I'd like to extend a very warm welcome to all of the audience for the ACMG Genomics Case Conference. This series has been planned by the ACMG Education Committee. Each month, the conference will rotate among laboratories that perform genome scale testing and the clinicians involved in the cases. Use of all case materials that will be presented have been formally consented for use with the patient or the guardians. Today's case conference is called PEGS, Expanding Exomes to the Underserved, and it will be hosted by the University of California, San Francisco. This activity is supported by an unrestricted educational grant from QIAGEN Bioinformatics and the Ingenuity Clinical Decision Support Platform. As a disclaimer, this presentation is intended for educational purposes only. The presenters today will be Dr. Anne Slovotnick, myself, Dr. Jessica Van Zippel, Marta Sabatini, doctor, Dr. Pierre-Marie Martin, and Shannon Rego. All of us work at the University of California, San Francisco, and are involved in the PEGS grant and in the Genomics Medicine Initiative. The following presenters have disclosures. The planning committee of the Education Committee at the College of Medical Genetics and the staff disclosures are shown on this slide. This session is approved for one continuing medical education credit. To obtain credit, those who paid for CMA will receive an email with questions and a certificate after the case conference. Those who paid for CMA will receive an email with questions and a certificate after the case conference. So now to begin the case conference, the overall learning objectives of this series are to describe the clinical utility of exome and genome sequencing tests, to identify clinical indications for exome and genome sequencing, to list determinants that are used to assess the probability of a variant's pathogenicity, and to elaborate on the importance of pre-test counseling and consent. In particular for this presentation, we would like the audience to be able to recognize the barriers to obtaining genetic testing in the underserved, to describe how representation in variant databases differs with ethnicity, and to discuss variables affecting the clinical utility of exome sequencing. To ask a panelist a question, please type your question into the Q&A box in the lower left corner of your control panel. We will answer the questions verbally at the end of the presentation. And so to start. So we're all aware of the effectiveness of whole exome sequencing. However, in the underserved and underrepresented population, there can be barriers to obtaining healthcare, and in particular, genetic testing. These barriers have been studied in papers describing the testing experience of the Hispanic population at increased risk of cancer. Barriers can exist at an individual level. There are differences in the awareness of testing, the knowledge of what the test entails, and there may be concern and worry regarding the effects of testing and what the process actually means. At an interpersonal level, demographic factors, the acculturation of the individual or family undergoing testing that is influenced by their knowledge of English and the time spent in this country, can also influence test uptake. Cultural beliefs can be particularly important. An individual may face life concerns that conflict with testing. They must just have too much else going on in their life at the time to consider genetic testing. But overall, it's known that a strong desire to help the family can also influence test uptake. Barriers to obtaining genetic testing in the underserved and underrepresented can also exist at the system level, and one of the major barriers to testing in this country is the cost and insurance coverage of genetic investigations. Many people who are offered this testing are concerned at the potential for insurance discrimination should they take up the offer. They may be referred to clinic and not attend due to fear and uncertainty of the effects of genetic testing. They may have concerns about privacy and lack of bilingual services and the preferences for face-to-face interaction. The overall data is still very limited. At UCSF, we're part of the CSER Consortium. That's Clinical Sequencing Evidence Generating Research Consortium II. This is the second round of U01 grants that is funded by NHGRI. One of the aims, very broadly, is to study the clinical utility of next-generation sequencing in underserved and underrepresented minority populations. There are a number of sites across the USA that are using a number of different types of genomic testing from panels to whole-genome sequencing to test a population that varies from pediatric cancer to adults with neurocognitive developmental disabilities. At UCSF, our program in prenatal and pediatric genomic sequencing aims to examine the clinical utility of whole-exome sequencing in 1,100 undiagnosed individuals, 300 prenatal cases, and 800 pediatric cases with structural anomalies and developmental disorders. We are sequencing our population as either trios or duos with biological parents in a CLIA-certified lab and have an expert clinical genomics board to sign out. Our sites are in the Bay Area and include the UCSF Benioff Children's Hospital, Oakland. That's represented by the red dot here. The UCSF Benioff Children's Hospital, Mission Bay, and the UCSF Betty Irene Morse Women's Hospital, represented by the green dot, and also the Zuckerberg San Francisco General Hospital. Our patient population in the Bay Area... Excuse me, I slipped a slide. Our patient population in the Bay Area is actually exceedingly diverse, and at one of our sites, the Zuckerberg San Francisco General Hospital, it's very close to areas with ethnic minorities and underserved populations. So far, the population that we're studying includes patients with conditions that have a likely genetic etiology that have undergone SNP analysis that have undergone SNP array or array CGH with a negative or non-diagnostic test result. At least one biological parent must be available, and we're looking specifically to study conditions with treatment, such as metabolic diseases or epilepsy. We are interested in trialing a rapid exome and recruiting patients from the NICU and PICU. And so far, out of the 36 individuals we've recruited from the pediatric arm of the study, there is a high uptake amongst underrepresented and underserved minority in such that all belong to this category. We will assess clinical utility in those who have a positive whole exome sequencing result six months after the return of results at a follow-up visit. We'll use all clinical evaluations that we have, and scoring will be performed independently by study clinicians who have not been seen, or sorry, have not seen the patient under consideration. So what's known about the clinical utility of next-generation sequencing? This has been studied in outbred populations who are not minorities or underserved. And it's now known that this test is effective and cost-effective. Increasingly, whole exome sequencing is provided earlier in the care of patients so that instead of a tertiary test at the end of a diagnostic evaluation, whole exome sequencing is currently considered almost as a first-line or second-line investigation in a group of children with high yield with non, such as neurocognitive disabilities because of the cost-effective of this investigation. However, in the medically underserved population, this has been much less studied. One study published in Genetics in Medicine earlier this year looked at the possibility for whole exome sequencing to actually exacerbate disparities, as many uninsured receive care at community health centers, and a minority of these offer genetic testing. In the study by Strauss et al., 79 probands were investigated with whole exome sequencing, but all but three were from the old-order Amish or Mennonite population. These were pre-screened by targeted allele detection for the disorders that were common in these population groups, and these were ruled out in the people who were subsequently entered into this study. The indications for testing were predominantly neurological disease. Chromosome microarray was able to find a diagnosis in 9% of individuals, and whole exome sequencing in a subsequent 44%. Interestingly, interest in secondary findings was high, and 98% opted to receive this information. The diagnosis was highest yield in patients who had other family members suggestive of a disenfranchised disorder, and negative whole exome sequencing results were still considered of clinical utility because it reduced the likelihood of a genetic diagnosis. Whole exome sequencing can also yield potential diagnoses with candidate genes. The results in terms of the percentage with a diagnosis were similar to those in the non-medically underserved population, but this population also experiences insurance as a major barrier to care. Interestingly, the standard evaluation for delays cost $19,000 prior to exome sequencing and achieved the diagnosis in only a third of individuals, whereas whole exome sequencing by itself early in care is much cheaper and achieves a higher diagnostic rate in this population. However, there's still little information about the utility of this test, and at UCSF, this is one of the reasons for our interest in this population. So now I'm going to hand over the presentation to Dr. Van Zippel. So I'm going to review the whole exome sequencing laboratory workflow that we have established here at UCSF. So for our specimen collection, for pediatric cases, we're accepting a variety of sample types, including blood, saliva, and buccal swabs. And as Anne mentioned, we're testing both trios and duos using a very standard, what's becoming a standard lab processing of genomic DNA extraction, library preparation, hybrid capture for the exonic region, and then sequencing on Illumina sequencers. In our cohort of patients thus far tested, we do have a majority of trios submitted, so about three quarters of our cases so far have been trios, and about one quarter has been duos. Following sequencing, we're using standard methods to convert the sequencing data into sequencing reads, and then aligning those reads to the human reference genome. And we are still using HG19 as our reference. For the alignment of the reads to the reference genome, we call variants where differences occur between the reference and our sequencing. And we're using a series of filtering parameters to narrow in at what may be the pathogenic variants in the proband sequence for each case. So we're using both technical filters to ensure that we're looking at high-quality variants, and we also rely heavily on population databases to remove common variants, and also use predicted deleterious filters to narrow in on what may be pathogenic variants. So once we have used those filtering techniques, we use a variety of inheritance modes to further analyze the data. And in any typical case, we look at all of these different inheritance modes, including de novo variants, compound heterozygous variants, which would be a recessive model. We also separately look for homozygous or hemizygous variants and also for inherited heterozygous variants. So for the inherited heterozygous variants, we use the patient phenotype to come up with a primary gene list and filter those inherited heterozygous variants with that gene list. Finally, if a patient has consented to wanting the return of secondary or additional findings, we use a specific ACMG59 gene list in the curated set of genes for secondary findings. Once we have our list of variants that require further curation, we use a variety of data sources to inform our analysis of the variants. And of course, at the end of the day, we're using the ACMG guidelines for declaration of variant pathogenicity. So we use that for calculating. As I mentioned, we do rely quite heavily on public datasets to understand population frequencies of variants. And one of the first population datasets that was available was 1000 Genomes. And while in comparison to datasets that are available now, it had a fewer number of individuals, it had a really nice diversity of ancestries in that population. And so we were able to see the frequency of different variants across different populations. And more recently with the release of EXAC, and then again with additional genomes added to the genome AD, the NOMAD database, we have an increase in the number of individuals in these public datasets, but and also maintaining the diversity of ancestries included. And why this is important is that having increased size and ancestral diversity in these public datasets allows for more stringent variant filtering based on the minor allele frequency of the variant. And so we're able to filter out variants that are too common to be causal in a given population or in any population. And the filtering power is really increased by the size and the ancestral diversity. But it also allows for the identification of genes or gene regions or domains that are constrained, where they show, for example, a lower than expected number of loss of function variants in a gene. So we were interested in being able to calculate the ancestral origin of our samples. So we recreated the principal component analysis originally published by the 1000 Genomes, where individuals of different ancestries were categorized into populations based on their genetic SNP data. And so we were able to recreate the same structure. Structure. And sequenced 50 samples of known genetic ancestry. And I know it's difficult to see with the small dots, but we were able to co-localize our samples with the open circles of known ancestries with those of the 1000 genomes. And part of why we're interested in doing this is that genetic ancestry, there's been an over-representation of European ancestry in clinical exome sequencing publications, and of course with our study here at UCSF, we're interested in understanding similarities and differences with diverse or under-represented populations. In a publication out of a large dataset from Decipher of 1000 exome sequencing families, and actually now there's a pre-publication of 7000 or more exome sequencing families. In this cohort, it's about 75% of British European ancestry, and while this is a really interesting and useful population without a recent bottleneck and also with low levels of consanguinity providing some useful understanding of utility in that population, we're really interested in also looking at an ancestrally diverse population at UCSF and further understanding what the clinical utility is in that population in terms of diagnostic yield and also similarities or differences in the contribution of different inheritance modules such as autosomal recessive versus de novo. And so with that, we'll move into three interesting case vignettes from our sequencing. Hi, my name is Marta Sabbadini and I'm one of the genetic counselors here at UCSF and I work both in the biochemical service as well in our whole exome sequencing clinic and effort here. So I want to start by presenting one case, one of our cases, which is a seven-year-old female with intellectual disability and developmental delays who presented with elevated CK and a suspected fatty acid oxidation defect. Briefly, she was born at term to a G1, T0 to 1, 29-year-old mother via normal spontaneous vaginal delivery and with an uncomplicated pregnancy that was remarkable only for gestational age towards the end of the pregnancy. She was a product of a consanguineous relationship with a family originally from Pakistan and the family history was otherwise unremarkable. At the time she was born, she was the first child to her parents and she now has a two-year-old brother who's seemingly unaffected. So soon after birth, she was flagged for a positive newborn screening with an elevated C4 and a concern for short-chain acetyl-CoA dehydrogenase deficiency or possibly isobutyl CoA dehydrogenase deficiency. Follow-up testing was non-conclusive and showed transient abnormalities both in the acetyl carnitine profile as well as urine organic acid. And while inconclusive, it pointed towards a possible fatty acid oxidation defect. So at 10 months, a probe assay in fibroblast was performed, but this was normal. At that point, the family was lost to follow-up. The family returned to UCSF when the child was around age six, when she was referred to neurology for anxiety, frequent falls, and dissociation at school. During her visit, the mother reported that she had frequent numbness spells and when she described this, she said she had tended to have during the spells difficulty walking with pain in the upper and lower extremities, slurred speech, drooling, and sometimes eye rolling back or poor eye contact. The neurology ruled out seizures and she was noted to be normal on neurologic exam, including her tone seemed to be normal. She was noted to have developmental delays and she was still not talking in full sentences at that point at age six. Around age seven, she had an IQ test that confirmed she had intellectual disability with an IQ below 70. Soon thereafter, she presented with an acute episode. So she returned from a trip to her family in Pakistan and she started having progressive weakness to a point that the mother went with her to the emergency room where she was noted to have an elevated CK of over 12,000, mild hyperammonemia, elevated LFTs, but her blood glucose and lactate were normal. She was admitted and during the admission, she was noted to have minimal abnormalities in biochemical laboratories that soon normalized soon thereafter. She had a SNP array with a normal female of 46 XX with multiple region of homozygousity distributed across the chromosome as expected with a history of constant greening. She was also noted to have hypothyroidism, but her cardiovascular assessment by EconiQG were normal. And while brain MRI was discussed, this was never performed. A whole exome sequencing was ordered at that point as a trio test, and this revealed a homozygous non-sense variant in Tango2. Both parents were found to be carriers and this was dislocated on chromosomes in the region 22Q11.2. This is within one of the region of homozygousity that were previously identified by SNP array and no secondary findings were identified. So, Tango2 is a gene that has been recently described as associated with disease. In particular, 16 patients have now been described in the literature with biallelic pathogenic variants in Tango2. These seem to be all loss-of-function variants with both truncating, deletions, and missense variants. Some seem to be recurrent, while others seem to be private. And all of this is consistent with autosomal recessive inheritance, as is observed also within our family. To describe the clinical feature of the Tango2 biallelic variant, kids usually present with developmental delay with regression, and all of them have intellectual disability, often with dysarthria, gait abnormalities, hyperreflexia, and myopathic faces withdrawing. Seventy-five percent of the individuals seem to have seizures. There are sometimes non-specific brain abnormalities identified. But other features are hypothyroidism and cardiac arrhythmias, and a few individuals have sensorineural hearing loss and intermittent exoscopia. This is consistent with some of the presentations that our patient had with dysarthria, gait abnormalities, and the hypothyroidism that was identified during her admission. Interesting, these cases often have acute metabolic presentation with profound muscle weakness and ataxia, disorientation, and often can be even to a comatose state. During these acute phases, they can have transient eschocardiogram abnormalities, which can be life-threatening. They usually have rhabdomyolysis with elevated CKA and LSDs, and they have other metabolic abnormalities. And again, this is consistent with what was seen in our patient. So Tango2, as we said, is a newer gene that belongs to the transport and goal-organization protein family. There are multiple isoforms, and the function seems still to be unknown. Interestingly, of the mutations, the pathogenic variants that have been reported so far, these are distributed across the exons, across the length of the gene, and they seem to be all loss of function. We have three truncating variants that have been identified, a spliced variant and two large exonic deletions. There's also a nonsense mutation that is recurrent that has been, and in vitro study, has been showing to also have a loss of function effect. Of the mutations that have been described so far, two of them seem to be recurrent. One is the missense mutation, the glycine-154-2-arginine, which has a manual allele frequencies of 0.26% in the Hispanic population, but with a total distribution of about 0.02% in the general population. There's the large exon 3-9 deletion, also seems to be more prevalent in the European population. One is not commonly observed in other populations. A few of the other pathogenic variants reported are present in exact, but at a very low allele frequency as expected for a recessive condition, and in no homozygous individuals. The variant we identified, which is a truncating mutation, was not described, has never been identified, is not present in either exact or nomads. So in summary, with this first case, we presented a seven-year-old female with intellectual disability with elevated CK and a previously suspected fatty acid oxidation that was undiagnosed. Whole exon sequencing allowed us to identify biallelic pathogenic mutations in the Tango2 gene in a family from an underrepresented population. Interestingly, had we decided to do a fatty acid oxidation panel, only a couple of current panels do include this gene, as this has been described very recently. In terms of clinical utility, this allowed us to put a targeted emergency protocol in place for her, including the need for ECG monitoring during acute episodes to prevent life-threatening episodes. This will allow to monitor for the three-year-old brother, which seems to be unaffected for now. And very interestingly, as I said at the beginning, this is located in the 22Q11.2 critical region, and it's something to keep in mind when individuals identified to have a deletion in this region also present with recurrent acute metabolic crisis and rhabdomyolysis. So thank you. So this was the first case, and now I'll pass it to Pierre for the second case. My name is Pierre-Marie Martin. I'm a clinical genome scientist here at UCSF. I'll be presenting the next case, a 17-year-old female with neoplanktotic delays. Pregnancy and birth was notable for preeclampsia and high blood pressure in mother. She now has intellectual disability and speech delay. She functions more similarly to a third grader. She has a vocabulary of 13 years old, and she's homeschooled. She also has motor delay, haplotonia, but this is mild. She's able to walk, swim, or horse-ride a few times a week. She has microcephaly, a suspected cortical dysplasia, but no seizure. She suffers from anxiety and has challenges with attention. She also has recurrent tachycardia. She's a tall girl with a marfanoid-like body with BMI under 2%. She has mild scoliosis, and additional clinical features include constipation and a history of hypermarble joints that is not resolved. Monodrome sequencing revealed a heterozygous truncating mutation in MEK-T2 that was not inherited from the mother and possibly the novel, but the father was not available to confirm the status. This mutation has been previously reported in four patients, three children, with intellectual disability and Rett syndrome. This variant was classified as pathogenic in ClinVar and is not in NOMAD. And before I talk more about this specific variant and this case, I'll do a brief introduction on Rett syndrome. So Rett syndrome affects 1 to 10,000 births. It's a severe neurodevelopmental disorder that affects mostly females. Males can be embryonic lethal. And a defining feature is a regression in motor and communication skill after normal early development. Another characteristic feature is a short-speak hand movement and loss of hand function. A typical onset and progression is normal development the first few months to a year of life, followed by a developmental stagnation. That's where the acquired microcephaly is going to develop. And this is followed by a rapid regression. That includes loss of hand skills, speech and communication skills, and the appearance of hand sterotypies. This is followed by a stationary stage, and later on, a motor deterioration that will lead to most of the girls to end up in a wheelchair by then 10 years. 1999 was the identification of variant in MEK-P2 as a cause of Rett syndrome. We know now that variants in MEK-P2 account for most of Rett syndrome, but other genes can be mutated as well, specifically in atypical Rett cases. It functions as a transcriptional repressor, but its biological role in the physiology of the disease remains unclear. In fact, many different mutations in MEK-P2 can lead to different phenotypes, from classic female from classic Rett to atypical Rett. That includes the late onset variant or the preserved speech variant, to mild mental retardation, to even normal phenotype with no symptom in some female carrier. In males, mutations in MEK-P2 are rare because it can be embryonic lethal, but when the males are born, it can lead to severe encephalopathy or severe mental retardation with some Rett feature. In a mosaic context, it can even lead to a classic Rett as seen in girls. One explanation for this viability is the X-inactivation status. When the X-inactivation is skewed toward the wild type, it's going to create a milder phenotype than when it's skewed to the diseased allele. The other explanation is the location of the truncation mutation itself on the protein. The earlier the truncation IP on the protein, the more severe the phenotype, and conversely. And this has been well studied over the last year, with a good correlation between the location of this truncation and the clinical severity and the age of onset. And again, an early truncation is going to lead to a more severe phenotype and an earlier age of onset. And to exemplify these viabilities, I have two videos that I'm going to present to you. One is a case with an early truncation that represents a severe case, and the other would be a late truncation that represents a mild case. So this first case had her first symptom at around one year old, and she lost a few words she had. She developed hand stereotypies, and she was never able to walk. Diagnosed at two years old, never able to walk or have a conversation. And the next case I'm going to show you is a milder case with a later truncation. This is a case that showed a first symptom around four years old, some speech delay. She developed mild hand stereotypies at six years old. She was actually diagnosed at 12 years old. And she has now a mild ataxia gait. She's able to walk, and she's able to have a conversation. What's your favorite thing to do? I like to ride my bike. Is that your favorite? I like to ride horses. I like candy. Do you like chocolate? Uh, no. No? No. What kind of kid doesn't like chocolate? Huh? Not me. So now we'll go back to the mutation we found in our case. So this variant has been found, reported already, in four different patients in three different kindred. You can see in the first patient here that it's a typical presentation with first a variant present in male, which is pretty rare as it can be embryonic lethal, and this variant being inherited from a healthy mother. And usually when the mother is healthy or the girl is healthy, it's suggestive of the X inactivation that is skewed toward the wild type allele. But here the X inactivation was random. So the combination of this random inactivation plus the presence of this variant in a male suggested that this mutation was male. And consistent with that is its location on the late location on the protein. Those two patients had motor delays, speech delays, but they didn't have the defining feature of retinome, which is classic retinome, which is a regression, loss of use of hand-in-hand stereotypes, and they had microcephaly instead of the microcephaly. They also were hypotonic. The second case reported is the same mutation but de novo, and the clinical presentation is very similar to these first two patients with motor delays, speech delay, no regression, no report of hand stereotypical loss of function, and no microcephaly and hypotonia. The third reported case has very few clinical information. It was actually diagnosed as a red syndrome because of hand stereotypes. Our case has the same mutation. It is not inherited from the mom, and we suspect that it's not inherited from the dad either based on those previous reports that suggest that the dad would be symptomatic of such a mutation. The clinical presentation is very similar to this other case with the same mutation with motor and speech delay, no regression, no hand stereotyping, no loss of hand function, and a microcephaly instead of the microcephaly. She also has a history of hypotonia. The X inactivation has not been done to this case, but we suspect that it will be skewed toward the disease allele as it is in the reported case. So to summarize, we have a 17-year-old girl with developmental delay of microcephaly, young fatty, and history of hypotonia. Oligosome sequencing was able to identify the causal mutation for this patient, which is a rare MEK-P2 truncating variant associated with a malatypical MEK-P2-related phenotype. Now I'll finish this presentation by saying that this case exemplified the utility of oligosome sequencing. Because based on the clinical presentation itself, a MEK-P2-related kinase variant wouldn't have been guessed, and molecular diagnostic would likely have been delayed or missed. And now Shannon Rigo is going to present the next case. Hello, I'm Shannon Rigo. I'm a genetic counselor working on the PEGS project here at UCSF, and I'm going to talk to you about our third case today, which was a 10-year-old. Is that working? There we go. This is a 10-year-old male with epilepsy, global developmental delay, and failure to thrive, who was recruited from the Pediatric Intensive Care Unit here at UCSF. This patient was born to a 23-year-old G1P1 female. The pregnancy was complicated by pregnancy-induced hypertension, decreased fetal movement, and IUGR. The patient was born at 36 weeks. Labor was initially induced due to pregnancy-induced hypertension, as well as a fetal non-stretch test that showed fetal heart rate deceleration. Labor was not progressing, however, so they proceeded with a C-section. The patient's birth weight was 5 pounds 1 ounce, which is 21st percentile, after a correction for gestational age. Cryptorchidism and micropenis were noted at birth. The patient was also hypoxic and hypoglycemic at birth. He was in the NICU for two months due to ongoing respiratory distress and a polymicrobial infection that led to concern for a possible immune disorder. Both parents are Caucasian. There is no consanguinity reported, and there's no significant family history. The patient's clinical course to this point is notable for hypoglycemia that persisted after birth. He was treated with diazoxide, and hypoglycemia resolved at 14 months. His epilepsy was diagnosed at four months. His seizures are typically multifocal on EEG, but they have been largely well-controlled on anti-epileptic medication. This patient has significant global developmental delays. He walks with a walker, and he's nonverbal. He has a history of failure to thrive. He's been G-tube dependent since two years of age, and he remains below the first centile for height and weight. He also has a history of duodenal strictures, which have likely resulted in his recurrent cholangitis. He has a history of chronically elevated LFTs, which are likely due to his chronic cholangitis as well. He does have a chart diagnosis of cortical blindness, though it is not entirely clear how this diagnosis was established, so we've asked that he follow up in a pulmonology as an outpatient. He's also microcephalic, with head circumference below the second centile. This patient was lost to follow-up for several years, but has recently been hospitalized for several months with status epilepticus in the setting of enterobacter sepsis and respiratory failure, requiring mechanical ventilation. He's been an inpatient for several months, and he is no longer intubated, but remains hospitalized due to cholangitis and worsening pancytopenia. This patient has had extensive genetic testing and screening. All of it's been uninformative. This includes cytogenic studies. He had a normal karyotype and normal chromosomal microarray. He's been tested for Smith-Lemley opus, cystic fibrosis, allergy syndrome, and PFIC. He's also had a number of biochemical studies, none of which were suggestive of the inborn error of metabolism. And his newborn screening was normal. He's also had MRI findings, which include bilateral malformed hippocampi and thinning of the corpus callosum, a liver biopsy and autoimmune workup that were done due to his elevated transaminases were normal. So this patient underwent exome sequencing while a patient in the pediatric intensive care unit. He was tested as a duo, so the proband and mother as a paternal sample was not available. And exome sequencing revealed a maternally inherited hemizygous truncating variant in the gene RNF113A. This is an eight-base pair frameshift deletion at codon 302. It was reported as a variant of uncertain significance because it has not been previously reported and also because there is limited evidence connecting this gene with the phenotype in the literature to this point. No secondary findings were reported for this patient. In terms of gene level information, the gene RNF113A codes for ring finger protein 113A. This gene is involved in the repair pathway for damage induced by alkylation. It has been reported to be associated with non-photosensitive trichocytodystrophy. And that is just in one single case report that described two male cousins who also had a maternally inherited truncating variant in RNF113A, and that was reported by Corbett et al. in 2015. This gene is predicted to have low tolerance for loss of function variants. Let's go to come back to that. On the variant level, this is a frameshift variant. It is located one codon away from the previously reported variant that was identified in Corbett et al. It is presumed to be rare as it is not present in 1000 Genomes, NHLBI-ESP, or the NOMAD population databases. And again, this variant has not been previously reported. Trichocytodystrophies are a heterogeneous group of multisystem diseases. A hallmark of trichocytodystrophies is sulfur deficient brittle hair, which can be confirmed due to tiger tail pattern of banding on polarized microscopy. It's particularly likely to involve neuroectodermal derived tissues. The phenotype includes brittle hair and nails, ichthyosis, growth retardation, intellectual disability, ocular cataracts, and susceptibility to infections, among other common features. About half of affected individuals have a photosensitive form of the disease. But again, pathogenic variants in ARRA and F113A appear to cause the non-photosensitive form of the disease. Most known photosensitive cases of trichocytodystrophy are caused by pathogenic variants in the genes ERCC2, ERCC3, or GTF2H5, which code for subunits of transcription factor 2H, a multifunctional complex involved in both transcription and nucleotide excision DNA repair. Pathogenic variants in the gene MPLKIP account for about 20% of the non-photosensitive trichocytodystrophy cases. And the function of this gene is not very clear. Thanks to Pierre for his help with this slide. This slide compares the features of trichocytodystrophies in general with those seen in previously reported families that we've been discussing from Corbett et al., who also had a mutation in ARRA and F113A, as well as with our patients in the third column. The first column summarizes the common features of trichocytodystrophies from 112 cases, with mutations in five different genes. The most common features are brittle hair, skin abnormalities, growth retardation, developmental delay, and microcephaly. The second column summarizes the features that were specifically reported in the previously reported RNF113A cases, which again were two affected cousins with truncating variants in this gene. And the third column summarizes the features that were specific to our patients. The previously reported cases in column two share quite a few features with our cases. Both have brittle hair. Neither have the typical skin abnormalities seen in trichocytodystrophy, which are more commonly ipsiosis. But both the previously reported RNF113A patients and ours have cutis marmorata. They also have failure to thrive, developmental delay, abnormal MRI findings, spinatal dysgenesis, and susceptibility to infection. Neither of the previously reported patients with RNF113A truncating variants, nor our patient have the most common ocular abnormality associated with trichocytodystrophy, which is cataracts, but it is possible they will develop cataracts down the road, and our patient will be continued to be followed in ophthalmology. X inactivation was 100% skewed in the female carriers in the previously reported family with mutations in RNF113A. But we've not yet done X inactivated, sorry, X inactivation studies on the mother of our patient. So we don't know how our patient compares in this sense. So to summarize, this case involved a 10-year-old boy with epilepsy, developmental delay, failure to thrive, transient neonatal hypoglycemia, microcephaly, cryptorchidism, and micropenis. He's identified to have a truncating variant of uncertain significance in RNF113A, which is associated with the non-photosensitive form of trichocytodystrophy. I'd point out that this variant was unlikely to have been identified without whole exome sequencing. It is, I believe, only currently available on two panels for trichocytodystrophy. The clinical and personal utility for this family includes no need for further invasive diagnostic testing, as well as now the recurrence risk is known. So I'm going to, this is the last of the three cases, so just to conclude, I wanted to summarize that there are significant barriers that exist for whole exome sequencing, especially in underserved and underrepresented populations. Exome sequencing is effective in terms of achieving diagnosis for these families, and we still have a lot to learn about the clinical as well as the personal utility of exome sequencing in these cases. Our references are here. We'd like to say thank you to the patients and families who we have discussed in these reports today, as well as to our PIs and the many collaborators who have worked with us to make this project possible, and also thank you to the NIH for our funding. And now we are glad to take questions. So we have some questions that have been typed in during the presentation. The first is from Dr. Pao Singh Kharlon, who's a cancer geneticist. And the question is, what would be advantages of substituting exomes with genomes? What are the costs for exome and genome sequencing? This is Jessica Vance, I'll respond to this question as best I can. So I think the real benefits of exome sequencing still continue to be on the relative lower cost for the sequencing itself, since it is about 2 or 3 percent of the genome. The sequencing costs are still much higher for genome versus exome. Of course, there's a few disadvantages of exome, mainly being the slightly less evenness of coverage across even exons in the exome, whereas in the genome it's really more even coverage. So every base of an exon really is covered effectively, more evenly in whole genome sequencing. And I think one thing that we're starting to really understand is it's not just single base pair changes and small insertions and deletions that we need to be able to detect, but we really need to be able to detect smaller copy number changes, like exon level or multi-exon deletions or duplications, and that type of analysis is obviously better done with whole genome sequencing. So we're certainly interested in understanding what clinical utility may be, how much higher it would be for whole genome, and then to make the calculation of whether or not the cost-benefit ratio can push you to doing whole genome versus whole exome sequencing. Thank you. Another question from the same person. For how long do you monitor variants of unknown significance after diagnosis? This is Pierre-Fyammery Martin. So usually when we can't identify a pathogenic variant at the end of our interpretation or at the end of our exome meeting, we identify BUS of interest. And first, the clinician runs them in GeneMatcher to see if there's any other information out there that has not been published. And we have a protocol to reanalyze the evidence in the literature six months after the initial interpretation. All right. So another question still on the utility of whole exome sequencing is from Dr. Wen-Li Cao from IU Health. Can whole exome sequencing replace chromosome microarray completely now? So I think it is important to hold the data available for whole exome sequencing is not truly amenable to analysis in the same capacity of chromosomal microarray. What I think people are really interested in understanding if whole genome sequencing could effectively replace chromosomal microarray and combine that also basically merge it into one test microarray with exome. And I think that I suspect that's on the horizon, but there's still some work to do to really understand the effectiveness of genome sequencing analysis versus the individual tests. OK. Thank you. And then a question from Dr. Celeste Eno from UCLA. As to the exomes being run, you stated that you need a negative CMA. But what about CMAs with BUS because it is rare but could very well be pathogenic? Do you still run the exome? So I think that's a great question. So if we have a variant of unknown significance on the array, we interpret the patient's phenotype, we reinterpret the patient's phenotype very carefully with respect to that variant. And we often still do exome because there can be dual diagnoses in people. It's sometimes hard to know if a CMA variant is causative. Any other comments? So I have a question for Marta. So the case you presented was an early fatty acid oxidation disorder that's now molecularly known but is very rare. In the case that we had a fatty acid oxidation disorder that was presumed but didn't have molecular genetic diagnosis, would you recommend ECG monitoring? So that's a really good question. And I have to say, definitely it does depend on the kind of fatty acid oxidation we are suspecting. This is not something we do, for example, for MCAD, but it's something we would do for VLCAD. So I know that in the cases reported for some people, the abnormalities in acylcarnitine were pointing towards a possible VLCAD diagnosis. So ECG monitoring may have been as part of the monitoring that was recommended. However, first of all, in this case, this is not what was suspected. We were suspected a medium chain or a short chain abnormalities. But also, I think what this case also highlights is the fact that, in this case, the monitoring, while for VLCAD, we usually do yearly monitoring for cardiac abnormalities, in this case, it's really important to actually have continuous ECG monitoring during an acute episode, which is not something that is normally done for all fatty acid oxidation. And another question for Shannon, was hair analysis done in your case? So hair analysis was sent to pathology for this case. And the result was somewhat interesting. There was the tiger tail pattern noted, but very, very, very faintly is what the report said. So we sent it, I think, to try to add more evidence to VUS that we actually really do think is very likely to be the causative of disease in this case. And it seems as though the results of that test do support that this is this patient's diagnosis. Do you think they would have commented on that, hadn't they known what we were looking in the test? The hair? The hair? At the pathology, do you feel like you would have learned about that? I'm sorry, I don't. It's like if we didn't know that we asked for that test specifically because we suspected this condition. Oh, do you think it would have been done otherwise? Yeah. Do you think? Yeah. You know, I have been looking through this patient's record to see if brittle hair was noted prior to the diagnosis, and I couldn't find that it was. So after knowing that this is likely his diagnosis, it has been validated by several of his providers that he does seem to have brittle hair. Now the question from Dr. Eno at UCLA, pre-notal exome question, do you report variants of unknown significance? Do they need to be likely pathogenic or pathogenic? Have you had success making a primary gene list based on the ultrasound findings? So this is a great question, and I think what we found is the phenotyping in pre-natal cases can be very challenging as compared to post-natal cases. So we believe that our ability to create a comprehensive gene list based on the phenotype is significantly hampered. In terms of reporting variants, we are reporting variants of uncertain significance if there's a very strong, if there's a moderate to high gene disease relationship. Great. And then another question from Ken at the University of Washington. One of the issues in underrepresented populations is the VUS rate. What are your expectations as regards to your VUS rate, and do you plan on addressing that specific issue in your publication? Yeah, so we're definitely interested in understanding, we also, we anticipate that the VUS rate may be higher in our underrepresented populations because there's not as much information known. Also in particular about potential founder variants in these populations that may not have been described. So this is something that we're interested also in working with the other CSER sites to combine information together in order to be able to really increase our understanding of founder variants or other variants. So we hope to be able to publish this information and make it available for other groups also as the study progresses. And then a final question from Dr. Bass at Kaiser Permanente. How many prenatal cases have you done, and what is your yield in comparison with postnatal cases? So I think I can answer this one. So we've done almost an equivalent number of pediatric and prenatal cases, but unfortunately our yield in the prenatal cases is much lower than the pediatric cases. So we've consented and recruited a number of individuals, but we really don't have enough has gone completely through to result session to actually give you the data, but we have found that it's much less likely that we'll get a positive pathogenic or likely pathogenic result in the prenatal cases. And we agree that that's probably due to difficulty creating a primary gene list and also knowledge of the phenotype before delivery. Any further questions? So I just wanted to say to the audience, if you're interested in learning more about whole exome sequencing, the American College of Medical Genetics just had an annual meeting in Charlotte last week, and there are meeting on-demand recordings with the proceedings of the meeting and the release of those recordings will be in early May, so they'll be available for purchase. Okay, thank you everybody for tuning in.

exome sequencing

genetic testing

intellectual disability

developmental delays

Rett syndrome

epilepsy

microcephaly

trichocytodystrophy

rare genetic disorders