Hello, everyone, and welcome to today's installment of the ClinGen Somatic Cancer in FIC Virtual Molecular Tumor Board series. Today, we have a group from Nationwide Children's Hospital who will be presenting comprehensive genomic profiling in an indeterminate pediatric tumor guides counseling and therapeutic strategies. I'm Beth Patel, and I will be moderating the session today. Just a quick disclaimer that this session is for educational purposes only. Please use professional discretion on every case, as you always do. So this case series arrived as a collaboration between the Variant Interpretation for Cancer Consortium, or VIC, and ClinGen Somatic Working Group, specifically the VIC Virtual Molecular Tumor Board Group and the ClinGen Somatic Working Group. If you're interested in getting involved with either of these entities, please visit cancervariants.org to learn more about VIC. You can join a mailing list, you can join working groups, or just learn a little bit more about what's happening in this consortium group. It's open to all. And then also the ClinGen Somatic Cancer Clinical Domain Working Group is a very active group as well. I enjoy being a member there. And to learn more about this group, you can visit clinicalgenome.org and look at the working groups. And then we're very gracious to have this partnership with the ACMG Genetics Academy, who have hosted these webinars for us and offered CME credit for the participants. So thank you so much for coming today. So just a quick introduction to our planning committee and ACMG staff that work together to make these presentations possible. So first and foremost, we have our CME Activity Director, Cordana Raza. She is from Children's Hospital in Los Angeles, and she is a very active member in many different leading consortium groups, like the ClinGen Somatic Working Group, VIC, you've probably seen her around CGC as well, if you're part of that. I'm Beth Patel, I'm a senior developer at the Mayo Clinic. So I develop NGS assays and focus on interpretation workflows. I've been a co-lead for the VIC Virtual Molecular Tumor Board for four years and an active member of the ClinGen Somatic Working Group. We have a fresh face to our team this month. Manuela Benari is a bioinformatician at Charité in Berlin, and she is joining me as a new co-lead for the VIC Virtual Molecular Tumor Board. I also wanted to acknowledge the efforts of Shruti Rao, who is the previous co-lead of the VIC Virtual Molecular Tumor Board. She really helped to pave the groundwork for this series. She was administrator for ClinGen Somatic Working Group as well, and just a really awesome person to work with. Additionally, Alex Wagner from Nationwide Children's is on our planning committee. He is a lead on VIC and also a member of the ClinGen Somatic Working Group. And then from ACMG, Dr. Max Muenke and Jane Radford have been very instrumental in keeping this series going, as well as Claudia. So the American College of Medical Genetics and Genomics is accredited by ACCME to provide continuing medical education for physicians. This session is worth one AMA credit or that equivalent. And to obtain credits, you need to register, and by being here, I know that you have all registered for this session. You will need to take a short quiz at the end of the session, and please allow 24 hours to view and print a certificate. Additionally, we will have a recording of this webinar available in two weeks through the ACMG Genetics Academy. Just a couple things about logistics of the session today. So to submit a question, please use the Q&A feature and reserve the chat box only for technical issues. And if you see a question that is of particular interest to you in the Q&A, feel free to upvote it by clicking the thumbs up. So I am very excited to introduce today's speakers from Nationwide Children's Hospital, Dr. Catherine Contrell, who is the Senior Director, the Institute for Genomic Medicine Clinical Laboratory at Nationwide, and an Associate Professor in the Departments of Pathology and Pediatrics for The Ohio State University, Dr. Kathleen Schaefer, who is a Clinical Lab Director in the Institute for Genomic Medicine, and Kristen Zaggio, who is a Certified Genetic Counselor in the Division of Hematology, Oncology, and Bone Marrow Transplants. So thank you so much for being here, and we look forward to your presentation. These presenters have no financial disclosures. And today's learning objectives, so at the end of the presentation today, we should be able to describe the genomic testing methods used to determine clinical management for this patient and their family, also explain how family history influenced the interpretation of genomic results, and discuss the importance of genetic counseling in the context of high-throughput genomic cancer testing in this case. And with that, I'll hand it off to Dr. Cottrell. Beth, thank you so much for allowing us to join and to present this patient's story today. We know that test methodologies in the setting of a clinical cancer diagnostics are varied. Histologic assessment of disease-involved tissue is a fundamental component and often the initial step in a diagnostic workup. Cytogenetic studies have utility in chromosome or locus enumeration, as well as in the detection of structural variation, though dividing cells are a necessity outside of FISH studies. Increasingly, genomic analyses are used to study variation at the molecular level. These studies may be focused or broad, and they may encompass DNA or RNA-based studies, methylation or expression analyses, or protein-based studies. As part of the strategic plan at Nationwide Children's Hospital, genomics has served as a pillar to accelerate advances in research and clinical care with dedicated strategic funding and resources to drive those initiatives. Translational research within the Institute for Genomic Medicine is one such avenue. In 2017, an IRB-approved protocol was developed for genomic profiling of patients with rare or treatment refractory CNS malignancies, solid tumors, or hematologic disease. In order to be eligible for the protocol, a clinical care team nominates a patient into our REDCap database, along with the reason for desired genomic profiling. Eligibility is then reviewed by our study team, followed by an approach for consent onto the study. Both tumor normal exome analyses and RNA sequencing of the disease-involved tissue form the fundamental basis for this genomic profiling protocol. The goal within this protocol is really to enable personalized medicine, so we can find variation which may inform diagnosis, establish a prognosis, or identify the potential for targeted therapeutics. Importantly, this protocol also enables the detection of germline predisposition, because we are using paired tissues. Whenever possible, parents are also consented onto the study, in addition to the probands. This also enables a collection of germline samples from those parents, and these are particularly useful in trying to discern variant inheritance when needed. These translational research results were initially discussed internally within IGM, followed by a formal presentation back to a multidisciplinary team at a molecular tumor board. Experience with decided clinical utility can then be clinically confirmed following submission of a clinical order, and then returned into the medical record to allow for use in patient care and decision-making. With over four years of experience now with this protocol, this has enabled us to establish a clinical validation for our paired tumor normal exome analysis, as well as a validation for our CNS methyl array classification. That will be internally orderable for our clinicians at Nationwide Children's this fall. We recognize that each of the patients enrolled onto the protocol are unique, and this enables us to develop an end-of-one approach. In addition to the fundamental studies of our paired tumor normal exome analysis, as well as our RNA sequencing of the disease-involved tissue, we're increasingly able to use other genomic techniques to really identify the genetic underpinnings of disease within these patients. As I mentioned, increasingly in the setting of CNS malignancies, as well as in the setting of our sarcomas, we're able to use the dKFC classifier following methylation analysis to better classify some of our tumors. Additionally, we can employ other techniques, such as TaqBio long-reap sequencing, to further discern isoforms that may be of interest in our disease-specific tissue, or to perform genome sequencing to further discern structural variation which may be occurring. Immune studies are also performed in certain cases. The menistering immune profiling may be utilized to further discern the immune profile within those disease-involved tissues. Additionally, anything that would be clinically relevant, such as in the setting of a germline variant of interest, can then be offered for at-risk family members, such as in the setting of cascade testing. On the basis of the types of methodologies that we're employing within our protocol, we are able to detect a wide range of genomic variation. Through the use of our paired exome analysis, we are aiming to achieve a high depth of coverage. Typically, we are averaging about 250x depth. We perform this off of both FFPE as well as frozen tissue. And then our comparator normal tissue can encompass blood, saliva, buccal samples, or fibroblasts. The high depth of sequencing allows us to achieve a high sensitivity to detect low variant to low frequency alterations. So, we're aiming to detect not only our germline variants out of that comparator tissue, but also our somatic variants, again, with high sensitivity. So, single nucleotide variants, small insertion deletion events, and copy number alterations are all discernible from the exome analysis. Additionally, to further enhance our ability to pick up copy number alterations, we have probes, which are spaced throughout the genome, to enhance our ability to pick up copy number alteration out of that exome. Structural variants are primarily gleaned from our RNA sequencing of the disease-involved tissue. Additionally, the RNA sequencing enables us to be able to look at alterations within expression. So, we're able to look at expressions both across our cohort, as well as in comparison to other pediatric tumor resources, such as the UCSD Treehouse database. And increasingly, as I mentioned, methylation is one of the tools that we're using to better classify these tumors. So, with initial patient enrollment in 2018, following approval of the protocol by the IRB in 2017, we've really seen tremendous growth each year with regard to patient nominations, as well as an expanded number of clinicians that have gained comfort and experience in nominating patients onto this protocol and engaging in result return. So, there's really been tremendous benefit for the patient population within our institution. In terms of a summary of the sequenced cases, so as of the end of July, we had sequenced 281 patients. You can see on the left of the screen a greater number of males than females. In terms of age at diagnosis, there is a range. We do see enhancements or enrichment within the adolescent young adult population. We also do have patients greater than 18 years of age that have been nominated onto the protocol, as long as they have engagement and care by a clinician at Nationwide Children's. Additionally, we do see a number of patients that are less than a year, including a cohort of congenital patients that have actually been diagnosed with cancer at less than three months of age. In terms of the breakdown of the typical diseases under study within this protocol, we had early involvement by our neuro-oncologists, and so about half of the cases that we've sequenced to date have represented CNS malignancies. About a third of those are now solid tumors, and then you see a lower amount representing our hematologic diseases. And in looking at an overall summary of our patient cases, we've accepted 357 patients onto the protocol as of the end of July. We've been able to consent 309 and sequence 281. You can see a breakdown across the cohort of those signings, which were felt to be diagnostic, or therapeutically relevant. In addition, among all of the patients that we've studied to date, we've found that about 95% of patient analyses have provided meaningful information. So that means we've been able to discern some of the genetic underpinnings that were contributing to that patient's disease. Thermaline signings were present in about 17% of our patient population. In looking at this one other way, we can break down medically meaningful findings across the cohort. So you can see at the very, very top, each of these represents an individual patient with the tumor type discerned by color with the key shown below. Some of our pediatric tumors are very, very quiet, so they only have, for instance, a single driver event. Those would be represented at the far right here. Some of our tumors under study are much more complex, however, and they have a variety of copy number alterations, fusion events, single nucleotide variants, and at times, germline cancer predisposition. And so one of these more complex cases will actually be presented now by Dr. Kathleen Schaefer. All right. So I will be presenting a patient that presented at our institution. This is an adolescent female who initially presented with epigastric pain, emesis, and abnormal liver enzymes. She had an abdominal CT performed when she presented with a large retroperitoneal mass with a large cystic component that was identified. This was ultimately demonstrating mass effect on some surrounding structures, including the stomach, the duodenum, and the pancreas, leading to biliary obstruction and extension of the pancreatic duct. At the time of presentation, she also presented with extensive hepatic metastatic disease. At that time, the radiologic differential was pancreatoblastoma or other pancreatic tumors. So we're going to walk through the timeline of our return of IgM results in the setting of both our translational findings as well as our clinical findings. So you'll see these headed with a banner that is either shown in green for our translational findings, as the patient has been nominated onto our IgM translational cancer protocol, or our clinical findings shown in blue. So starting at day zero, when the corneal biopsy was taken, day one, there was an Archer solid fusion panel that was ordered from our pathology. This would potentially help with diagnosis of this particular tumor type, and we'll get into that in just a second. From a translational standpoint, day three, following that needle core biopsy, the clinicians in the lab had a conversation, noting that this particular tumor was difficult to classify, and this might be a good one for protocol nomination. You'll see that even about three hours later, this patient was ultimately nominated for our IgM translational cancer protocol, for which she was ultimately accepted onto protocol and sequenced. So this slide breaks down the pathology assessment. So the initial pathology diagnosis was a high-grade, poorly differentiated neoplasm, so a very nonspecific diagnosis. On the right-hand side is a table of the IHC markers which are performed. So those shown in red are the positive markers that we see that were performed with the initial IHC staining. At the time of when we sequenced this tumor, just to point out in regard to the rest of our clinical and translational testing, we had 100 percent tumor cellularity. So this was a relatively high tumor content tumor for us to work with. Additionally, in parallel with both our clinical as well as translational testing, there was an outside path consultation or multiple outside path consultations to really try to get at what is the spinal diagnosis of the tumor. So right around the time that we finished up our translational testing, there was a final pathological diagnosis due to positivity of BCL10, trypsin, and chymotrypsin, which ultimately led to a diagnosis of pancreatic HNR cell carcinoma. So that's what the blue highlighted IHC stains are demonstrating here. So the first testing that was ordered on a clinical basis was the ARCHER targeted RNA sequencing panel. So this is targeted RNA sequencing performed using an anchored multiplex PCR, also known as AMP technology, which allows us to identify both novel as well as known fusion partners. Our solid tumor fusion panel encompasses 115 genes associated with solid tumors. And what you can see from our output here in the table is we identified a fusion of TRMM24 and BRAF. The five prime fusion partner, TRMM24, fused at exon 10 to the three prime fusion partner, BRAF, fusing at exon 9. And this is graphically demonstrated in the middle of the slide. Using our ARCHER metrics, this was classified as a strong oncogenic fusion event due to the number of unique start sites, the percent of reads, and the unique reads supporting the event. ARCHER also takes into account if the event has been previously identified within their own database known as QIVER. Important for the mechanism of this described fusion, so BRAF fusions in particular, this fusion was predicted to be in frame. And this is important because you would want to retain that kinase domain of BRAF, ultimately to lead to constitutive activation of the RASMAP kinase signaling passage, which we'll get into a bit later when we discuss the translational results. This particular fusion in and of itself is not diagnostic of any particular tumor type, but it has been described in cancer. So there are reports of it being described in colorectal cancer, non-small cell lung carcinoma, and cutaneous spitsoid melanoma. So following these initial results from pathology, the diagnosis was altered to be a high-grade, poorly differentiated neoplasm with a TRN24 BRAF fusion. So still pretty nonspecific diagnosis. We classify this particular fusion using the AMP-ASCO-CAP somatic variant interpretation guidelines. So this is a four-tier classification system, which you can see demonstrated here, with Tier 1 and Tier 2 demonstrating variants of strong and potential clinical significance. Tier 3 would be variants of unknown clinical significance, and Tier 4 would be benign or likely benign variants, which in our lab we do not report out on a clinical or translational basis. The main difference between somatic and germline variant interpretation would be that somatic variant guidelines are interpreted on the basis of the tumor type under study, as well as if the particular variant has an impact on therapeutic, prognostic, or diagnostic implications. So in this case, we're working with a tumor that is not a particular diagnosis. It's very nonspecific. And looking at TRN24 BRAF fusions, we can implicate that there are therapeutic implications for this. So there is a study that has looked at TRN24 BRAF fusions in particular, and it did demonstrate through in vitro analysis and cell line work that this particular fusion does activate the MAP kinase signaling pathway, and it does demonstrate responsiveness to both MEK and RAF kinase inhibitors, which allows us to classify this as a Tier 2 level C due to the evidence provided by multiple small published studies with some consensus. So moving into our translational studies, as Kathy mentioned previously, we did perform paired tumor normal exome sequencing, which allows us to look for both germline as well as somatic variation. So in this case, our comparator germline sample was the peripheral blood, and the disease-involved sample was from that needle core biopsy from that midline epigastric tissue. RNA sequencing was also performed from that disease-involved sample. So going back to our timeline of return of results, looking back on this translational side up at the top, on day 8, the patient was consented onto the protocol and peripheral blood was drawn. Day 9, the peripheral blood was extracted. So this would be our germline sample, and we were actually able to use nucleic acids that were initially extracted from that clinical ARCHER that was ordered. So we did not have to obtain any additional tissue for extraction for our translational studies. By day 14, we were able to communicate a preliminary report for that ARCHER analysis, which identified the TRM24B RAF fusion, with the final ARCHER report being reported by day 22. So I'm going to pass it on to Kristen, who is going to introduce our family and the pedigree. Yeah, so we wanted to briefly look over the pedigree and family history, because this ended up being critical in some of the interpretation of the subsequent results that came from these IgM studies. So typically, we meet with families and get a family history very early on in this process, at the time of consent onto the study. And in meeting with this family, so to orient you to the pedigree, here is our patient, the proband with the epigastric tumor. She had one full brother who was 10 years old and healthy. And then you can also see the family had lost one prior child to stillbirth. Looking at the parents, parents had reported that they were understandably concerned about cancer risk based on previous cancer history in their family. So on the paternal side, the patient's father was in his 40s and healthy. He had one twin sibling who was also healthy and had no cancer history, nor did the paternal grandparents. But in the paternal great-grandparents, there was a history of pancreatic cancer at an older age, at 80 years of age. That individual had a sibling with uterine cancer at 60 years of age and another sibling with multiple myeloma at 45 years of age. And then the paternal great-grandfather also had a history of cancer, although the family didn't have much information other than it had occurred at an older age. But more striking for us was the family history of cancer on the maternal side. So the patient's mother is illustrated there with the dot pattern and was currently 42, but had a history of breast cancer, which was bilateral and triple negative at 38 years of age. And at the time of her cancer diagnosis, she had undergone genetic testing via a germline clinically available panel. So this was a commercial panel that was initially a 28-gene breast cancer predisposition panel with a reflex to an 84-gene pancancer panel, which had been completely negative. Her mother had also had a history of breast cancer, interestingly, at the same age, 38 years of age. Her cancer was unilateral. But it's worth noting that the maternal grandmother also had a diagnosis of schizophrenia, which complicated the family dynamics and made it difficult for us to get any additional information or medical records from her regarding her diagnosis. She was adopted, and so there's no additional information on family history beyond the maternal grandmother on that side of the family. So unfortunately, although we have a suggestion that looks clearly to be an autosomal dominant pattern of inheritance, we don't have any more extended family on that side to help interpret the information that we had. And otherwise, on the maternal side, there's a history of some uterine cancer at an older age in a sibling of the maternal grandfather, but otherwise pretty unremarkable cancer history. So certainly, as we were consenting and enrolling this family, we were viewing it through the lens of this potential suggestion of an autosomal dominant breast cancer predisposition in the family. So we'll revisit this pedigree later on, but just wanted to sort of present this up front as we're thinking about this case. So moving forward to our germline analysis, as was sort of alluded to, we did identify a germline finding in this family, or in this proband initially, which was in the gene ERBB2, which is also known as HER2. This was a missense variant at amino acid 842, which results in a change of a valine to an isoleucine. This is found in a heterozygous variant in low frequency, so you can see this is right around 50% VAS. In this case, because this is a germline finding, we do use the ACMG criteria for germline variant interpretation. So this results in a five-tier classification from pathogenic to benign. And really, there's various discrete criteria that you would apply or not apply to ultimately result in that final classification. So what we're showing here is we actually were able to apply four different criteria, which resulted in a classification of likely pathogenic for this particular variant. So running through those classification criteria, so PS3 is a pathogenic strong piece of criteria, which we were able to apply because there are functional studies shown for this variant, which has demonstrated it to be activating in breast cancer cell lines. It has also demonstrated sensitivity to a kinase inhibitor called duratinib. We were able to apply two different pieces of pathogenic moderate criteria. So PM1, that is located in a mutational hospital, which we'll show in just a bit. PM2 is a population frequency piece of moderate criteria. And this was that it was absent from NOMAD, which is a large population database. So this is looking at a low, minor allele frequency in the population. PP3 is a pathogenic supporting piece of criteria, which is looking at computational evidence supporting a potential damaging effect for the protein. So in this case, there were 26 in silico tools that were available for us to evaluate, of which 22 or 85% predicted a damaging effect on the protein, which was, from our laboratory, enough evidence for us to apply PP3 for this criteria. So having these four different pieces of criteria ultimately classified this as likely pathogenic. So when you think of ERBB2, it is frequently described in the setting of somatic disease. So what you're seeing on the bottom is a screenshot from cBioPortal, and you can see there's a handful of somatic hotspots. So we're showing here a serine 310. The one that is highlighted by the large red arrow is our variant of interest, the valine 842 isoleucine. So it is a mutational hotspot for somatic disease, but it is less frequently described in the germline setting. However, we were able to find two different reports of this variant described in association with familial breast cancer. Providing further support for ERBB2 germline pre-exposition, this is a similar story where there is a somatic hotspot. So this is a G660D variant that has been described in a germline setting. So again, what you're seeing is a screenshot from cBioPortal, again, demonstrating that this G660D variant is seen as a hotspot, which is denoted here by this little fire from cBioPortal. It has been described in bladder cancer as well as colorectal cancer and endocervical cancers. So it is described as a somatic hotspot, but there's also a report of this particular variant being seen in the germline setting for a family that has lung adenocarcinoma. So you can see we've adapted this pedigree here from this particular publication, where they have performed targeted sequencing and identified individuals who are heterozygous for this ERBB2 G660D and who also present with lung cancer. So while this might not be currently a germline predisposition gene, it seems to be a candidate gene for a cancer predisposition. And Kristen will talk a bit more about this and the canceling implications for this family. Moving a bit forward to our somatic analysis, so you can see our filtering strategy at the bottom of the screen here for how we filter for somatic variants for tumor normal exome sequencing. This mainly relies on Mutech2, which is a way to evaluate looking for artifacts or past variants. We also look at location within the gene. So splice type, for example, would be filtered out if they are too far away from the end of the splice region. Again, we're looking at minor low frequency, looking at no match frequency. We're also looking at tumor variant allele frequency, the number of supporting reads, and then all of our variants that pass our filtering strategy do also undergo a manual review with an integrative genomics viewer to double check that they are not artifactual in nature. So what you're seeing here is there are actually seven small nucleotide variants and small insertion deletion alterations that did end up passing our filtering schematic. All of these genes using, again, that AMP-ASCO-CAP somatic variant classification are classified as a Tier 4. So Tier 4 are benign or likely benign variants. Again, these are variants we would not report back in a clinical setting or a translational setting. And all of these are variants that have been not described in association with cancer. So from the setting of this somatic analysis, looking at small nucleotide variants or indels, we did not have any variants to report back to our clinical team. Moving forward to our copy number alterations, what you're seeing here is a whole genome view of our copy number plots. So to orient you with how this plot is generated, this top half of the plot here is showing copy number change. So this zero line here is a log2 scale with data being normalized to the two copy number state, with the red line indicating different segments across the genome. So when you see a deviation above or below that zero line or that two copy number state, that would be either a gain or a loss. So this would be a gain of chromosome 1q. This would be a loss of chromosome 7p. At the bottom of this plot is our variant and low frequency. You can think of this similar to a SNP plot from a microarray, for example, where in this case, we're looking for loss of heterozygosity, where this would be a variant low frequency of 100. This would be a variant low frequency of zero. So this would be a heterozygous variant for variants that fall at either end of the scale. And then in the middle would be our heterozygous variance. The deviation between that would be loss of heterozygosity. So the best example shown here is this loss of chromosome 7p, where you would have a single copy and you see that corresponding LOH. So this particular copy number profile of a gain of chromosome 1q, a loss of chromosome 7p, and this loss of chromosome X, really is not consistent with any particular tumor type in general. What you're probably not able to make out very well from this whole genome view is this cluster of dots shown here at chromosome 8, which I actually have marked as an amplification. So if we zoom in and look at chromosome 8 in particular, again, we can make out that cluster of dots shown up here, which is suggestive of an amplification. Our single chromosomes view has a middle plot here, which is called a helper plot and allows us to look at what genes are located within the regions of gain or loss. So this is indicating that there is a focal amplification of the gene MYC within this region shown here. So we report out to our clinical colleagues a table of all corresponding copy number changes. So, again, you can see here that focal amplification, including MYC on chromosome 8q24, as well as that 1q gain, that 7p loss, and that chromosome 8 loss. In this case, we use ACMG CGC classification criteria, which is specific for somatic copy number variations to help us classify these particular variants. Again, this is in the setting of a particular tumor type, which in this case, this is still being classified as a high-grade, poorly differentiated neoplasm with 224 VRAF fusion. So in this case, we classify these as a Tier 2 variant. So variants with some clinical significance, and these are recurrent or acquired variants that have been observed in different neoplasms, but not specific to a particular tumor type. So many of us are probably well aware that MYC amplifications can be described in many different tumor types, MYC amplification in medulloblastoma, for example. We also see 1q gains in a variety of tumor types. Ependymoma is another one, for example, that has prognostic implications. So given these findings, we share these with our clinical colleagues in a preliminary setting, and they opted to reflect this to clinical testing. So a clinical FISH test was ordered to confirm that MYC amplification, which is what you're seeing on the screen here, whereby we have an LSI probe or a locus-specific probe targeting MYC, shown in red, and then a centromere probe targeting chromosome 8, which is shown in green. So again, we're seeing two copies of centromere for chromosome 8 in each of the cells that are being tested. And then we have one copy of chromosome 8 in each of the cells that are being imaged. And we're seeing that amplification of the red signal shown in what is a homogeneously staining region pattern in these cells that are shown here. So we evaluated 100 cells, and 82 of those demonstrated this MYC amplification. So this data was able to be performed in the clinical laboratory and reported within the patient's medical record. So going back to our timeline, by day 31, we were able to share some preliminary data, including that MYC amplification. Once that MYC amplification data was shared with the nominating providers, by day 46, that MYC FISH was ordered from Pathology. So the next step of our translational studies are RNA sequencing-based analyses. This allows us to look for fusions, and we do this using a composite approach of seven different fusion callers, which are shown at the bottom of the screen. We do this because each fusion caller really has their own advantages and disadvantages. And you can see, just by looking at the Venn diagram, which is shown on this slide, that each fusion caller will call a number of different fusions, with, for example, Cicero calling, in this case, 1,400 fusions, while Ariba only calling 10. The advantage, though, of using the composite approach that we are using is that we look for overlaps between all of these fusion callers. So really, in our case, only one fusion overlapped between all seven fusion callers, and that happened to be the TRM24 BRAF fusion that we also identified by our Archer fusion panel. So going back to our TRM24 BRAF fusion, we are now visualizing this in a bit of a different way, looking at the protein domains. So we have the full protein domain of TRM24 shown at the top, and the full protein of BRAF shown at the bottom. And as I mentioned previously, that kinase domain of BRAF is really the domain that we want to make sure that we are retaining to elicit those mechanistic effects and constitutive activation of the RASNAP kinase signaling pathway, and that's shown in this gold color here. So this third figure is the Chimera TRM24 BRAF fusion, and we are seeing we are retaining that entire kinase domain, and we are losing this regulatory domain of BRAF. So ultimately, this fusion results in downstream constitutive activation of those pathways that we previously described. So at this time, we still did not have our final diagnosis. However, as we mentioned at the very beginning of my part of the talk, we did have a final diagnosis right around the time that our data had finished sequencing, and that final diagnosis is a pancreatic acinar cell carcinoma. So going back and retrospectively looking at our data, pancreatic HNR cell carcinomas do recurrently harbor BRAF fusions. So I'm just presenting one paper that looked at comprehensive genomic profiling of these particular tumors, which reported that about 20% of these tumors harbor either BRAF or RAF1 fusions. Other papers, depending on which ones you read, may say this could be up to 60% of pancreatic HNR cell carcinomas. The most common BRAF fusion is SND1-BRAF fusion. And notably, TRM24-BRAF, from what we can tell, has not been described in this particular tumor type, although the mechanism of all these BRAF fusions is expected to be similar. And then I also pulled in this figure from the supplement because I thought this was an interesting way to look at the genomic data of all of these individuals. And I have one patient in particular highlighted by this red box. Oh, let me go back, I apologize. So some of these names might not come through very well, but this top arrow is highlighting the BRAF, with these yellow boxes highlighting a BRAF fusion. This bottom box down here is highlighting MYC, with this green box highlighting a MYC amplification. So this is a patient that has a similar genetic profile. It's what we're seeing from our patient. They also harbor a couple of different alterations that we were not identifying. So CDKN2AB loss, which would be on chromosome 9P, as well as a TP53 alteration. So while this is not an exact match to our patient, this is relatively similar to what we were seeing from our program that we sequenced. I also wanted to point out, though, that for MYC amplifications, there's really only two patients that harbor a MYC amplification. So this is a very rare finding within this cohort of individuals. So moving forward to the therapeutic implications for this patient. As we mentioned back with the ARCHER analysis and with our RNA sequencing-based analysis, the TRMM24 BRAF fusion has therapeutic implications for this patient. So focusing on panels C and F here from this particular manuscript, which looked at TRMM24 BRAF fusions in its cell line, we see that TRMM24 BRAF fusions do activate downstream pathways through phosphorylated ERK. And then when treated with a RAF kinase inhibitor, so it's Serapinib, or a MEK kinase inhibitor, U0126, we do see inactivation of phospho-ERK or loss of phospho-ERK. So this particular fusion is sensitive to either RAF kinase inhibitors or MEK kinase inhibitors, or MEK inhibitors, which would be something that could be under consideration for our individual. So by this point, we now have all of our data generated from a translational basis as well as a clinical basis, and it's time for us to present our data back in a molecular tumor board setting. As Kathy alluded to, this can be done really in multiple different forms. So we have a neuro-oncology tumor board, we have a sarcoma tumor board, we have an embryonal tumor board, as well as a more general tumor board that is encompassed by many different disciplines. But in all these cases, this is really a multidisciplinary setting, which includes clinicians, radiology, pathology, laboratory genetics, radiation oncology, as well as genetic counselors and surgical colleagues. So what you're seeing here is an image that shows what this used to look like pre-COVID, where we were able to visualize both radiology and genetics simultaneously. While now this has been moved to virtual meetings, this is still being done in a more integrative approach. So our patient in particular was treated with a combination of gemcytogenin and paclitaxel, as well as that MEK inhibitor, Tremetinib, for four months. So you can see, looking at the pathway here, RAF would indicate BRAF, where she is activating the pathway due to the fusion event with downstream activation of MEK, which is being targeted by Tremetinib. She did have initial interval response with this therapy, showing decrease in size of the primary pancreatic tumor and metastases, but ultimately did end up progressing within three months. So moving forward with our timeline, by day 52, we were able to issue our final report for the MEK fish, as well as by day 52 is when we had completed all of our reports back to our clinical teams, including both clinical as well as translational results. Given these discussions, the clinical team did opt to confirm the ERBB2 germline variant. This is performed using a targeted Sanger sequencing approach, which is demonstrated here, looking at this highlighted variant, which indicates its heterozygosity, shown here at the variant base pair, which is shown by the arrow, again, confirming this ERBB2 in the germline. So we were able to write a report and put this into the patient's medical records as she can undergo counseling. So finishing up our timeline here, day 74 is when we had discussion of this germline etiology of the ERBB2 variants. Day 84, this order was placed for confirmation of the variant with a day 100 being the final report was issued. So with that, I'm gonna pass it on to Kristin Zasio, who is going to discuss the counseling implications for this patient and her family. Okay, so genetic counselors typically get involved on the backend when there is a clinical finding in the germline that may be relevant and have counseling implications for the patient and or the family. So in this case, I just wanted to go back and revisit the pedigree to start. And in knowing the information that we now had regarding this germline ERBB2 variant, which has some preliminary evidence suggesting it may be related at a minimum to hereditary breast cancer and potentially other familial cancers as well. There were a couple of individuals in the family that we were thinking about, certainly our patient, but in addition, our patient's mother with the history of bilateral breast cancer, which at this point was still unexplained from a genetic perspective. And then in addition, certainly thinking about the maternal grandmother who also had breast cancer at a similar age, younger than expected. And then the family also raised concerns about our patient's brother, who was a healthy 10-year-old male, certainly less concerned about a breast cancer predisposition for him than they would be for a female. But understandably, given everything that the family had been through with their daughter being diagnosed, we're quite concerned about any cancer predisposition risk to their son. And so we had anticipated that they may have some questions about what this could mean for him as well. So thinking about the information that we had going into meeting with the family to review results, we knew that there was a germline ERBB2 variant initially identified on research-based testing, clinically confirmed in a CLIA lab. And at this point is when the family was officially notified of results. Because we had enrolled parents on study as well, we were able to confirm that the variant was maternally inherited. So again, adding to our suspicion that perhaps this could be related to those reports of hereditary breast cancer in patients with germline variants in this gene. But we also knew this is a variant in emerging gene. There's limited information available about cancer risks, whether there could be any association with a pediatric cancer with this gene, although all we had to go on was the reports of hereditary breast cancer and hereditary lung cancer primarily. And there's also no information on surveillance for this gene right now. So we don't know what cancer screenings would be recommended. We don't have any official guidelines to recommend earlier screening in a person who is positive for this variant. It's something that we know we're still actively learning about. So in reviewing this information with the family, they were certainly interested. They were obviously not strangers to cancer in the family and to wondering about hereditary cancer risk based on the history and the mother and the maternal grandmother. And initially the mother had very strong feelings about testing the patient's brother. They're a very savvy family, but the mother felt that she would want to know any information that could be available to the family, whether it's useful now or in the future. And her perspective was really that they had been blindsided by their daughter's diagnosis. And they really wanted to be ahead of anything that could be a potential cancer risk to their child or their future grandchildren. And because at this point, the patient's diagnosis and prognosis were not good, the family really was invested in having any information they could about the patient's sibling. However, the father was initially a little bit hesitant to pursue testing in the sibling. And certainly from a clinical perspective, we supported that in that we didn't have much to do if we found a positive result in the sibling. But after some extensive counseling, first with the family, just parents, then parents and our patient, and then ultimately the parents and the 10-year-old sibling, it became evident that it was quite important to the family that they know this information about the brother and that they have some sense of control over the situation, even if we don't have any surveillance or even a lot of concrete information about the variant to offer. They really were interested in having something that they could just tuck in their back pocket and potentially use at a later time if more information becomes available. So after extensive conversations and conversations with this child who was at the age of ascent, the parents and family did elect to pursue testing of the ERBB2 variant in the 10-year-old brother. Unfortunately, they were not able to test the maternal grandmother due to complications around her mental illness. And so at this time, she still has not been tested. Just to back up, and I think this was a question that came up in the Q&A as well regarding whether this gene change was on the panel that mother initially had. That was the commercially available breast panel a couple of years ago. We certainly had that question as well, and that was something that the family brought to us. So we conducted a review via Concert Genetics platform of the currently commercially available pan-cancer, hereditary cancer panels, as well as breast cancer panels. And among 119 products that contained 851 unique genes, ERBB2 was not included. And so after extensive searching, we found that ERBB2 was not included on any of the commercially available products that we queried. And so what this told us for the family was that it's not that the wrong panel was ordered or if something different would have been ordered for the mother, perhaps this would have been found. Nothing was missed. It just is new, and there's not a lot of evidence yet that has led to it being added to clinical panels yet. And it also told us for the family that this was not something that if we had elected not to test the brother, that they could go out and potentially order as a direct-to-consumer test, for example, because it isn't clinically available yet for the family to do that. So we really were holding the decision-making regarding whether or not this family would be able to access this information, which had become quite important to them. So whenever we're thinking of testing children for adult onset conditions, certainly this one is a little bit different in that we have limited evidence for this variant in terms of what the long-term cancer risks will be, but we sort of viewed it in the same frame. There are some special considerations that come up in the pediatric setting. The concerns certainly respect for siblings' autonomy or for the patient's autonomy. We think about this even in testing our patients with cancer, but certainly in pre-symptomatic testing, this is a concern. We wonder about coercion of the patient being tested. Is it something that's only important to the parents? Insurance discrimination is certainly a concern. We know that the Genetic Information Nondiscrimination Act does protect against discrimination from life insurance and employment in most situations, but it does not have the same protections for things like life insurance or long-term disability insurance. So certainly we worry about those things in pre-symptomatic testing, particularly when there's more evidence surrounding cancer risk for the gene, but that's certainly something that we thought about in this case and worked through with the family. Again, there's no clear interventions at this point if we find a patient is positive for this variant, and we really spent a lot of time reviewing that with the family, that we don't know that this variant is connected to the sibling's cancer, and we certainly don't know that this variant is going to pose any cancer risk to this patient as they get older. Right now, the relationship between the variant and disease is unknown, and so in the same vein, we also worry about providing false reassurance to a family if a patient is negative because we don't know that this explains the cancer diagnosis in his sibling, and so certainly we can't give them reassurance if a test is negative. But on the other side, we have attention given of late to the psychological benefits of knowing and the parental right to know, and so at this point, what we've recognized is that deferring testing is also a choice that impacts the child, and I think there's been a lot of evidence of late to suggest that perhaps the psychological harms that many of us have anticipated happening to children have not happened and have not panned out in terms of genetic testing for adult onset conditions, and that in some families, the ability to know and help with their coping process just by having information not withheld from them can be beneficial as well. And evidence about penetrance, natural history of gene changes, and hereditary cancer is constantly evolving and changing, and so for some families, just having information in their back pocket that they can use at a later time may be beneficial, and we're certainly learning in pediatric clinics that some gene changes that we've always traditionally thought about as adult onset, in fact, may not be truly adult onset. For example, we've seen colon cancers in patients with Lynch syndrome, which was traditionally thought of as an adult onset colon cancer syndrome, and so our knowledge and our understanding of these variants is constantly evolving. And ultimately, family dynamics, coping with a child's illness, realizing that their daughter, unfortunately, her prognosis is quite poor, and so having some control over a situation and not feeling that information is being withheld from them can sometimes impact a family's coping as well. And certainly, this is not a decision that's made by one person. It's something that we have a team, including pediatric psychologists, family psychologists, certainly the oncology team and genetic counseling teams all work with the family to help understand what makes most sense in regards to that family's particular situation. This is just a quick review. This is actually a borrowed slide, just reviewing what limited guidance we have currently from our professional societies. And, you know, traditionally the mindset was that we withhold pre-symptomatic and predictive testing of minors for adult onset conditions until at least the age of 18, which is the age of legal consent in our country. But certainly we have seen more of a shift to making exceptions when the diagnostic uncertainty poses a significant psychosocial burden for a family, or when the benefits of knowing may outweigh withholding information and what we perceive to be the benefits of that for a family. And so certainly we weighed these things as well as we were helping the family make decisions about testing. So just to sort of summarize where we were, again, the traditional approach is to avoid testing for adult onset cancer predispositions in children, but certainly with the expansion of tumor normal profiling in these pediatric cancers, this question is gonna continue to be posed. And perhaps in more extreme examples than this one, where there's still some uncertainty regarding the variant, but we will be identifying genes that are considered to be adult onset cancer predisposition genes. And we will have families who are pushing for this information and siblings, and it's going to pose challenges for genetic counseling for those siblings and for other members of the family. So I think this really highlights the importance of pretest counseling for families to really emphasize these potential outcomes upfront. This family had been counseled at length upfront and had really had some time to make decisions about how they would handle uncertainty, what they would do if this type of result came up. And so I think that was really beneficial to them in being able to make decisions rationally that fit their family versus emotionally, given everything that they were going through with their daughter. So revisiting the pedigree again, the family did ultimately feel very strongly about testing the brother, and he ultimately did test positive for the variant as well. Again, we still have not been able to test the maternal grandmother. However, we are working with the mother's providers to see if we have tumor tissue available that could be tested for her. She initially did not have any tumor testing at the time of her diagnosis. She was treated at a smaller hospital. And so that wasn't something that was routinely done, but we're hoping that perhaps that may be something that we can revisit in the future, both from a somatic variant standpoint, as well as in light of this finding. And as far as the brother is concerned, the parents have had a very reasonable approach to that and said, okay, we know. We don't yet know what that means, but certainly we have the information. And if more information comes to light, we would like to know about that as well. So that is where they left things with the brother at this time. So just to briefly summarize, I know we're up against time. This presentation in this case has really highlighted the benefits of complementary clinical and translational studies working together with both teams in order to provide these services for patients and for the entire family when a germline variant is found. Certainly translational can inform clinical confirmatory testing and the decisions around that. And there's a clear benefit to pair germline and somatic analysis that's highlighted in this case, certainly to understand which variants are germline and which are somatic and to help take action more quickly on those variants when they're identified. And ultimately broad genomic profiling does enable the identification of rare events and a spectrum of genomic alterations, all of which can present challenges and opportunities for the teams that are involved in caring for these patients. So with that, I think we just have our acknowledgments and then with the limited time we have left, I'm sure we'll be happy to answer any questions. Thank you. Thank you. Hi, thank you so much for that presentation and really interesting case. It's really great to hear kind of a little bit more about germline cancer predisposition in these cases. Just a quick announcement and then maybe we'll get to one question before we have to sign off. So ACMG Annual Clinical Genetics Meeting is March 23rd through 25th. Abstract submission opens on October 1st and those abstracts are due November 19th. So we welcome anybody who wants to submit abstracts to submit them starting later this week. We do have some questions in the Q&A. Thank you so much for going through and answering these as you've been able to. So a couple of interesting questions here. One is how much of your NGS is automated versus manual? Can you give us like the 10,000 foot, 30 second summary of how that works? Sure, I can comment on that. So it is moving towards being a bit more automated, but right now it is a highly manual process. So essentially we get a Excel sheet at this point that has germline variants of interest on it that are annotated with a variety of different links to databases. We have uranylal frequency, population frequency, those types of things. The same for the somatic, which are derived from UTEX, but it is, again, largely a manual process at this point in terms of taking a deeper dive into those variants. Great, thank you. And then I guess another question is how are germline changes handled? Do you only report germline changes when they're associated with cancer predisposition or something cancer related, or do you report kind of secondary findings that may not be related to cancer? So the patients have the option to consent to receive secondary findings as well as carrier findings. So that is an option that they can choose. With regard to what we report out in the setting of germline, really it is planted towards cancer predisposition or disease predisposition. So we have a number of different ways that we filter that, but there was a gene list established by Zhang et al. from a St. Jude publication. We also look at CDC related genes. So we're able to narrow down the list of relevant findings that we would report out. And that would be pathogenic, likely pathogenic, or variants of uncertain significance when they're deemed to be relevant. Great, thank you so much for the fantastic presentation, for answering the questions. I see that we're four minutes over already. So we'll go ahead and conclude the session, but thank you so much for sharing your expertise with us today. And thanks everybody for joining. Thank you.

clinical case study

genomic profiling

pediatric tumor

germline variant

somatic copy number alterations

counseling

therapeutic strategies

genomic testing

family decisions

pediatric tumors