Doesn't mean you shouldn't test them, but you need to use your clinical judgment in  thinking how much testing do we do and how long do we persist. Then, and I think with what we're able to do today, up to almost 50 percent of patients  who present in the clinic with early onset and often severe pediatric epilepsy can receive a genetic diagnosis, which is huge.  Ten years ago, we were diagnosing just a small fraction of patients, primarily with SCN1A, and this has been a giant leap forward.  But that means that there are 50 percent who still remain without a diagnosis and we need to think about how we continue to both  identify what might be their diagnosis and then test for it in the clinic. And then I just want to highlight that in the late October, the National Society of  Genetic Counselors published guidelines for genetic testing in epilepsy, which I think is huge. And we've talked a little bit, some of us, about getting coverage for testing and one  of the ways to convince insurance companies to cover a test is to have both the evidence and the literature but also guidelines that are backed by some of these national societies  and these were endorsed by the American Epilepsy Society as well. And their recommendation was to do exome or genome as a first-tier test in patients with  epilepsy and as an alternative you could do a panel, but it should have at least 25 genes on that panel and include C and V analysis.  The other important piece, as I've alluded to, is that the testing really should include pre- and post-test counseling and hopefully many of you have access to that, but I think  it is very important to think again about how do you counsel patients about what you might find, what you might not find, and what the implications may be for your, not only  for your child, but for the family as well. So diagnostic odysseys can take many different forms and about half the time they actually end up without a diagnosis.  So there are many reasons that you may get to the end of a, or what you think is the end of a diagnostic odyssey and not have a diagnosis.  The first I don't like to put up there because I'm a geneticist, but it might not be genetic, right?  So you need to think about what the other, I think everything is genetic, but you know, are there circumstances that you, when you go through the history that may have actually  contributed to this and maybe genetics is only a part of it or not at all? I've already talked about this. If the gene wasn't on the panel, you didn't test for it.  It may be that it is genetic, but we don't yet know, we haven't yet discovered that particular gene or variant as a genetic cause.  And then that variants can be missed for technical reasons. These tests aren't perfect, they're the best we have today, but I'll give you some examples of why we might miss things.  And then importantly, variants can be misinterpreted. So I think variants of uncertain significance, I don't consider misinterpretations.  I consider that you have to carefully look at them and you may change the interpretation based on your clinical or scientific knowledge.  But also sometimes testing companies just give you the wrong interpretation. And so if you're really convinced that this gene or variant is the cause in your patient  and they say it's not, you reach out to your colleagues who do research or who are geneticists who might be able to help.  So like Dr. Mailchap, I'm going to give a couple of examples of diagnostic odysseys. And these are patients that I've seen in genetics clinic over the years.  This is a little boy who at about six months of age presented with maybe some delays, probably a little bit like we heard about Ben, like just kind of on the cusp, but around eight  months presented with infantile spasms. And he had a chromosome microarray, I'm showing you here a karyotype, but that microarray  showed 47, actually XYY, that's a typo, but so he had an extra Y chromosome. But this is about the time that I saw him and said that this is not the reason that  this child has infantile spasms and developmental delays. So we were lucky enough to have a large gene panel at the time available to us. And we did a 327 gene panel.  It came back with four variants of uncertain significance, but we did follow up testing and the ones we thought maybe there was a chance it was the cause, but all of them were  inherited and we thought we don't have an answer yet. So over the next few months he had a lot of additional testing, metabolic, imaging, and other, all of which was non-diagnostic.  And so when he was about two, whole exome sequencing was more readily available and we decided to send that test.  And within two months, which was fast at the time, we got a diagnosis with a de novo MEF 2C variant in this child.  And we gave the parents the result and by the time we actually saw them back in clinic  they'd connected with a Facebook group and were delighted to have a diagnosis and a connection. Interestingly, I went back because I said, MEF 2C, this is a known epilepsy gene.  Why did they not report this on the panel? Turns out it was added to the panel just between the time that we sent the panel and we sent the exome.  So it actually was not, again, it was not on the panel. So there's no reason that we would have detected that variant. Another reason to do exome first.  This is one other example of a young boy who had infantile spasms and developmental delay at about five months of age with progressive microcephaly.  The neurologist actually sent the genetic testing first and did a gene panel that was available at the time, probably the same one that Ben had, 53 genes.  Also did a mitochondrial panel, all negative. He was continuing to get worse, so they convinced the insurance companies to pay for exome sequencing only in the patient.  And he had one variant of uncertain significance in a gene that was not known to be an epilepsy gene, but it was reported. And eventually we said, well, this is all we have.  We need to know, let's persist, and eventually sent the parents for exome sequencing. And that VUS was de novo, but we still didn't know if that was the cause.  And unfortunately, he passed away at four years of age. I then saw his brother who presented at about 10 months of age, not as severe as his older  brother, but he was clearly having some delays. He had microcephaly and eventually developed some seizures.  And in 2016, so two years later, I said, I just think we need to test, we just need a whole exome in the brother.  They probably have the same thing, but this will give us more information. And so we did. And they came back with a diagnosis in the brother. I said, well, that's odd.  But it was a recessive disorder with two variants in the gene Tango2. So we said, again, his brother must have the same thing. So let's go back.  Can we ask them to go back and look at the brother's exome sequence? And indeed, he did have the same thing.  But the reason it wasn't reported two years earlier is that Tango2 was not a known disease gene. And so it's new knowledge that contributed to this diagnosis in 2016.  As you can see, this was published in February of 2016. So the brother, we were able to take advantage of that knowledge.  And the second reason probably is that one of the mutations was actually a large deletion in Tango2, which at the time the first exomes were done, they probably were not analyzing  for copy number variants. So they may have found one of the variants, but probably not the other.  So we had the advantage both of technological and analytic improvements in the test, as well as improvement in knowledge about which genes cause epilepsy.  So again, another example of why to continue to go back, go back to the testing company, do the new test, because there's always new information contributing to your interpretation.  So I just want to talk really briefly about some emerging genetic testing approaches and options.  This one is actually today a clinical test, so less emerging than actually available. But methylation profiling or epi signatures have become a very useful, I think, genetic  or epigenetic test. These signatures have been developed for a large number now of genes that cause primarily neurodevelopmental disorders.  So I would say today they're less useful for the developmental and epileptic encephalopathies,  but they're now working on all of those genes to see if there is a methylation signature. And what I mean by methylation signature is that you do not a sequencing test, not a copy  number variant test. You actually do a methylation array that looks at methylation of the DNA across the genome. And you can see patterns. I give the example here of CHD2.  So children who have a mutation in CHD2 that's disease-causing have a methylation pattern when you look at the results of their test that looks very different than children who  don't have a mutation in CHD2. And one of the nice ways that this test can be used is that if you have a child who has  a variant of uncertain significance in CHD2, you can send the methylation profiling. And if they have the signature for CHD2, then that's probably a pathogenic variant.  If they don't, it's probably a benign variant. So that's shown here. I don't think I can point, but shown here in the aqua-colored little cluster of patients there.  So they all have the same methylation pattern. They all have CHD2 mutations. All of the pink patients there across the top do not have CHD2 mutations.  The one little blue bar in the middle there is a patient with a CHD2 VUS. Clearly falls in the group without CHD2 mutations, so it's probably a benign variant.  And you should keep looking for the genetic cause in that patient. It can also be used if you've never found a variant in a patient.  You can send this test, epi-sign test, and ask, is there a methylation signature for any of the genes that we know have signatures?  And sometimes it can actually be used as a diagnostic test. We are also using DNA methylation analysis to look for methylation aberrations that cause epilepsy.  So this is something we're doing purely on a research basis right now. But as you can see, all you need to know is that one red line is a patient who looks  very different than all of the black lines. And that patient has a difference in methylation at a specific gene in the genome that may  or may not be causing abnormal expression of that gene and causing epilepsy. RNA-seq is another approach that is available clinically in limited situations.  This is where we're not looking at the sequence of the gene, but we're looking at the expression of a gene.  And so you can look at the amount of the gene that's expressed or whether there are splicing changes in the expression of a gene.  And this is often used if you have either a genetic change that you can't interpret but looks like it might affect the splicing of the gene.  You go and test whether it affects the splicing of the gene. You can also use it if you have a recessive change in a gene that causes a recessive disorder,  but you've only found one variant and not the second variant. But if you go look at the expression of the gene and it's clearly decreased or altered,  this may lend support that there's probably a mutation, a second mutation that you just haven't found yet.  A caveat is that the gene must be expressed in the tissue that you're doing the expression analysis in. So often this is done in, the expression analysis is done in blood.  And we know that when we're thinking about epilepsy, this is a brain disorder and not all of the genes that cause epilepsy are expressed in blood.  So again, another important thing to think about ahead of sending the test. And then we've already heard about whole genome sequencing, where you're sequencing  not only the DNA that codes for genes, but you're sequencing 100% of the DNA. And this is available clinically in some situations.  It's most often used today, I would say, in rapid genome sequencing in NICU patients of various presentations.  The important thing about genome sequencing is that you can get different types of variants you don't get with exome sequencing.  So you can find changes in the non-coding or regulatory DNA, though I would say the interpretation, that's my little shrug there, it's much harder in non-coding DNA than it  is in exons. You can find repeat expansions, structural variants, rearrangements, depending on which sequencing platform you use.  It's still a little bit more expensive than exome sequencing, but that is rapidly coming down and there have been some announcements in the past month or two even of new approaches  that make it much cheaper. But importantly, these often require follow-up studies to actually determine whether the variant you find is pathogenic or even real.  And I will just say also that most of the genome sequencing, when it's sent clinically, still the analysis focuses on just the 1% of the DNA that codes for genes, but a genome  is a better exome. Actually, it can detect things a little bit better than exome sequencing, even in those same regions.  And then the second whole genome sequencing approach that is emerging and not yet into the clinic, but I think will be certainly in the next few years, is long-read sequencing.  So most of the genome sequencing that we talk about and probably that Dr. Patel used is what we call short-read sequencing, uses technology from Illumina.  So you're sequencing little pieces of DNA at a time, and then you have to stitch them together, right, and map them back to the genome.  Long-read sequencing means just that, you're sequencing longer pieces of DNA at once. And so what that does is it allows you to look at big stretches of DNA and see things  like deletions in the middle of it, or repeat expansions in the middle of it, or rearrangements that you might not be able to detect by short-read sequencing.  So this is, I've just seen a couple talks recently, actually, where long-read sequencing seems to increase the diagnostic, or you can come up with a diagnosis in about 5 to 15%  of patients where exome or genome sequencing failed. So it's definitely going to, again, chip away and help us get diagnoses in a few percent  of patients where we haven't been able to do that before. And then these are just ongoing, this is ongoing research that we're actually doing in my lab,  but many labs across the world are doing similar approaches. So thinking about our unsolved patients, we're happy to hear from you if you have DEE patients  especially who you've been unable to get a genetic diagnosis. We're taking approaches that I've already talked about, whole genome sequencing, including  long-reads, RNA sequencing, in this case we're trying to do it in organoids developed from patient cells so that we can get a kind of pseudo-brain expression, and then methylation  studies that we're using in a variety of ways. And then I just want to touch on, so most of what I've talked about is really available  for the severe early-onset pediatric epilepsies, but there have been important advances also in testing for focal and generalized epilepsies, and many of you are aware of this.  So genetic testing is certainly important in the focal epilepsies, especially in familial forms or specific phenotypes where you know that there's a genetic cause.  More and more it's being used in presurgical workup, and then testing tissue from surgical resections has led to some really nice results.  These are just a handful of the genes that we know are important in focal epilepsies, many of them in the mTOR pathway.  And then a very interesting, I think, emerging result over the last few years has been looking  specifically at tissue and finding that there are mosaic variants, so similar to, you can think of it as like a tumor where you have changes in the genome as the tumor develops.  In these focal cortical dysplasias, you may find a genetic variant that probably actually drove the abnormal development of that focus. This requires tissue.  It requires specific sequencing approaches, but there are clinical tests available for this. And then there are other approaches to say, well, what if we suspect it, but we don't  actually have brain tissue? Are there other ways to get at this? And there are some nice results emerging for using CSF, for example, or by using EEG electrodes  and getting DNA from those tests to actually test that tissue that comes off of there. GGE has been a tough nut to crack, even though we've thought for a long time that it's the  most genetic epilepsy. There are rare familial causes where we know there are specific genes. And as I alluded to earlier, there are copy number variants that are significant risk  factors for generalized and other epilepsies, including 15q13 deletion. So you might consider if you have a family, especially if they have kind of GGE+, so plus  autism or intellectual disability more than you expect, you might consider this test. The other important approach that is emerging that you may have heard about is something  called polygenic risk scores. So polygenic risk scores is not clinically available or useful for epilepsy yet, so I'm just going to put that out there. But we're getting there.  So polygenic risk scores looks at lots of different genetic variants that have a very, very small impact on risk. And this is from a genome-wide association study.  Each colored bar represents a chromosome, and what you need to know is that all of the dots above that red line represent a significant difference.  So variants that are more frequent in patients with epilepsy than patients without. But they're present in patients without epilepsy, so they're clearly not highly penetrant.  But we can start to take combinations of these and say, well, what if a patient has 12 of  these risk factors and they add up to a significant enough risk that it might be contributing to their epilepsy?  So these polygenic risk scores are being developed for different types of epilepsy, for different ethnic backgrounds, which is important.  So these tests have been done in white European individuals, and the results don't necessarily apply to other population groups. But we're getting there.  And there have been studies published now on polygenic risk scores in various types of epilepsy.  And this is just not for you to digest, but we put this table together for a chapter that we wrote recently with current approaches to genetic testing, emerging approaches to  genetic testing, and what they detect and what they don't. And as you can see, it's complicated.  But we're hopeful that as we get to more long-read-hole genome sequencing and tests that can detect  multiple different types of variants, that we can narrow in on the tests that we need to do to detect the most variants with the biggest impact in our patient population.  So to finish up, genetic testing has clearly had an impact on clinical care, especially in the pediatric setting. So who should you test?  I would argue in pediatrics, you should test most of your kids who present with epilepsy, but especially those with severe early onset and or epilepsy associated with intellectual  disability, autism, or other syndromic features. Rare familial cases in GGE and focal, as well as rare kind of really specific phenotypes where you know there's genetics involved.  Early diagnosis is key. I would argue that exome is the best test that we have readily available today.  Genome sequencing panels are great, and chromosome arrays should still be considered, especially if those other tests are negative.  If you have a very specific phenotype, like Angelman syndrome or Rett-like syndrome, there are specific panels for that that you can do.  So you can certainly consider that as opposed to exome first, although exome will pick up all of those as well.  And then as we move into the future, you know, whole genome, I think it's really exciting. So genome sequencing is online and becoming increasingly available.  All of these other tests that I briefly touched on are moving into the clinical space quickly. And like I said, hopefully we're going to get to a point where we may have one test  for all or maybe just two or three tests for all instead of that 10-year diagnostic odyssey that some of your patients have experienced. And I'll stop there. Thanks.

Video Summary

Heather Mefford, a clinical geneticist, discusses the role of genetic testing in diagnosing pediatric epilepsy, particularly severe cases. Mefford outlines available genetic test types including chromosome microarray, gene panels, whole exome, and whole genome sequencing. She emphasizes that genetic diagnoses can end lengthy diagnostic journeys, help with prognosis, and inform treatment or recurrence risk. While recent advances allow genetic diagnoses in up to 50% of pediatric epilepsy cases, there remain unsolved cases potentially due to untested genes, undetected variants, or unknown genetic causes.<br /><br />Emerging tests include methylation profiling and RNA sequencing which aid in difficult cases. Chromosome microarrays can identify copy number variants, while sequencing focuses on DNA sequence changes. Whole genome sequencing, including techniques like long-read sequencing, could improve diagnostic yields further. Mefford notes that ongoing research aims to solve unsolved cases and stresses that genetic testing should be prioritized for patients with severe and early-onset epilepsy, leveraging whole exome sequencing as the current most effective method. The talk reflects optimism in reducing diagnostic odysseys with future technological advances.

Asset Subtitle

Presenter:

Heather Mefford, MD, PhD, is a physician scientist and Member in the Center for Pediatric Neurological Disease Research at St. Jude Children's Research Hospital. Her research is dedicated to gene discovery in pediatric neurological diseases, focusing on epilepsies. Her work helped define the genetic landscape of rare developmental and epileptic encephalopathies using cutting edge genomic technologies. Using engineered and patient-derived cells, her group develops cellular models of genetic epilepsy for characterization and testing potential targeted therapies. She was a co-PI for several of the Epi4K consortium projects and co-chairs the ClinGen Neurodevelopmental Disorder Clinical Domain Working Group and Epilepsy Gene & Variant Curation Working Groups.

Keywords

genetic testing

pediatric epilepsy

whole exome sequencing

chromosome microarray

RNA sequencing

diagnostic odyssey

methylation profiling