With California's embrace of cell-free DNA screening and the increasing use of gene sequencing, pregnant patients have questions about what they can learn from these tests and what to make of the findings. In this talk, perinatologist Annalisa Post, MD, discusses the information that various prenatal genetic tests can and can’t provide, and how to assess accuracy, offering a valuable tool for individualizing results. She also explains the differences between whole exome and whole genome sequencing and when rapid whole genome sequencing – the latest option – has particular value.
So, this was set up to talk kind of about whole exome and whole genome sequencing since we'll talk about this is becoming a lot more common, so I think it's really great just to kind of know where we are and what's available now, but I just decided to do kind of a review of genetic screening and testing cause I think it's always worth going over again and there's been just massive changes in the last year that we're all still trying to catch up to and changes keep coming, so. I will go through kind of a just a broad overview. And I have a genetic counselor who's just next door, and she said I can text her if anything comes up that I can't answer, if there's specific questions about like California prenatal screening program or things like that, so I'm happy to call her in if, if she can help us. Um, so I'm gonna go over kind of screening versus diagnostic testing, go into screening a little bit, and then go through the diagnostic genetic testing options and where we are now with what we call next generation sequencing, which is really the whole exome and whole genome sequencing. And I mentioned my genetic counselor who gave me some of these slides as well. So, first off, I just wanted to review this analogy that I use a lot with patients, and I think it, it's kind of helpful when thinking about what genetic testing can or can't tell us in different kinds of genetic testing is this kind of books analogy, and You can describe the Genetic information, so DNA is a series of books like a shelf full of cookbooks, and each, each chromosome is essentially one cookbook, and altogether it tells us all the information about how to make a human baby. And An Euploid is would be a missing book in this example, or an extra book, and this is something that will in more detail talk about, but that's gonna be what a karyotype will tell you, you know, is there a book missing on your bookshelf essentially, or an extra book that shouldn't be there. Then if you were to open up those books and zoom in and look at the pages of the books and see is there any page torn out or some extra pages that got stuck in that shouldn't be there, that's thinking about what we call copy number variants, and that's what a microarray will detect, and that's like most of the things you think about when you think about specific genetic conditions that have some sort of numbers attached to them, like the 22Q 11.2 deletion, which is the George syndrome. And then when we talk about zooming in still further, sometimes just single gene changes or single nucleotide changes or a small number of nucleotide changes can result in enough of a change of function of a protein to cause a disease, and those are what we call single gene conditions, so there's just countless single gene conditions, but many of the skeletal dysplasia, spinal muscular atrophy, Tay-Sachs, tons and tons. So to find this, you actually cannot detect these even with a microarray. So when we're doing these kinds of testings, we're doing what we call sequencing, actually looking at the letters arranged, and most targeted sequencing is just let's look for this specific condition or this specific few conditions, and that's what all sequencing is going to be, so single gene or maybe sequencing larger portions of the whole genome. So then to go into genetics and screening versus diagnostic. I know we know this, but screening is anything that comes from mom's blood. It's not diagnostic, meaning it's not the gold standard. We're not actually able to say this is what this fetus has, and this means that abnormal results should In theory, always be confirmed with a true diagnostic test. Now, there's exceptions to that. Sometimes you have a fetus that has this clear picture of all these anomalies that go along with trisomy 18, and you get trisomy 18 on cell-free DNA. Do we technically have to confirm it with amniocentesis? You should, but in that case with the ultrasound findings, we'd feel pretty confident, but it's still important to remember that. Diagnostic test is anything where we actually get the fetal cells. So we have an amniocentesis, we have CVS or even intheria, percutaneous umbilical blood sampling, and then all that range of testing that we can do are all considered diagnostic tests. So, of course, serum screening has changed so dramatically in the last year with the move from the California prenatal screening program, also called serum screening, sequential screening or integrated screening, all these different ways it was cut and doing the NT and first try quad screen, all that stuff. And that was an Andyloidy screen but also gave us some other information, sometimes good, sometimes just more confusing, like a low pap A level or a high AFP level, um, but now, of course, in California and in many other places, we've moved towards not using the serum screening and instead doing self-free fetal DNA screening as the first line test, not just for high-risk patients like it was just a year ago, but now for all patients. And of course this goes by many different names. I always write it because I'm Annoyingly pedantic like that as CFDNA, self-revealed DNA, um, but It's also we all the time call it NIPS NIPT. I put this little asterisk here cause I think NIPT is becoming the most common way, and again this is totally pedantic, but it's not actually accurate because tests are actually supposed to only refer to diagnostic things. It's technically a screen, but I agree that calling it an IPT rather than NIS is better, so like, I don't have a problem with that. It was always weird when we were calling it the NIPS test, and I'm not a fan. So, for what, for better or for worse, it's usually called NIPT cell-free DNA, whatever, all the same. So, as we know, we're taking a sample of mom's blood in this case. In mom's blood are fractions of her own DNA. All of our blood has tiny bits of DNA from cells that have gone through apoptosis and are just being broken down, but also placental cells. So you get this mix and then it is tested in lots of different ways that are. Constantly evolving, but we're essentially looking for genetic information and looking for is there extra of a chromosome, missing parts of a chromosome, and we're increasingly able to do more testing. Can we actually look for single gene conditions or microdeletions, and that's expanding rapidly. Certainly, that'll become mainstream in the future as well, but for right now, it's still most used just for annuloidy, the, the standard 1318, 21. I put this in and I'd like to just leave and go to this website briefly cause I think this is a really important exercise to do. Self-revealed DNA screening is a screening, like I said already. The predictive values, so you get a result and how sure are you that that is correct. It varies because positive and negative predictive values vary with the prevalence of the disease in the population. So, if you are Getting a result that's positive for Down syndrome in a 48 year old person. There's a higher prevalence in that population, just because we know the risk is a lot higher in that population, so that makes it much more likely to be a true positive. If you do a test for trisomy 13 in a 21 year old, That's a very low prevalence of trisomy 13 in that population, and therefore a positive result, even though the test is still the same test, is actually much less likely to be real. This is just a statistical reality that's really important. So, there's this website that's extremely helpful, the perinatal Quality.org. And we use it all the time if we're specifically counseling people about genetic results from, from NIPT as it says, um, the specific numbers you shouldn't ever cling to because they make some assumptions about the test sensitivity and specificity, but it really is, I think, pretty dramatic. So I'm just gonna pull it up just to share this and go through the exercise real quick. So, you go, this is what you get when you go to this website. Usually we're talking about the positive predictive value, so you get an abnormal result, what's the likelihood that's real. And let's put in that example I mentioned. So let's put in trisomy 21, and let's put in her age, and let's say 40-something, so we're gonna say 45. These numbers here, the sensitivity and specificity, that is the test performance, that's not the same as the positive negative predictive value. So it sets these. You can adjust them. Sometimes if you look at the result, it'll tell you, if you look, you can find like the specific lab and you can change it. Most of us don't bother, but you should know that changing that will change the numbers a little bit. One of the reasons why this is not totally accurate, but it sets it as kind of a general test numbers that makes sense for NIPT. Now we hit calculate. So, 45 year old positive trisomy 21 on NIPT 98% chance that that fetus actually has Down syndrome. OK, and that's kind of what I feel like we Expect for these tests. When you get a positive result in a person who's high risk, it's probably true. But let's go ahead and change it. So, Let's go to trisomy 13. Trisomy 13 is interesting. So the test is best for certain things, and it's best for trisomy 21 specifically, in part because that's the way the test is designed, but also because different chromosome problems act differently, and trisomy 13 positive results, there's this much higher chance of something called confined placental mosaicism. Meaning that, yes, there's trisomy 13 in some cells of the placenta, but not in the fetus itself. So that is one reason why sometimes the results are not right, and that's much more true for trisomy 13 as compared to trisomy 21, medium true for trisomy 18. But also 202013 is just way, way, way more rare, so that's gonna affect it as well. So if we put in a 45 year old, still leaving her 45 with 2013, let's see what that looks like. OK, they don't have enough data. Let me make her 40. I did this earlier with a 40 year old at work. So, it's only 50/50, which I think is pretty remarkable. Now, again, if this fetus has holoprosencephaly and severe growth restriction and a heart defect, OK, it's probably trisomy 13, we can adjust to this mentally and, and counsel patients, but I think this is important. And then the last part I'd like to do is let's go in and now make her. 20. And calculate, cause this is just, to me it blows my mind, 6% chance that it's real. So this is a really important thing to just keep in mind, especially for non-Down syndrome and especially for younger women, is that It feels like a really good test. It has really good negative predictive value, and we could go through that, but just across the board when you get negative results, it's very, very, very, very, very, very likely to be negative, but it does not perform the same way for positive results, especially in these situations, so I just really wanted to go through that exercise cause I think it's really dramatic. OK, moving on to what this talk is actually about. So, I talked about uh screening for annuity primarily. As you guys know, we can screen for other things with self-refetal DNA, sex chromosome differences, microdeletion screening is becoming increasingly screened for. As of now, those are not part of the state program, as you know, so patients would either have to bill to insurance or do a self-pay, um. In the future, these, at least some of them are going to become incorporated. There is a plan in the next year to incorporate the sex chromosome differences, so that will happen. I don't know about the timeline for microdeletion screening, but again, as we get more data, as things expand, these are going to be incorporated more in the future. So let's move towards diagnostic tests, so we actually need to go in and get a sample of something fetal for these tests. First, just to start, very basic karyotype. This is what's been around the longest. This is literally taking this jumble of one cell's chromosomes that you'd see under the microscope on the left here. Doing arts and crafts and lining them all up so that they match with their little labeled bands, you can identify which is which, and then looking to see is there an nuploidy, I'll go back to that, which is an extra chromosome or a missing chromosome here there's tetraploidy X, so extra again, this is the books, right? So there's two extra cookbooks of the chromosome X here. They could also detect if a chunk is missing, so this is actually pretty subtle, but there's the top band of this chromosome 5 is missing here on the short arm, so it can detect large deletions or duplications. And then it can also detect translocation, so in this case, part of chromosome 9 has become stuck onto chromosome 22. So, these things we can see with karyotype, and I'm gonna go back a little bit, but Preotype takes a little while because you can't just do it on your sample. You have to take the cells you get, make sure they're alive, and culture them, and they actually do multiple cultures, cause they need to get cells, enough cells, they need to get them to the correct mitosis stage of metaphase so that they have these organized chromosomes that are about to split but not yet split. And get enough cells so that they have a good representative sample. So it's, it's a process that actually is more labor intensive often and more expensive than some of the newer tests because of all of this. So, moving on to microarray, microarray, we are zooming in on the chromosomes essentially to, we're now moving to the pages of the book. So, these are called copy number variants. Are there chunks missing or chunks duplicated? And this, like I said, is responsible for a lot of the genetic conditions that we see that are not annuloidy. It doesn't require culture usually if you have enough cells. It doesn't require live cells, so it's faster, and it's also the preferred test if you have a demise because it doesn't matter if the cells are healthy, you just get some blood, you get a sample of the placenta or fetal tissue, and extract the DNA and you can do microarray. So it's faster, it doesn't require live cells, and it's obviously gives you far more information than a karyotype. So, unless you're specifically looking for trisomy 21 to confirm a positive NIPT for example, microarray is often the preferred test. Now, microarray can have results you don't expect as well. We actually have a pretty good study that showed that in low-risk patients with normal ultrasounds, if you just do a microarray, an amnio in a microarray, you'll get positive results in 1.6 patients. Now, some of those are going to be results that are Of uncertain significance. Some of those results are really minor. Those results may be for things that have really variable penetration and are really hard to counsel about, so they're not necessarily all like clearly severe findings, but it's, it's something positive and 1.6% in this study. Now, As time goes on and we learn more about different genetic results, these numbers change because things that were previously, oh, this is a variant of uncertain significance or likely pathogenic, it may later become reclassified as, oh no, that's a normal variant, and so these, these numbers may change over time. If you have an anomaly, obviously much more likely. If you take all anomalies together, there's a 6% chance of positive microarray results, but again, that number varies so much, you know, if, if the anomaly is certain heart defects, it might be a 50% chance, for example, so obviously the results are gonna be higher with an anomaly. So now moving to the main thing we want to talk about today, which is sequencing, so. Some sort of actually zooming in now on the bottom picture, now we're at the letters, right? We're actually looking at the words and letters, are the letters arranged properly, and so this requires what we call next generation techniques where we're actually going letter by letter and literally sequencing the genetic material, and we can do this in different ways. We could do targeted, meaning we're just going to look at a few specific things or one specific thing. Or you can do a much more expansive sequencing of large portions of the genome. Targeted sequencing, like I said, might be single gene, it might be a panel of genes, but you need to have something to guide you. So, maybe this parent is known to have a condition and she wants to know if her fetus has this condition as well. And so you're gonna go looking just for that condition, but she doesn't want to do a full microarray, or in this case, microarray wouldn't work, but, you know, we only want to know this one specific thing. Maybe you think it's skeletal dysplasia, so there's panels that are developed that are a skeletal dysplasia panel that will sequence the genes that we know we may find something on for skeletal dysplasia. There's panels for fetal congestive heart failure, there's panels for like neurofibromatosis. There's, there's various panels that have been developed to sequence the relevant areas. Whole exome and whole sequence, whole genome sequencing take this to the next level. Instead of targeted sequencing, now we're just looking through everything, or at least large portions. So again, it's a nucleotide by nucleotide sequencing of DNA. Whole exome sequencing sequences only the part of DNA that actually gets coded into proteins, that actually turns into things that our body uses, and remarkably, only 2% of the genome is the exome or exons, but this is the part that in general we think of as important, that's the actual portion of DNA that if something's wrong, would be a problem. Whole genome sequencing says let's just sequence the entire thing. We're gonna sequence the coding, coding and the non-coding things. For whole exome, it actually takes an extra step because how do you know if it's an exon or an intron? How do you know if it's, you know, in the exome or the genome, you actually have to have a first step where targeted probes that will attach to the coding regions at the beginning of all of the different exomes or exons are attached to the DNA and then that portion is amplified and expressed, so. It actually takes more steps cause you have to first identify the exome, then sequence the exome. Well, genome, you don't have to identify any portion, you just sequence everything, and that's relevant to how we're using it now to talk about it in a moment. So, fetal whole exome sequencing is something that is relatively new to be used extensively, but really in the last few years, we've really ramped this up, specifically at UCSF, but other places are following suit as well, and it's pretty exciting. So, doing study-based, so just purely research-based fetal hu XM sequencing at UCSF started around 2018, um, and they actually started a study specifically for fetal non-immune hydrops because these are really tricky cases. This is babies with hydrops where it's not because they have anemia, it's not because they have syphilis, we can't find anything on a microarray. But clearly they're very sick and they're born and there's something very wrong with them. So they start doing prenatal whole XM sequencing and we're able to find in this series that was done for a couple of years with patients from all over the country, a diagnostic variant in a third of cases, which is actually really high yield for doing a genetic test like this. So, that's now much more broadly used for specifically non-immune eye drops. Throughout this, they recognize that this also can be useful in other cases where we don't have a clear diagnosis, but there's something that we suspect may be genetic. So at UCSF specifically, we now do institutional full XM sequencing for more fetal anomalies, all with negative microarrays. So, We kind of go do a case by case analysis. Sometimes it's not usually appropriate if it's something that we know not to typically be genetic, it's a mild finding, but for fetal anomalies that are concerning or even something like very severe early growth restriction, we will offer these patients the option of whole exome sequencing. Now, it's traditionally been very slow and that's a major limitation. It used to be said to be 2 to 3 months and that was pretty real. It's gotten a lot faster, but right now, it still tends to take about 4 weeks, sometimes a little less, but it's, it's pretty slow. It's, it's slow, and you have to bear in mind this is after the patient's already had an amniocentesis and a microarray, and so the whole process takes a long time. So using that information to guide. Termination decisions or sometimes even to get that before baby is born, it's, it's still a challenge, but it can sometimes the results can be really helpful in guiding the baby's postnatal care. So we've been doing this and this expansion to fetal anomalies is really just in the last year or two, so it's pretty exciting. We'll do it obviously for babies being cared for at UCSF or moms that are getting their care there, but we also do it for a lot of the other uh community clinics as well who are doing an amniocentesis, say at the Alo Valley Perinatal or whatever. They have sent us some of these samples for exome, uh, to our geneticist there. Um, here's a website here. I, I won't go to it now, but it just has a lot more information about the fetal exome sequencing program at UCSF including references to some of the papers, the one that I referenced, and just, um, patient information as well, that can be really helpful, so it's worth checking out. And then the what we're moving into now is whole genome sequencing. So why might we do whole genome instead of whole exome? Well, like I mentioned before, even though it's more expansive, it's actually easier and faster because we don't have to target, you just sequence everything, so it's actually going to be faster. Could it also be better? Well, there is a very small chance of detecting something that's not in the exons. It's in what we call the introns, the DNA that doesn't get turned into proteins, that just hangs out inside the cell nucleus. But sometimes we think that that may actually be directing. Other things, so that maybe that DNA is important in how the exons get expressed, for example. So in theory, we know it's possible to find something in the introns, but that's very unlikely, and that's not usually the purpose of the whole genome. Maybe in the future as we learn more, we'll find more things in the introns, but it's still pretty rare. So the primary benefit right now is that it is going to be a faster study, and in some ways maybe a simpler study. Because of this, we're actually moving, we just started a program in the last couple of months at UCSF that extends also to other programs again that, you know, we kind of are used to working with and that send samples to us for doing rapid full genome for prenatal patients and postnatal patients. So, this is skipping over microarray and going straight to a whole genome and getting results pretty quickly. The results were still often officially saying up to 4 weeks, but the actual results, once they actually get them onto the sequencer, can be resulted in 3 days, and then they still get analyzed and discussed with the temageneticist, so it takes a little bit more time, but it's very, very fast, which is really important postnatally when you have a baby and you don't know there's something wrong, we don't know what it is, what kind of special care does this baby need, so it's been very helpful already. And now we're also able to do this for select prenatal patients. It's not always appropriate if we think. This is gonna be tricemeatine or Deor or something, you should start with that testing at least now because this is still quite expensive. It's cheaper than it was, you know, doing a whole exome. With the more expensive side of it is like still a Up to $14,000 study, really, really, really, really expensive. These more rapid ones that we've worked to bring the cost down on are still over 4000, but that's actually really cheap for this kind of sequencing, but of course, you should only use it if it's, if it's the right choice to use. So, we usually, for all of the whole XMR whole genome patients, there's a specific consult with one of our trained genetic counselors who specifically does these consents as well as consulting with our geneticist to make sure this is an appropriate patient. But more and more commercial labs are also starting to do rapid testing, so Gene DX is able to do whole exomes, sometimes it's cheaper for the patients, um, and has a pretty rapid turnaround time. They're not doing whole genome officially, but they're definitely moving in that direction, so it's a lot of changes that you guys can expect. To see on some of your patients. And I think that's actually it. I mean, my last slide was just about, I, I touched on this already, but It's, it's kind of confusing. You feel like a diagnostic test, it's diagnostic, it should give you a yes or no answer, but of course it's not always the case because especially as we do more sequencing, we find differences where we don't know what they mean. So, you really can end up with a lot of these in the middle variants of uncertain significance, and like I mentioned, as time goes on, as we do more sequencing and more sequencing and more correlation to different cases. We'll be able to take those variants and put them more into the likely pathogenic, likely benign. And so the numbers get better over time and how we can interpret them. But it's still tricky, and there's a lot of art of genetics that goes into interpreting these results, cause they're not interpreted in a void, they don't just take an exome or genome and say, here are the findings. The findings are interpreted within the context of the phenotype, so they need to know like, why are we doing this? Is it severe growth restriction and a brain anomaly? OK, well then these would make sense because these genes have to do with the way the brain is formed and the skeleton. So, they're interpreting these results and because of that, of course, these are things that we're gonna get better at with time, but we're good enough now that it can really be really helpful for certain patients, so it's very exciting.
