Hello, and thank you for joining us today. My name is Greg Fisher, and I am a human molecular geneticist at Prevention Genetics. I'd like to talk to you today about genetic testing in the realm of autism spectrum disorder and how testing for autism spectrum disorder, or ASD, can have an impact on the care of a patient. And so learning objectives with today's presentation is to describe the genetic diversity behind ASD, or autism spectrum disorder, and the growing list of gene candidates. Explain the clinical utility of genetic testing for ASD and other related syndromes. I'm then going to identify some examples of genes that are associated with ASD that include secondary clinical features that may result in a change in clinical care of the patient. And then finally, observers will recall the clinical utility of genetic testing for ASD and other related syndromes based on testing here at Prevention Genetics. And so autism spectrum disorders really fits in the category of neurodevelopmental disorders. And these developmental disorders have significant degrees of genotypic overlap. But neurodevelopmental disorders themselves are a very common cause of disorders in individuals from the ages of 3 to 17 years old. The prevalence is around 17%. And according to DSM-5, neurodevelopmental disorders include things like intellectual disability, or ID, global developmental delay, communication disorders, ASD, ADHD, specific learning disabilities, and motor disorders. Without excluding DSM-5, cerebral palsy and epilepsy are also included in the context by some groups when thinking about neurodevelopmental disorders. And so I like to think of these conditions as basically overlapping circles, these Venn diagrams, where you can have individuals that have classic ASD, they can have ASD and ID, and they can have ASD, ID, and epilepsy phenotypes. And so for today's presentation, I just want to focus on ASD. So ASD's frequency has continued to increase over the last decade. You can see that in the figure on the right, with the most recent data describing a prevalence of 1 in 54 individuals. In 2015, reports of healthcare dollars spent on the treatment of ASD was estimated to be around $268 billion, which exceeds the cost of stroke and hypertension. This is projected to expand to $461 billion annually by 2025. These numbers might actually be outdated. It could actually be a little higher than that. And at that point, it will exceed the cost of diabetes and ADHD. And so ASD is really a classification used to characterize conditions that are described as having the following features, varying degrees of social impairment, varying ability to communicate, propensity for repetitive behaviors, and restricted interests. Now we know that there's a strong genetic component to autism based on early twin studies that were done several decades ago. And I think it's important to recognize that with a prevalence of 1 in 54 individuals, we all recognize that ASD is common. However, an individual's specific genetic etiology is rare, as it's reported that no single genetic etiology accounts for more than 1% of autism overall. And so when we think of autism, we understand that it is genetically complex and heterogeneous based on the number of genes that are implicated. We can have a single allele that's involved. A lot of times this involves de novo variants, and that is going to be the focus of my talk today. But I think we need to recognize that there's also recessive forms of syndromic autism that have been described. And then as the number of alleles increase and we start to think about common variation, this is really where our understanding of autism starts to decrease. And so it's going to take larger studies and additional scientific work to really understand how common variation contributes to the very complex phenotype that is ASD. And so if we think about single allele forms of autism, we really need to think about de novo variants. And so the first type of de novo variants that I want to talk about with you today are single nucleotide variants, or SNVs. These SNVs arise three times as often on the paternal allele, and there's a 1.3-fold increase in the mutation rate with every 10 years of paternal age. And when we talk about ASD and de novo SNVs, or single nucleotide variants, we really are talking about an enrichment of likely gene disrupting variants. And so these are nonsense, frameshift, splice site, or intragenic, maybe small deletions that really disrupt the function of that protein product. And so a lot of studies that really characterized likely gene disrupting de novo variants have been carried out almost a decade ago, and they're still being studied. And so these studies were really carried out with simplex families, and so families where a proband is diagnosed with ASD but has an unaffected sibling. And so these cohorts were looking at the genetic burden between affected siblings and unaffected siblings. And so if we look at Sanders et al. from 2012, what we found was there's clearly an enrichment of de novo variants. And so we see that with missense variants, and we see that with nonsense variants. We do not see that enrichment with silent variants. Two years later, a larger study was published looking at the same phenomenon and observed the same features, where there was an enrichment in de novo likely gene disrupting variants in individuals with ASD compared to their siblings. And we can see that here with nonsense variants, splice site variants, indels, and frameshift variants. And so overall, this really suggests that these de novo likely gene disrupting variants are of increased prevalence in this group. But copy number variants are also important when talking about individuals with ASD. So in 5.8% of probands, there's at least one rare de novo CNV event that's identified. And if we compare that to unaffected siblings, where we only identified such an event in 1.7% of individuals. ASD individuals with respect to CNVs, the burden is greater regarding the CNV number, so how many CNVs are identified, how big are those CNVs, or the size, and what types of genes are impacted by those CNVs. And so really, we're looking at genes that are important in neurodevelopment or brain function. And so when we look at this comparison, we see there's a three to five-fold higher burden of CNVs in ASD probands compared to controls. And so if we look at the figures on the bottom, if we look at the proportion of samples that have greater than one rare CNV event with respect to deletions and duplications, we see definitely an increase in individuals with ASD with regard to deletions. The number of CNV events that are identified in an individual's genome, we see as well an increase in this with deletions compared to their siblings. And the number of genes that are implicated or impacted by that deletion that are related to neurodevelopmental phenotypes is also enriched in individuals with ASD compared to their siblings. So this next slide is kind of shocking. I think we're all used to thinking if we identify a variant, we then test other affected family members to see if that variant is segregating in the family and use that information to help us identify what's actually happening and is the genetic cause in a family. But with autism, it's a little bit different. So if we look at a simplex family where there's an individual with an ASD diagnosis and maybe a sibling that's unaffected and compare that to a multiplex family, so a family where multiple siblings have an ASD diagnosis, what we see is that the number of de novo events per genome is the same in these individuals. So there's no enrichment in de novo events in a multiplex family compared to a simplex family. But what we find is that in 69.4% of multiplex families, these affected siblings carry different, rare, highly penetrant ASD variants. And so what that means is what I had said earlier is that everyone's autism is slightly different at a genotype level, including in the same family. We see shared de novo events in approximately 20% of sibling pairs. And when we see something like that, we really start to think parental mosaicism. There is some information that we can glean from when individuals might share the same variant that might be responsible for the phenotype. And a lot of times this has to do with the phenotype of the patients. And so one study that was performed in 2015 found that if you looked at a particular scoring metric of behavior called the ADOS score, or the A-D-O-S score, and looking at communication, social, and social communication aspects of this assessment, you could see that there was really no statistical difference in siblings with respect to this scoring system that shared the same variant. In individuals that had very different scores on this assessment, they didn't share the same mutation. And so what that says is that deep phenotyping can really help a provider understand whether two individuals in the same family might have the same variant, or if those individuals may have different variants. And so you may consider performing exome sequencing or large panel testing on each individual that's affected in the family. And so now I'm going to move on to the impact of ASD testing on the clinical care of the patient. I've given you some examples of the variants that are implicated in ASD. Now I want to talk a little bit about how those variants impact care. And so I think it's important to start out by recognizing that with molecular testing for ASD, there sometimes is an aversion to testing in the community. And that's because ASD is seen as a difference and not a disorder. And there's quite a bit of work that's been published to talk about that. Recently, there's been a lot of publications about the neurodiversity movement. And I think that some individuals who go through life with an ASD diagnosis think that when someone says there's a risk of ASD, it has somewhat of a negative connotation. And so as a result, there's a fear of intervention. There's a fear of intervention. Well, I think it's important to always remember that genetics is not the complete story. There's always an environmental component to ASD. However, that being said, there are many benefits that can come from genetic testing for ASD. The first being that ASD and its related conditions are often part of a broader medical condition. And those broader medical conditions can have clinical consequences beyond that ASD phenotype. One report in 2016 described that comorbidities occur in approximately 70% of individuals with an ASD diagnosis. So active surveillance and early intervention with this genetic knowledge can allow for certain interventions to occur before any symptoms actually develop in the patient. A genetic ideology might implicate a specific biological mechanism. And so, for example, potentially pharmacogenomics. If you understand the molecular mechanism behind an individual's ASD phenotype, you may be able to utilize existing pharmaceuticals that are optimum for treatment of certain behaviors. Variants also are associated with specific cognitive and behavioral profiles. And so with that information at hand, you may have more informed information to direct your treatment. And it might actually even allow access to treatment because of that molecular diagnosis. So there's many families that can't get some of the care or assistance that they need without that molecular diagnosis. And then finally, being able to understand the variants or genes that may be at the heart of an individual's ASD diagnosis can allow them to network with other individuals that have similar genetic profiles. And I think at this point, we're all becoming quite familiar with things like Genome Connect and GeneMatcher that allow this to happen. And so I don't expect everyone to memorize or read everything on this table. But what's on this table are various C and D events that are associated with various neuropsychiatric phenotypes. And it also describes some of the extra neurological features. And so while these individuals may present with ID or autistic features or even ASD, what I want to point out is that these deletions are also associated and duplications are associated with cardiac phenotypes. They're associated with epilepsy and seizure phenotypes. They're associated with kidney and renal abnormalities and other features including eye abnormalities, obesity, recurrent ear infections, muscular hypotonia, ataxia, and hearing loss. The same is true with genes associated with ASD phenotypes. And so the same holds true with respect to cardiac abnormalities, epilepsy and seizure phenotypes. Remember that overlapping Venn diagram that I talked about in the beginning of the presentation. Other features that include recurrent ear infections, obesity, impaired vision, hypotonia, cryptarchidism, and gastrointestinal symptoms. And so all that being said, I want to give you two examples today of variant events that when identified led to a change in clinical intervention. So the first one that I want to talk to you about is 17Q12 deletion. So this is a highly penetrant CNV and is one of the 10 most recurrent pathogenic deletions in children with an unexplained developmental disorder. So these children present with ASD, schizophrenia, ID. They present with macrocephaly and characteristic faces. Some of those features are displayed in the figure on the left. And they also present with mood anomalies. These individuals also present with prenatal echogenic renal calluses, prenatal hydronephrosis, unilateral or bilateral kidney cysts, and then maturity onset diabetes of the young type five or moody. So as a result of identifying the molecular cause in these individuals, their revised clinical intervention can include consideration of the nephrotoxicity and weight gain of medications that are associated with mood stabilization. Considering moody, monitoring glucose levels throughout their life, and considering some of the kidney abnormalities, monitoring renal and endocrine function throughout their life. The second example that I wanna talk to you about today is 15Q13.3 deletion. This deletion has variable penetrance and some carriers are even phenotypically unaffected. However, it's one of the most common recurrent deletions that's associated with schizophrenia. And this deletion encompasses CHRNA7, which is a nicotinic cholinergic receptor or an NCHR. So these individuals present with ASD, schizophrenia, ID, seizures and epilepsy, and minor facial features, but extra neurological features can include rage outbursts, psychosis and aggression. And so with identification of this deletion, some of the pharmaceuticals that can be used to treat some of these neurological features can be tailored to the genes that are disrupted. So galantamine is an allosteric modulator of NCHR. So that nicotinic cholinergic receptor. And so it's been shown that this drug can ameliorate some of the rage outbursts. And so as a result of identifying this deletion, behavioral management can be tailored and you can have an informed pharmacotherapy that can result in different drugs knowing that this gene is disrupted. So next I wanna move on to some case examples of cases that we've seen here at Prevention Genetics. And so the first case that I wanna talk to you about is a nine-year-old female whose provider ordered a singleton intellectual disability and epilepsy and autism panel here at Prevention Genetics. This is also known as the IDEA panel that includes over 2000 genes. Later there was parental follow-up of the variants that were initially reported for this patient. And so this patient presented with delayed milestones in childhood, autistic features, retronathia and micronathia, intellectual disability, bilateral ptosis, recurrent ear infections, and eczema. A previous CMA was negative, but it was noted in the clinical notes that a unifying diagnosis was unknown based on the current presentation in the patient. And so we identified a heterozygous de novo missense variant in ACTB based upon follow-up parental testing. And pathogenic variants in ACTB are associated with autosomal dominant Baritzer-Winter syndrome. And so individuals with this condition present with intellectual disability, metopic ridging and trigonocephaly. 100% of individuals have bilateral ptosis and hypertelarism. They can have short stature, contractures, some cardiac abnormalities, and coloboma. The spectrum of severity varies, but furthermore, these patients, as I mentioned, can have some cardiac abnormalities. And so congenital cardiac defects are actually identified in approximately 33% of patients. Those can include PDA, ventricular or atrial septal defects, bicuspid aortic valve or aortic stenosis, and mitral valve and tricuspid regurgitation. They can have kidney abnormalities and more rarely B-cell lymphoma and can benefit from speech therapy and physical therapy. And so as a result, this individual, they did potentially identify a cardiac defect. And so a screening was then carried out to monitor that cardiac defect. So there's an example of a child that also had genetic testing carried out on the right where they did identify a cardiac defect, a minor one because of ASD testing. It's a very good story, so I would encourage everyone to follow the link if they have the time and watch the video. So the second case that I wanna talk to you about is a 12-year-old female whose provider ordered a singleton analysis, again, with the Prevention Genetics IDEA panel. This patient presented with short stature, moderate intellectual disability, global developmental delay, autistic behavior, some hip subluxation, pain and sensitivity, and strabismus. There was some previous testing, which was all normal, and a previous CMA did identify a duplication with a nested deletion at 8P22. And so what we identified was a heterozygous frameshift variant in ADNP, and ADNP is associated with autosomal dominant Hell's mortal van der AA syndrome, or more easily referred to as ADNP syndrome. So these individuals also present with ASD, intellectual disability, some motor and speech delay, seizures, and various dysmorphic features. And so some of the other actionable clinical intervention that can occur because of an ADNP diagnosis includes screening for cardiac defects, but this one's rather interesting in that there may be some new pharmaceuticals on the horizon that could treat some of the speech defects that are associated with this condition. And so there was a paper published in late 2018 looking at a mouse model of ADNP syndrome and looking at these dendritic spines. So you have your synapse where the cells are communicating with each other, and what was identified in these mutant mice is that mice that have a disrupted allele for ADNP have less of these dendritic spines. And so what they were able to assess was these ultrasonic vocalization calls of distress, or USVs. And so this is almost like an equivalent of mice talking to each other. And so what they found is that mice that had a disrupted allele of ADNP actually had less of these chirps that were occurring. And so they could induce these chirps by temporarily removing the mice from their mothers that would cause stress. And the mice that had a disrupted allele had less of these chirps. And so they were able to show that with treatment with their compound designated NAP, they could actually restore these chirps to greater than actually wild-type levels. So stay tuned with that one. That might be something interesting in the next few years. And so finally, I wanna close and just talk a little bit more about the clinical utility of testing at Prevention Genetics for ASD and other related syndromes. And so I think those of us that work in the field of neurodevelopment and specifically ASD are familiar with the 2013 guidelines that were released by ACMG. And so in 2013, this paper reported that chromosomal microarray and karyotype had a 13% diagnostic yield, fragile X being one to 5%, and then finally single gene panel and exomes being approximately 20%. So recently, there was a really good review of diagnostic yield of these tests in the context of neurodevelopment published in February of this year. And now we're finding that the diagnostic yield of things like chromosomal microarray and karyotype for ID, ASD, and other developmental disorders is somewhere between 15 to 20%, whereas fragile X is more around 1%. And finally, gene panels, exomes, genomes is approximately around 25% or higher depending on the study that you look at. And so I think when looking at different panels that are available at commercial and academic laboratories, what one recognizes immediately is that there's a vast difference in the genes that are included on these panels. So if we look at the summary that was published in 2018, you'll see that we have some panels that have 28 genes or we have some other panels that have almost 2,600 genes. So why is there this variation? Well, I think it's important to recognize the difference between a panel, a large panel, and an exome and a genome. So panels are really those genes that have a well-established association. So those genes have multiple literature references describing their association with ASD, but they're not always carried out in a trio. So you might be missing de novo variants or you may not recognize that those variants are de novo. Whereas a large panel that has thousands of genes has those genes with a well-established association and those genes with suggestive evidence. So they're on the cusp of becoming well-established genes. So as a result, you're casting a much wider net of candidates. At Prevention, we found that these large panels are a really good alternative for clients that are unable to access exomes, whether it be due to insurance limitations or provincial limitations. And these panels also include genes that due to internal laboratory data or other collaborations that laboratories have, they've been able to implicate those genes and include them in their large panels. And then finally, you have exomes and genomes, which are all inclusive and are definitely the goal for all patients. However, we recognize that access and approval is not as easy for these tests as it is for a panel. Exomes, genomes, and large panels, as well as small panels, may include CNV detection. There's some reports that this method is more sensitive than CMA. There was a report in 2015 that found 95.6% of the small CNVs that were detected by a whole genome sequence were not necessarily detected by high-resolution microarray. But that's been five years ago already, so things are always changing. But what I think it's important to point out is that whatever panel you choose from whatever laboratory you decide to go with, it's important that they frequently update that panel because ASD is a very active area of research and there are new genes coming out every month that are implicating or describing new cohorts of individuals. And so at Prevention, our Intellectual Disability, Epilepsy, and Autism, or IDEA panel, is updated twice per year, and the vast majority of our clients order a singleton analysis. In the last few years, we released an option known as Patient Plus where we perform sequencing on the proband and then do targeted testing in each parent to identify whether a variant is de novo or inherited. And we've seen an increase in the number of our clients that are pursuing Patient Plus analysis compared to a trio where we do full sequencing on mom, dad, and proband. If you look on the right in the bar graph, you'll see our overall percent positive result for the IDEA panel is approximately 16%. But you can appreciate that when we have trio testing or Patient Plus testing, we see a marked increase in the number of positives. And so this is because we're able to identify those de novo variants. It's quite interesting to point out that our Patient Plus analysis is nearing trio analysis levels for the percentage that are positive. So this testing approach also allows us to identify those de novo variants. But regardless of the testing structure that's selected for your patient, clinical phenotype is critical. That really helps in analysis of the variants that are identified and being able to really report those variants that are most likely associated with the patient's phenotype. And so I just wanna point out that our IDEA panel was previously known as ASD-ID or the Autism Spectrum Disorder, Intellectual Disability and Epilepsy panel. It's the same genes. It's just an easier name to say. It rolls off your tongue a lot better than ASD-ID. We recently included various epilepsy genes in this panel because of the phenotypic overlap. As I mentioned recently, or earlier in my presentation, we update this panel twice a year. And this does include exome-wide CNV. And so because of this, we've had some providers that are really happy about this testing approach because it gives the full picture of what's happening with the patient instead of doing it in a tiered approach. If just doing CMA, for example, you might stop if you identify a pathogenic finding and miss the other diagnoses. And so we also have a smaller Autism Spectrum Disorder panel, as well as various other intellectual disability, epilepsy and ID panels at Prevention Genetics. And so in conclusion, with respect to the genetic diversity of ASD, there is a higher burden of de novo variants, specifically CNVs and likely gene disrupting SNVs in individuals with ASD. To try to understand how common variation plays a role in ASD, larger cohort studies are necessary and are ongoing to help resolve this. Identifying molecular cause of ASD for a patient can impact their clinical care. So I gave some examples of that today and how those molecular diagnoses can change therapy for that patient. And then finally, the clinical utility of genetic testing. There's various genetic testing options from small panels to exomes and genomes. And there's variability in the genes that are included on those panels based on their reported evidence. Overall, a molecular diagnosis is identified in approximately 25% of cases, but frequent updates of these panels are important. So regardless of laboratory, check to see how often the panel is updated. And so with that, I wanna thank you all for your attention today. And any questions that you have can be directed below at my email address or via phone call. Thank you.

genetic testing

Autism Spectrum Disorder

patient care

genetic diversity

gene candidates

clinical utility

neurodevelopmental disorders