Welcome to the ACMG Genomics Case Conference. This series has been planned by the ACMG Education Committee. Each month, the conference rotates among labs that perform genome-scale testing and the clinicians involved in the cases. Use of all case materials that will be presented has been formally consented for by the patients or their guardians. The title of our talk today is Next Generation Sequencing for Somatic Variant Detection in Patients with Overgrowth and Related Disorders, and we are from the Genomics and Pathology Services at Washington University School of Medicine in St. Louis. Thank you for joining us. This activity is supported by an unrestricted educational grant from QIAGEN Bioinformatics and the Injunuity Clinical Decision Support Platform. This presentation is intended for educational purposes only. The presenters today are myself, Julie Nydig. I'm the Medical Director of the Cytogenetics and Molecular Pathology Laboratory at WashU. Letitia Love-Gregory, who is one of our faculty and project coordinator in Genomics and Pathology Services, and Megan Corliss, who is the Genetic Counselor in the Genomics and Pathology Services. We have no disclosures. The Planning Committee, Li-Ming Bao, Monica Giovanni, and the staff, Jane Radford, Claudia Barnett, and Mike Watson also have nothing to disclose. The educational credits are granted through the ACCME. There is one hour of CME for this presentation, and credit can be obtained afterwards. To obtain credit, you will receive an email within 24 hours of the live event that provides the link to complete the CME certificate once you've completed the linked questionnaire. There are overall Genomics Case Conference objectives that have been set out for this particular series, and for our particular talk, we have significant learning objectives for you today. We'd like you to illustrate the utility of next-generation sequencing and the detection of low-variant allelic frequency somatic alterations, review the genes interrogated on the testing panels, and understand the relevance of these genes to the phenotype of the patient, establish that somatic etiology of overgrowth disorders necessitates genetic evaluation of the involved tissue rather than blood, and demonstrate the frequency of genetic variation associated with somatic overgrowth disorders in a clinical laboratory cohort. To ask us questions, please type your question into the Q&A box at the lower left corner of your control panel, and we will answer those verbally at the end of the presentation. Now, to go on to our talk itself, somatic variation is well-recognized in the arena of cancer genetics with tumor mutation profiling, but more recently has been discovered as the cause of a number of congenital disorders. The idea that somatic variance could be a cause of sporadic birth defects as well as skin disorders was first hypothesized by Happel in 1987. The first paper showing somatic mosaicism in these overgrowth syndromes include the Lindhurst et al. paper on AKT variants in proteus syndrome. Our somatic overgrowth panel developed out of cancer assays that included driver cancer genes. Pathogenic variants found in mosaic overgrowth disorders are often mutations that are well-recognized in cancer, and these oncogenic mutations cause cell proliferation. The majority of genes associated with somatic overgrowth disorders are in the shared PI3K, AKT, and TOR pathway. You can see why this would make NGS testing the most likely to pick this up, because both contain admixture of either tumor normal cells or normal cells and cells that have postzygotic mutations, and the mutations may, in fact, be limited to the affected tissue. Obviously, we need high depth of coverage in order to uncover these postzygotic mutations. There is significant clinical overlap when looking at these disorders. Skin, vascular, and neurological abnormalities are found in addition to fibroadipose or... Oops, I'm sorry. I don't know why that happened. Hmm, okay, hopefully it won't happen again. In addition to fibroadipose or other tissue overgrowth, again, the majority of pathogenic variants found in affected patients are in the genes of the PI3K, AKT, and TOR pathway. PIK3CA-related overgrowth syndrome, PROF, encompasses a number of disorders that may all have underlying PIK3CA somatic mutations. The clinical testing is performed here at Washington University in St. Louis. We have seven different gene sets available to order. The somatic overgrowth gene set includes 18 genes. In addition to that, we have sets for just PIK3CA-related overgrowth, McKeown-Albright, NEVA syndrome, Curry-Jones syndrome, Mufuchi syndrome, and Rhizopathy. I will mention that the NEVA syndrome genes are nine genes that aren't necessarily in the 18-gene panel. We use the target hybrid capture approach, paired and 101-based paired sequencing on an Illumina HiSeq 2500 with a minimum DNA input of 50 nanograms. For affected tissue, we mean tissue from the patient where the overgrowth is occurring. We do request a secondary blood sample for a comparison to the constitutional genome. Buccal swabs are only useful when the patient has clinical findings in the head, including brain malformations, vascular malformations, or lipomas of the facial tissue. In addition, we have tested skin, brain, ovarian, and bone tissues. The limit of detection is approximately 3%, and we can look for a mutation that might be found in a variety of syndromes. Now, I'm going to turn the rest of the talk over to, first, Leticia Love-Gregory, who will take over, and then Megan will follow. Hi, my name is Leticia Love-Gregory, and I'll talk a little bit about the case review workflow here at GPS. So, to review these cases, we make use of the software Clinical Genomics' Workspace, or CGW. This is somewhat of an all-in-one cloud-based platform that allows for NGS data acquisition, QC reporting, as well as variance identification, annotation, pre-classification. In addition, we can manually review the variance alleles using IGB. Within this workspace, clinical reports are also drafted and sign-out occurs. As a part of the initial review process, this is performed by our variant scientists, molecular genetics fellows or trainees, and genetic counselor. During this process, we review all of our resulting specimen and NGS QC metrics, as well as any available family or clinical history information. We then draft the clinical interpretations for either pathogenic, likely pathogenic, and the most relevant variants of unknown clinical significance. After this is done, of course, after manually reviewing the variant classification levels to ensure that these are in accordance with the 2015 ACMG AMP Guidelines. As many of you know, the AMP Guidelines is basically a five-tiered evidence-based system that takes into consideration population frequency, variant allele function, be it predicted by either various in silico tools, or based on experimental findings from peer-reviewed literature or case reports. Another round of review is then performed, and sign-out is performed by our board-certified faculty. In some cases, additional testing in the form of Sanger sequencing is performed on a peripheral blood sample. And in these patients, this testing is done to assess whether a variant is of germline origin or truly a somatic alteration. When we take into account all of the various overgrowth-related gene sets, to date, we've tested upwards of about 194 cases, and pathogenic or likely pathogenic variants have been identified in approximately 53.6% of these cases. When we exclude the macune Albright specimens, this detection rate increases slightly to about 56.1%, and we have a fairly low rate of variants of unknown clinical significance. As you would expect, in disorders or somatic overgrowth-related disorders, these patients are identified and cared for by various numbers of physician specialties. And as reflected in the slide that you're seeing here, of course, many of these tests are ordered through the genetics clinic and dermatology. However, these patients are also identified or tested or ordered by neurology, pathology, plastic surgery, and other specialties. To determine the frequency of pathogenic or likely pathogenic variants across the SOMA gene sets, what I wanted to show and demonstrate here is that out of 136 SOMA cases that have been tested here, 70 of those patients were positive for either pathogenic or likely pathogenic variants. And as you can see in the pie graph, the majority of these variants occurred in the PIK3CA gene, which accounts for approximately 70% of these variants. In addition, PIK3R2 as well as PTIN and RASA1, AKT, IEH1. When we take into consideration the resopathy cases that we've performed here, we've performed one. This patient in particular was positive for RASA1. This patient was positive for pathogenic mutation in the N-RAS gene. Three of the 14 McKeown-Albright cases were positive for variants in GNAP, which is one of the G-protein nucleotide encoding genes. Taking a look here at the 13 MEVUS tests that have been ordered, what we found is that 11 of those patients tested positive for pathogenic or likely pathogenic variants particularly in K-RAS, here, H-RAS, as well as PIK3CA. And again, in one of the guanine nucleotide binding encoding genes. We also provide the opportunity to order custom assays within this test. And as you can see here, of the 10 patients that were tested using a custom order, four of those patients harbored pathogenic or likely pathogenic variants, either in PIK3CA, one person was positive for a variant here, IDH1, H-RAS, and GNA11. One of the most common questions that we tend to get for this testing is in regards to the specimen type. And a lot of questions, one of the questions tends to be how informative is a buccal swab or a non-invasive sample as well as peripheral blood? And what we wanted to demonstrate using this slide is that the invasive or when an invasive tissue is ascertained being either fresh tissue, fixed tissue, or fibroblasts collected or generated from these affected areas, what we found is that the detection rate for pathogenic or likely pathogenic variants is significantly higher in the invasive tissue compared to the non-invasive. So as you can see here represented by the blue portion of the column, 54.1% of the variants tested in this data set were pathogenic, 6% likely pathogenic, and about 40% were non-pathogenic or non-likely pathogenic. Another point that I wanted to emphasize using showing this data here, which is also part of this that's been previously published, is that although we're looking at 188 tests here excluding McKeown-Albright, this really represents 180 patients. There were eight individuals here that were tested multiple times. And within that data, which I don't show here, what we found is that nearly half of those eight individuals the initial non-invasive specimen was negative followed by a positive invasive test or the sample in which a pathogenic variant was identified. And I make that point to reiterate that the most informative tissue for this particular test is of course the affected tissue or specimen from the affected area. I'm gonna now sort of turn this over to Megan Corliss who will present several interesting cases that were recently identified in the lab. Hello, everybody. Thank you, Leticia. So this last slide that I wanted to present before I get into the case presentation, this slide schematic shows the alterations, all of the alterations that have been detected in PIK3CA in our laboratory cohorts. And this was actually updated as of October 1st, so it's fairly recent. And I think where this is important to note is that the hotspots in PIK3CA at 453, 542, 545, 1047, you can see here that the hotspots are well represented within our cohort. I think also what's more important for this slide is that you can see is that we've detected pathogenic or likely pathogenic alterations in PIK3CA throughout the entire gene. So we've identified alterations in almost every domain of the gene. And I think this speaks to the importance of the design of the assay in that we interrogate the entire coding region of all genes on our panels. And that's important because not only can we pick up these hotspot alterations, but we can also pick up alterations in other portions of the gene. So what I would like to do now is go through six different cases from our lab cohort. And each of these cases were chosen for specific reasons, which we'll go into in more depth as we go through each case. So case presentation number one, this was a 49-year-old female. She was 49 at the time this specimen arrived here in our lab. She was tested in 2015, so she's now over 50. And this woman was actually sent, she had clinical findings that included macrodactyly of her right hand, so specifically the fourth and fifth digits of her right hand. And she also had some partial overgrowth of her right chest. This patient had actually had a number of medical evaluations throughout her lifetime. Her initial genetics evaluation was performed at 18 months of age. And this evaluation was actually prompted by her mother. Her mother had noticed that specifically the fourth digit of her right hand was larger respectively compared to the other digits. And there was a portion of her palm that also looked larger. So her mom actually initiated this genetics evaluation. At the time of her initial evaluation, she was given a diagnosis of atypical neurofibromatosis. Pretty quickly following that genetics evaluation, at 20 months of age, she had her first surgery. So at the time we received her sample in our laboratory, she had had nine surgeries to date. Interestingly enough, the pathology from those earlier surgeries, none of them were ever consistent with neurofibromas. And she did not have any of the cutaneous findings of that disorder. So it wasn't actually until much later in her life that she underwent a follow-up genetics consultation at the age of 26. And at that time, she was actually given the diagnosis of mild proteus syndrome. You can see here in these pictures, these are pictures representing her affected right hand prior to surgery. Up here in the left-hand corner. So I think everybody can appreciate the amount of tissue overgrowth on her palm. And then the picture taken after her surgery. And this was one of her more recent ones, the one in 2015. And a portion of the tissue that was removed was actually what we tested. For this picture on the top of her hand, again, you can see the picture of her hand prior to surgery. So again, the fourth digit is the most severely affected. Fifth digit is affected as well. And then the picture following her hand surgery. So the testing that we performed, again, it was a soft tissue specimen. So it was a portion of tissue that was removed during her surgery from her right fifth finger. And that was sent to us to have the SELMA gene set performed. So at the time that we did this patient's testing, was a couple of years ago. And at that time, it was an 11 gene panel. We did identify a missense variant in AKT1. So it was a glutamic acid to lysine amino acid change at codon 17. And as you are all aware, this is the alteration that is detected in greater than 90% of patients with Proteus syndrome. For those of you who may not be familiar with visualizing NGS data, you can see the picture here on the slide. The reference allele, the wild type allele is the C. And then the alternate allele, the variant allele is the T. So you can see that's represented in these colors at the top of the picture. The blue represents the wild type allele and the red represents the alternate allele. And you can see in this box here that this specific alteration was detected at a variant allele frequency of 21% in this patient's tissue. Following the detection of that alteration in her affected tissue, we performed comparative analysis on her peripheral blood sample by Sanger sequencing. This analysis did come back negative, further reinforcing that this was a somatic alteration in this patient. After sending these results back to the clinician, a number of clinical recommendations were made based off of this result. One, it served to educate both the patient as well as her primary care physician and really provided molecular confirmation of her clinical diagnosis. It also allowed for surveillance for some of the other findings that you can see with Proteus syndrome. So both skeletal, pulmonary, and soft tissue manifestations as well as these patients can be at an increased risk for pulmonary embolism and DVT. Additionally, like we mentioned earlier, this patient had gone through multiple surgeries and had this lifelong diagnosis that she struggled with. So it also allows for continued psychosocial therapy and support related to her diagnosis and complications of her ongoing surgical needs. So case presentation number two is a nine-year-old male. And this individual was seen due to multiple telangiectasias and multiple small capillary malformations on his body. At the time he was evaluated, his family history was essentially negative. His parents did not report any family history of genetic disorders or birth defects or anyone really with similar skin findings as this patient had. Hereditary hemorrhagic telangiectasia as well as ataxia telangiectasia were both in the differential for this patient. However, they were not very high up on the list, mainly because he had no other clinical findings or symptoms of either of those disorders. There was no reported history of hemorrhage or bleeding. His parents did not report any issues with gait or ataxia or any speech issues. So when we received this sample in the laboratory, the indication for testing on the requisition was vascular malformation. And the clinician ordered the full soma gene panel, the 18 gene panel. This was performed on a fresh tissue skin biopsy sample that was actually taken from the patient's forearms where many of these vascular macules and telangiectasias were located. For this case, we did detect a pathogenic variant in the RASA1 gene. So we detected a novel four base pair deletion. Again, for those of you who are not familiar with this picture below, these white lines here are representative of the four bases that are deleted. This was detected at an allele frequency at the time that was consistent with somatic origin. So the allele frequency for this alteration was essentially 20%. Following the detection of this RASA1 deletion in the patient's affected tissue, we again did have a blood sample from this particular patient. So we sent that blood sample for Sanger sequencing and it actually returned positive. So based off that result, it is very likely that this is a germline alteration in this patient, which a lot of RASA1 mutations are. Following the return of this result to the clinician, the clinician brought the family back in and he reviewed the results with the family, did an evaluation of both parents to look for subtle findings of this RASA1 deletion, which neither of the parents had. And at that time, they decided to undergo testing mainly for recurrence risk. For RASA1 deletions or mutations, 70% of affected individuals have inherited this mutation from a parent. 30% are de novo. So for that piece of information as well, the parents did decide to both undergo testing. We recently received those results back, and both parents did test negative for that deletion. Case presentation number three is a four-year-old male. Indications for testing for this patient were macrodactyly of the toes, developmental delay, and seizures. And the full soma gene set was ordered on a fresh tissue biopsy from this patient. For this case, we did detect a likely pathogenic variant in the PIK3CA gene. So this was a glutamine to arginine change at position 546. And the reason we chose to actually highlight this case is that we found this case very interesting. It was an alteration that we had not detected previously in our lab cohort. It also really speaks to the usefulness of the ACMG AMP guidelines for these variants. By applying three moderate as well as three supporting criteria for pathogenicity is how we arrived at the likely pathogenic variant status. Interestingly, with this patient as well, we had two other patients in our cohort that had alterations at this amino acid, but it was actually a glutamine to a lysine change. So this was the first time we had seen the glutamine to arginine change in a patient. Case presentation number four is a one-year-old male. He initially presented to medical attention due to left-sided stiffness and preferential use of his right hand versus his left hand. This patient actually presented to neurology first, and during that neurology evaluation, he underwent an MRI, and the MRI noted hydrocephalus, Chiari malformation, and polymicrodiarrhea. Following the MRI and the neurology consult, he was then sent on to genetics for a genetics consultation. During that genetics exam, they noted a vascular birthmark on his scalp, a cafe au lait spot on his leg, and extensive hair growth across the upper portion of his back. When we received this case in our laboratory, the clinician was very suspicious of an alteration in the PIK3CA gene. For this case, the PIK3CA gene only was ordered initially. The sample we received in the laboratory was a buccal swab specimen for this patient. Along with that primary buccal swab specimen, we also received a peripheral blood sample as a comparator. We did run the testing on the buccal swab, and the initial result for PIK3CA only was negative. It's always a difficult thing in our laboratory because it's difficult for us to tell whether the buccal swab is just not an effective specimen for this patient, or if additional genes needed to be ordered, and we weren't really interrogating the right gene. So, in this instance, we returned the negative PIK3CA result to the clinician, and they actually came back asking us to reflex testing to the full SOMA gene set, which was the 18 panel. Within that reflex, we did identify a pathogenic mutation in this patient, and it was actually in the PIK3R2 gene versus PIK3CA. We actually identified a glycine to arginine change at position 373. Again, you can see the reference allele here, the G, and the alternate A allele. This variant was detected at a frequency of almost 30%. With this sample, again, we had a peripheral blood sample from the patient, so we did send that off for Sanger sequencing, and the comparator sample, the comparator peripheral blood did come back negative for that alteration in PIK3R2. I like to illustrate this case because, again, I think it's very important to show the immense overlap in clinical phenotypes with these patients, and essentially why it's more important to order a panel on these patients versus a single gene test. This is also important because it shows, really, the utility of when buccal swabs may be useful. When we get calls in the laboratory regarding specimens, we highly recommend affected tissue specimens as the primary sample. Where buccal swabs become very useful is if the patient has any findings of the head or neck or any brain abnormalities, because at that point, a buccal swab could potentially be an affected tissue specimen. Case presentation number five, so this is a one-year-old male, and the indication for testing was suspected linear nevus sebaceous syndrome. For this case, the patient had actually already had a prior biopsy, so this specimen that we received for testing was in the form of formalin-fixed paraffin-embedded tissue. This is something that we're very comfortable with in terms of extraction in our lab. Our main assay here at GPS is a cancer assay, where essentially almost every specimen that we test is in the form of an SFPE block with tumor. For this specimen, it was representative of a scalp lesion that had been excised from the patient. In this particular case, it was one of the smaller gene sets that was ordered. It was the nevus gene set, which is a nine-gene panel. Again, we did detect a pathogenic KRAS mutation in this patient, so many of you may recognize that this is a hotspot mutation in KRAS, the G13D alteration, and this alteration was detected at a variant allele frequency of 7%. For this case specifically, again, it really provided molecular confirmation of this patient's clinical diagnosis. The vast majority of patients with linear nevus sebaceous syndrome will have alterations in one of the RAS genes. The last case I would like to go through is case presentation number six. This was a six-year-old female, and the indication for testing was Mifuji syndrome. Again, this patient had already had prior biopsies, so the sample that we received was in the form of a formalin-fixed paraffin-embedded tissue specimen, and this specimen was representative of a hemangioma that had been biopsied prior to testing. We were sent that formalin-fixed paraffin-embedded tissue specimen, and when this order came into the laboratory, it was a custom panel ordered by the clinician, and the clinician wanted to interrogate the IDH1 as well as the IDH2 gene. As you can see here, again, we did detect a pathogenic variant, so we detected the pathogenic IDH1 arginine to cysteine change at position 132, and this case is important for a number of reasons. So one, this alteration was detected at 3.3%, and really our limit of detection for this assay is 3%. There are rare instances in cases where we can report something lower than 3% if it's something that is a known pathogenic mutation, typically in a hot spot, and fits very well with the patient's phenotype. Additionally, for this case, where this is important is that back when we received this order, this was a custom order by the clinician. We had not yet advertised the option of the Mifuji syndrome gene set through our laboratory. This clinician had a patient and had been researching in the literature and wanted to test these two genes and had actually noticed that they were part of our cancer panel. This clinician was very savvy and knew that they would need deep sequencing, had actually reached out to us to see if it was possible to test this specimen, which we agreed to, and again, we're able to provide molecular confirmation for this patient. So before I go through this summary, I did want to make a couple of notations regarding testing in our laboratory. The one piece of important information, really, for us is the importance of clinical information. As we've mentioned many times, there is a significant amount of overlap between patients with these disorders, and it is very useful to us in reviewing these cases, as well as to the clinician who signs out the cases, to really have detailed information on the patient. It also makes us more confident in those calls, especially when we're looking at variants that may be novel or that are not known pathogenic or reported variants. So in summary, overdose syndromes have overlapping presentations, which again, can make selection of a single gene for analysis difficult, which is why we feel it's very useful for many clinicians to order the overarching soma gene panel. We went through six different individual patients that really had a variety of clinical presentations, and five of these six patients had somatic alterations in genes associated with overgrowth disorders. The one exception to these cases was the individual with the RASA1 deletion, which is very likely germline, since it was detected in his peripheral blood sample. A number of pathogenic variants were found, and most of these are found in affected tissue samples. So I think it really hits home why we highly recommend sending an affected tissue, even over buccal swabs occasionally, because the likelihood of finding that disease-causing alteration is much higher in an affected tissue. And then lastly, somatic variants with low frequency are best detected via NGS analysis, because we really do get the high depth of coverage and can detect these low variant allele fractions, typically down to two to three percent. We have provided here a short list of references. These are kind of our go-to papers in the lab here when we're reviewing cases and analyzing data from the NGS data. This was created as a handout for you all as a PDF, so you can find that under conference materials. We're more than happy to give you additional references if anyone is interested. You're always more than welcome to reach out to us, and we can provide those to you as well. We would like to thank ACMG for allowing us to do this presentation and present all this interesting information to you all. We would like to also thank our past and present staff here at Genomics and Pathology Services, as well as Genome Technology Access Center staff, that is our wet lab staff here at Washington University in our laboratory. And we would also like to immensely thank our patients, as well as our ordering clinicians and institutions. So for questions now, we're going to open up the question box. So if you want to ask a question, please type the question into the Q&A box in the lower left corner of your screen. So we're going to start here. The first question is from Hal at Kaiser Permanente. How reliable is your analysis using paraffin-embedded tissue? Do you want me to take this one? Go ahead. Okay. So as we mentioned earlier, so our SOMA panel actually developed out of our cancer assay here at WashU. We had some current clinicians here in genetics who needed deep sequencing of some of their patients, and they knew those genes were on our cancer panel, so that's how everything started here. That being said, our main assay is a cancer assay, and pretty much every specimen that we test is a formalin-fixed paraffin-embedded tissue specimen. I can tell you the, so our most recent data, which was from October 1st, we had tested a total of 35 cases that were formalin-fixed paraffin-embedded tissue, and 25 of those reported pathogenic or likely pathogenic results. So about 70% of those patients with the FFPE specimen had a pathogenic or likely pathogenic variant. Next question is Albert Gold from Niterra. You mentioned Gorlin, he thinks, but patch 1 and patch 2 aren't on your panel. Perhaps I am mistaken. Correct. We do not have patch 1 and patch 2 on our panel, so Gorlin is not part of that at this time. And it's usually not a somatic variant, usually it's constitutional in the Gorlin patients. The third one from Albert. He actually asked us to skip that one later on. Average coverage of our NGS panel from Han Shin. Han Shin Lin from Western University, the average coverage of your NGS panel, our average coverage currently is 1500 to 2000x. And another question, for the skin sample, did you get fibroblast cell culture or direct skin tissue? Do you want me to take this? Yes. Okay. So this is a good question. We get asked this a lot. So the skin samples are often fresh tissue skin samples, not fibroblasts. So typically 3 to 4 millimeter punch biopsies. There is some discussion, and I'm sorry, I don't know if there's specific papers in the literature about this, but there was discussion between us and the lab here about fibroblast samples and whether the mutation can actually be lost in the fibroblast sample due to long term culture. We don't have a perfect answer for that. We've had fibroblast samples where we've detected pathogenic and likely pathogenic mutations, and we've had fibroblast cell cultures that were negative. But the vast majority of specimens that we receive are fresh tissue skin biopsies. The next question is, do you use other techniques to confirm findings? How about those mutations with less than 10% allele frequency, and how do we confirm those? Do you want to take that? Okay. So for that one, we do not. So Sanger sequencing is really the only follow-up test that we will do on these patients on the peripheral blood. Your comment about mutations with less than 10% and actually probably less than 20% is noted. So it's not that when we test these peripheral blood samples that it's an absolute yes or no, it's in the blood or it's not. What we're really doing more for that is trying to determine whether it's germline, since the recurrence risk obviously will be much different for that. So we feel most alterations, if they're germline, they're going to be picked up via Sanger sequencing because they're going to be at a variant allele frequency that's higher than 20%. There's the next question from Barbara at Emory is, do you measure methylation changes? And the answer is no. No, we do not. The next question from John Brat is, how do you distinguish mosaicism at 3.3% from sequencing artifact? Do you want to take that one? Let me go ahead and go right up front. So in terms of, I think there's a number of things that we do here. We have essentially, we have various filters that we use. So we look essentially, our system that Leticia mentioned earlier, the clinical genomicist workspace, what's nice about that system is that we, any of us who are reviewing cases, because it's often not the same person, it's often many different ones of us that are reviewing cases, we can make comments when we've seen variants in multiple cases and multiple runs that could be artifactual. We also look at things like where it occurs in the genome. If it's messy in that area, that may be another reason why it could be an artifact. If it's in an area of the genome maybe where it's a homopolymer run, then that's something that we look at as well. Leticia, do you have anything to add? I think, yeah, we definitely try to look at the, visualize the variants, the reads for each of these cases. And if a variant is near 3.3 percent, we're going to be looking at that manually just to make sure that it's not in a region in which there is high homology as well. So there are a number of factors that are taken into consideration even beyond what's built into our pipelines for filtering these variants and identifying artifacts. And there's a huge depth of coverage for this assay. I think the depth of coverage is different from other next-gen assays, specifically because we are looking at cancer cells and we're used to looking for mosaicism. So for, this is from Minji. For cases with suspected linear nevus sebaceous syndrome, have you detected pathogenic or likely pathogenic variants in uncultured tissue from the lesion or normal tissue? Off the top of my head, so, yes, we have. So the number of cases that specifically that I can think of that had linear nevus sebaceous syndrome as the indication, to my knowledge, those were all fresh tissue specimens, so uncultured. I believe there was one or two that were also in the form of an FFPE block, but, again, uncultured specimen. And not normal tissue. Right. Not normal tissue. We don't even really accept normal tissue unless it's an instance of the clinician not being able to get a peripheral blood sample or something that can be used as comparator. Yeah. Okay. So from, I'm probably going to missay your name, so I apologize in advance. Yuan Xue. ACMG guidelines are for variant interpretation intended for germline variants. Do you use the same guidelines to interpret somatic variants? I think you should take that one. Sure. You can stay there. I hope you can hear me there. In the soma cases in which we're interpreting variants, indeed, you're correct that the ACMG guidelines are developed for germline variants. But in the case of soma, for somatic variants on the cancer side, we definitely use a different system for classifying our variants. However, we do use ACMG guidelines to determine whether these variants are pathogenic or not. And we think this is actually appropriate in this setting. Yes. And part of the reason why it's appropriate is because these patients have syndromic diagnoses. As opposed to cancer diagnoses. And so even though we're looking at somatic variation, we're looking at pathogenic variants that are affecting tissue formation that is not necessarily tumor growth. Diane from prevention. What is the average and range of mosaic variants between non-invasive and invasive tissues? So, Lee, I can tell you altogether from our cohort. And I'm sorry I don't have these separated out by non-invasive versus invasive. But the lowest allele fraction that we've reported was 2%. And I should note, too, since we're talking about that, is our limit of detection really is 3%. But by manual inspection and looking in the case, if we find an alteration at a known hotspot position, and to date, anything that we've reported under 3% has been PIC3CA. So those are known pathogenic mutations. And it also needs to fit very, very well with the patient's phenotype. So that's one thing that's important. But we have reported alterations as low as 2%. And then we've reported alterations as high as 45%. So there's definitely a range of what we've seen in terms of the variant allele frequency. From Laura at CHOP. Given the mosaicism of the RASA1 deletion in the affected tissue, how do you know this is germline in this individual? We don't know for sure. But the fact that most RASA1 alterations are often germline, and we did perform Sanger sequencing on the peripheral blood sample, and it was detected in the peripheral blood sample, we're making the assumption that it is germline at this point. But I don't think we can say 100%. It could be something that's still somatic, but possibly present in multiple tissues. I don't know if you want to add to that, Julie. No, I think that's perfectly answered. Do you do a larger panel and use software to subselect a smaller panel based on order from Soheil? So this is a target hybrid capture assay. So it actually encompasses 177 genes. So the capture set is the same as our cancer capture set. So it's 177 genes, and some of those are carved out specifically for SOMA, but the majority, obviously, are shared between cancer and SOMA. So essentially, yes, we are doing a larger panel because we're capturing all 177 genes, but that is selected down into the order gene set. So it is something that we actually, we upload something called a CSV form at the time the case is accessioned, and that causes it for us only to see the order gene set. So we don't see anything else in the capture set. It's masking anything that we haven't ordered. Exactly. And we just don't see those at all. Variants could be unmasked from other genes if requested, but at this point, we are basically working off of the specific gene panels that have been preselected. Next question, and this appears to be the last one, so anybody else who wants to get into questions, please go ahead. Jiaqi from Quest Diagnostics asks, is the cutoff different between invasive tissue and non-invasive tissue? I think you're asking there the cutoff in terms of our limit of detection, and the answer is no. So really, 3% is kind of our hard cutoff, again, unless we manually inspect something and find a known pathogenic mutation, then we very rarely will report under 3%. But that 3% doesn't matter whether it's an invasive or a non-invasive tissue. If we find something that's pathogenic or likely pathogenic and it fits well with the patient's phenotype, that's reported regardless of the specimen type. Okay. I don't see any other questions here, so I think... So there's one more slide, so we just want to say thank you for allowing us to present these really interesting cases. We're more than happy if anyone has additional questions. Please reach out to us. Our website is gps.wustl.edu, so gps at w-u-s-t-l dot e-d-u. You can reach any of us on that website. There is an email for the lab that comes through to us. Thank you to QIAGEN for supporting this. And a recording will be available in the ACMG Genetics Academy. We do ask that you all mark your calendars for the next case conference. This conference is scheduled for November 15th from 2 to 3 p.m. Eastern Standard Time. The title of this conference is Proceeding from the Non-Diagnostic Exome, Cases from the Stanford Site of the Undiagnosed Diseases Network. And this is going to be hosted by Stanford University School of Medicine. I will make one more comment before we end. If you are getting CME credit for this, please stay on. You're going to have an evaluation that comes up, and that needs to be completed in order for you to obtain that credit. Thank you.

next-generation sequencing

somatic variant detection

overgrowth disorders

ACMG Genomics Case Conference

genetic evaluation

clinical reports

pathogenic variants

clinical laboratory cohort

detailed clinical information