Good afternoon. Welcome to the 2020 Pediatric Section Meeting, Scientific Section on Tumors. This year's session planners for this session were Dr. Paul Clemo of St. Jude's and Dr. Todd Hankinson of University of Colorado. The moderators are Annie Derpo from Nationwide Children's Hospital of Columbus and myself, Samuel Cheshire from Primary Children's Hospital. Before I introduce our first speakers, I would like to remind you that the speakers will be available in the Conversation Lounge in the Remo platform after their sessions for a Q&A and further discussions, whereas the main meeting is run in this digital platform. This tumor session will have a number of interesting presentations from leaders in our field, and we'll start with the influential paper session. The first paper, Surgically Relevant Clinical Trials being Conducted by Pediatric Neuro-Oncology Consortia, is being presented by Dr. Kathleen Dorris. Dr. Dorris is an Associate Professor of Pediatrics, Hematology, Oncology, and Bone Marrow Transplantation and the Pediatric Neuro-Oncology Fellowship Director at the Children's Hospital of Colorado. She is an expert pediatric neuro-oncologist, site PI for the PNOP, and a planning member for PSYOP. Her research is focused on the development of new drugs to treat cancer, and she is also a member of the experimental therapeutic teams at her hospital. The second paper, Influential Research Publications in Pediatric Neurosurgical Oncology, is being presented by Dr. Nolan Gupta. Dr. Gupta is a Professor of Neurological Surgery and Chief of Pediatric Neurological Surgery at UCSF Benioff Children's Hospital. He is an expert in the evaluation and surgical management of children with brain tumors. His primary research focuses on the role of inflammation in brain tumor progression, and he is a Principal Investigator of the Brain Tumor Research Center at UCSF. We look forward to all the presentations. Hello, my name is Katie Dorris, and I am one of the Neuro-Oncology and Experimental Therapeutics Physicians at Children's Hospital Colorado. My talk today will focus on early phase clinical trials and the roles that neurosurgeons may play in them. First off, I have no conflicts of interest to disclose. The process of drug development for a new medication or biologic agent is lengthy and expensive. Several phases of clinical trials are routinely conducted, beginning with a Phase I trial to evaluate safety and to establish a dosage for future use in the next phases of testing. Phase II trials are designed to identify any early evidence of efficacy and continue to evaluate safety. Phase III clinical trials compare the safety and effectiveness of a new treatment against standard therapy for a disease, and results of these trials are often presented to the FDA for licensing. Phase IV clinical trials may be conducted by companies after FDA approval to detect any rare or long-term side effects in a much larger patient population. Often, thousands of compounds must be screened to identify one drug that finally achieves approval by the FDA. For oncology drugs, the success rate of bringing a drug to market is much lower than drugs developed for other indications. Why do so many oncology trials fail? That question needs a much longer talk to explore fully. In brief, the models used to test drug effects on cancer cells are imperfect. As the molecular biology of tumors is increasingly understood through basic and translational laboratory work, the majority of the oncology drugs in development currently target presumed molecular weaknesses of tumor cells. This shift means that the tissue and tumor effects of a targeted drug must be well characterized early in the developmental process. Some drugs are tested using animal models made from cell lines that have been in existence and culture for a long time and may no longer retain the characteristics of the original tumor. Efforts to create animal models directly from patient tumor samples are made, but that approach is costly. Some types of tumors do not have xenograft models or established cell lines for drug testing due to technical challenges associated with their generation. Even with the best model systems, animal biology can never fully replicate the human system. Differences in drug metabolism and side effect profiles between animals and humans can limit how well preclinical work predicts our trial experiences when new agents are introduced into humans. Finally, the anticipated mechanism of action for tumor inhibition based on preclinical work may be incorrect, and drugs, especially for drugs that have multiple effects in cells. The added obstacle of identifying drugs that can penetrate the blood-brain barrier and get into CNS tumors adds to the challenge for neuro-oncologists. Often, CNS penetration of compounds is not tested early in development in animal models due to expense, and instead, predictions are made based on features of the compound that should or should not allow it to cross the blood-brain barrier. As this audience knows, the blood-brain barrier is complex, and our predictions of how it will behave are often incorrect. Furthermore, there are blood-brain barrier characteristics that are different between humans and animals. Neurosurgeons are important to the success of clinical trials for patients with CNS tumors in many ways. First, the tissue that's obtained at standard of care surgeries is critical to optimal patient care and is frequently needed for clinical trial enrollments. Increasingly, studies require tissue availability to screen for molecular alterations specific to the mechanism of action of the relevant drug as part of eligibility criteria. Additionally, there are essential neurosurgical rules for patient care on trials, including acquisition of tissue during a clinically indicated surgery after a presurgical course of drug to evaluate CNS tumor drug levels and evidence that the drug is acting on the tumor in the anticipated way. Another neurosurgical role is to deliver drugs by a method that bypasses the blood-brain barrier for agents that are known not to penetrate CNS tumors when delivered systemically. Phase zero standalone trials may be conducted before phase one studies or added as components to phase one or two trials, at which time can be called different names such as surgical cohorts or target validation arms to collect pre-treated tissue that can be analyzed for drug level or biomarkers that might suggest the drug is acting on the tissue as anticipated. One example of the value of a phase zero trial involves a medication called Adavacertib or AZD1775, which is a WE1 inhibitor. WE1 is a nuclear protein that has multiple functions but primarily is involved in the regulation of the G2M cell cycle checkpoint. Pre-clinical work from the colleagues in our neuro-oncology group at Children's Hospital Colorado had focused on the use of AZD1775 for treatment of medulloblastoma. Specifically, work in the Babaker lab demonstrated that survival of mice with medulloblastoma implanted in the brain was no different between cohorts treated with gemcitabine only versus gemcitabine and AZD1775. However, in animals with medulloblastoma tumors established in the flanks, the combination of AZD1775 and gemcitabine prolonged survival significantly. The conclusion from this group and from others doing work with this drug and brain tumor model systems was that AZD1775 does not cross the blood-brain barrier. A group led by the Barrow Neurological Institute conducted a phase zero trial in adults with relapsed GBM in which AZD1775 was administered to the patients 2, 8, or 24 hours prior to a clinically indicated surgery. In that trial, however, AZD1775 did show good brain tumor penetration at pharmacologically relevant tumor concentrations. One of the main reasons for the discordant results related to the animal and human trials is the presence of a transporter in rodent brains that shuttles AZD1775 out of the tumor. If researchers had relied entirely on the animal modeling experience, AZD1775 would not have been pursued for CNS tumor indications. Currently, a phase one trial of gemcitabine and AZD1775 in combination is being developed through the Pediatric Brain Tumor Consortium for pediatric patients with relapsed medulloblastoma. Similarly, two other biologic agents with concern for poor blood-brain barrier penetration have been investigated at Children's Hospital Colorado for different CNS indications. First, tocilizumab, an IL-6 receptor antagonist for cranio based on preclinical work in Hankinson Lab, and trastuzumab, a monoclonal antibody that blocks HER2 based on preclinical work with at the Foreman Laboratory at Children's Hospital Colorado. Both of the trials have treated patients prior to clinically indicated surgery and then collected tumors to evaluate for presence of drug. Both studies have identified the drug in pre-treated tumors and that work has prompted future investigation of either IV tocilizumab for cranio or IT trastuzumab in combination with GMCSF for posterior fossa ependymoma. Numerous national and international consortia have been formed to conduct phase one to two clinical trials involving pediatric neuro-oncology patients. Many of the trials in these consortia have neurosurgical involvement. To start, the Pediatric Brain Tumor Consortium is an NCI-funded consortium that recently has increased its membership to 16 organizations. There's no time to go through each trial individually, but the Pediatric Brain Tumor Consortium has a number of trials ongoing, many of which have surgical components. Next, the Collaborative Network for Neuro-Oncology Clinical Trials, or CONNECT, is a group of 19 organizations across the world, and again, many of these trials have surgical components and are focused on newly diagnosed high-risk patients. The Pacific Pediatric Neuro-Oncology Consortium has 23 member organizations across the world, and again, many of these trials have number of studies that involve target validation components or convection-enhanced delivery of agents. The National Pediatric Cancer Foundation is a philanthropic organization that has 26 U.S. partner institutions and supports early phase studies for multiple pediatric oncology diagnoses. This group has one brain tumor-specific clinical trial open currently, but has other concepts under evaluation. The PEPCTN network involves phase 1 studies through 20 different U.S. institutions and is focused on all types of pediatric oncology indications. Once promising therapies are identified through phase 1-2 studies in these early phase consortia we've just reviewed, many of the treatments are then translated into phase 3 studies in the Children's Oncology Group. ClinicalTrials.gov is a very useful website for completed, ongoing, and upcoming trials across the world. In conclusion, phase 0 clinical trials that are standalone or surgical or target validation components of phase 1 or 2 studies offer an opportunity to obtain CNS tumors that have been pre-treated with a drug to evaluate pen drug penetration and pharmacodynamic endpoints. Numerous pediatric oncology consortia conduct clinical trials focused on children with CNS tumors and of note these consortia accept clinical trial submissions or concept submissions from non-member sites and ClinicalTrials.gov is a useful resource to identify ongoing and upcoming clinical trials across the world. Thank you. I would like to thank the organizers from the AANS-CNS section on pediatric neurosurgery for the invitation to speak today. The topic I will be discussing are influential publications in neuro-oncology. I will select two papers in particular that address a link between the solutions that we have for pediatric brain tumors and the challenges that are faced. The first paper is by Fangusaro and colleagues and it was conducted by the Pediatric Brain Tumor Consortium as a clinical trial. This is selumetinib in children with BRAF abberant or NF type 1 associated recurrent refractory or progressive low-grade gliomas. The inclusion criteria of this clinical trial included pilocytic astrocytoma that had two of the most common BRAF mutations or alterations, KIA-1549 and the BRAF-V600E mutation. An additional stratum 3 were those tumors that were in patients with NF type 1. Selumetinib is a MAP-ERK kinase 1,2 inhibitor and is a one member of a class of biologically targeted agents that are directed towards this pathway. By way of background, the MEK-ERK pathway is a common and very large pathway. This is a very simplified schematic. The pathway is generally used for control of proliferation and growth and is often activated in many tumor types. This pathway is also activated in patients with tuberous sclerosis and the use of everolimus in that group is probably the most classic example of a biologically targeted agent that has achieved success. The summary of the study, which was one of the very first to show and to test and show the results of a specific molecularly focused agent, showed that about a third of patients in stratum 1 and a little bit more than that in stratum 2 had a partial response. In particular, there was a better outcome in those patients with BRAF fusion as compared to mutation. This trial clearly shows that activation of this pathway and more importantly inhibition of this pathway plays a partial role in oncogenesis of pediatric low-grade gliomas. This has led to the development of a number of phase 3 trials through the Children's Oncology Group for this disease. It is likely to be the first of many trials that will focus on the activation of pathways in a variety of tumors. This leads, of course, to the question of what are appropriate biologic targets. This is an important and helpful slide to think about pediatric tumors. This is taken from a characteristic of a mutational burden across adult tumors. The number of somatic mutation and the mutation frequency on the y-axis is in the log scale. As you can see, many adult tumors have a very high mutational burden. DIPG and pediatric low-grade glioma actually fall far to the left in terms of mutational burden and often have very few recognized mutations. This represents a challenge in terms of identifying targets for treatment. The second paper that I would like to discuss is a complex examination of the metabolic and biochemical regulation in pediatric low-grade ependymoma. This is by Anthony Michael Raj from Michael Taylor's group at the Hospital for Sick Children. By way of background, what has become clear with diffuse midline gliomas like DIPG and thalamic tumors is that while the somatic mutation burden is low, the key problem in those tumor types appears to be a series of alterations that link epigenetic regulation with the oncogenic profile of the tumor. This is meant simply to illustrate the fact that histones, which are important in packaging of DNA, are themselves tightly regulated by both methylation and demethylation, as well as other types of modifications. In particular, on the bottom schematic, we can see that the K27 position is regulated by a variety of both methylases and demethylases. It's the important thing to understand, of course, that methylation results usually, not always, but usually in inhibition of gene transcription, whereas demethylation results in a lifting of that repression. Now, more specifically, the methylation and trimethylation of the K27 position is something that is carried out by the polycomb group of repressive complexes, and this is opposed by a group of demethylases, in particular the JMJD3 group. What appears to happen, at least in DIPG and other midline gliomas, is that the mutation at the K27 position to methionine sequesters the methylation complex in that, and as such, the PRC complex is unable to place methyl marks on histone, resulting in an alteration of gene transcription. So this is a regulatory change in terms of a single mutation causing widespread changes in gene regulation. This has been, in the past, recognized as a pathway that occurs in other tumors such as ependymoma. In this study, the investigators demonstrated that the metabolic regulation of the epigenome is what drives the oncogenesis of this tumor type. In particular, there appears to be a key role for hypoxia in favoring the growth of posterior fossa ependymomas. These tumors do not show any clear copy number alterations, but hypoxia maintains an aberrant post-translational modifications at the K27 position. In particular, the metabolism of these tumors adjusts the activity of histone demethylases. Although the numbers of actual clinical specimens was quite small, what the authors clearly showed is that outcome was actually tightly linked to a pattern of gene regulation that was linked to a degree of hypoxia. In other words, the presence of hypoxic conditions leading to a characteristic gene profile was associated with a much worse outcome. So the conclusions from both of these papers is that we are in the early stages of testing and implementing targets against specific molecular pathways. And this is a cautionary tale, however, because single pathway inhibition is unlikely to achieve cures or durable control, but combination therapy is preferred. The other challenge, of course, is that pediatric glial tumors, as demonstrated by diffuse midline glioma and also now ependymoma, their oncogenic drivers of these tumors occur not at the somatic DNA level, but in fact at the epigenetic and metabolic levels. And alterations in these areas will require agents that are targeted against these specific pathways in order to achieve durable control. Of course, this is a challenge for the future, and this, at the same time, the wealth of information available regarding the biochemistry and oncogenesis of these tumors suggests that there will be progress made with these tumors once these agents are developed. Thank you very much for that very interesting discussion. I have one question for Dr. Doris. What do you feel the impact of precision medicine is going to be where each patient is going to have different candidate medications that could potentially treat them, and also how it may impact the need for more biopsy specimens, especially at the time of recurrence? That's a great question. I think as we are finding that drug development is focused really on targeted agents that will target the susceptibilities of tumors specifically, we're going to find that there's going to be a transition either to focusing on molecular changes to tumors regardless of what the underlying histology is versus trying to find very large groups of hospitals that can pull together to bring enough patients of one diagnosis type with that histologic change. It's particularly challenging in the pediatric oncology community just due to our numbers being smaller than adults. I think there are a lot of shifts. Absolutely, as I mentioned in the video, the neurosurgical involvement is key both to assess the penetration of drugs, making sure they're acting in the way we expect them to act in the tumors. Potentially, those surgical cohorts of treating patients for short periods of time before a planned surgery will help inform the drug development process dramatically. Certainly, also as we shift sometimes to drugs and agents that aren't able to cross the blood-brain barrier, there'll be a lot more potentially convection-enhanced delivery approaches or just more novel approaches to treat tumors. A lot of changes, but we will need to partner as the oncologists with the neurosurgeons a great deal as we try more novel approaches to treat a lot of our tumors better. One question from the audience. Dr. Gupta, the question is regarding utilization of vaccines formulated from an individual tumor. Do you think vaccines have added benefit for higher specificity to an individual's tumor, thus sparing side effects underpinned by biological similarity to healthy models? It's a good question. I think it gets to the point of the number of mutations and targets that are available in pediatric tumors. I think one of the conundrums for pediatric tumors is that a lot of the immune-based therapies that have been used in adult trials have relied upon adult tumors, which carry a much—GBM in particular—which carry a much higher mutational burden. I think that, yes, the answer is that I think there's an immunotherapy area in pediatric oncology is an area ripe for innovation. There are trials, for example, that have explored targeting the K27M mutation in DIPG specifically as an example of that, but one of the fundamental issues is that the second paper that I highlighted is that the actual mutational burden in pediatric tumors is very low, and it may be that the majority of pediatric tumors that are difficult to treat, like DIPG, are in fact disorders of epigenetics and metabolic dysregulation as opposed to mutations that will lead to structural epitopes that can be recognized by immune therapy. But nonetheless, I think that the advent of strategies like CAR-T cells and monoclonal antibodies towards agents is a path forward to try to both individualize treatment to specific phenotypes and genotypes of tumors, but also to reduce off-target effects. So, a corollary, you think that stimulating the innate immune system, such as NK cells or macrophages, might play a more important role in pediatric brain tumors? Yes, I think that there is a—these provide a—you know, I think there's a role, particularly in the recurrence setting, for modulation of the immune system. Again, I think that the ironic thing is that a lot of the tumors that are difficult to treat at presentation are often low-grade or lower-grade, and the—it's unclear to what degree the innate immune system is active at that early phase. But I think that the other point that I think I perhaps didn't emphasize in my talk is that there—you know, the single agents like selumetinib, I think, have shown that there's a potential for a partial response, but I think, really, long-term control and control of these tumors is going to require a multi-modality approach for each patient. A single agent is not going to work. I think modulation of the immune system, cytostatic agents, immunotherapy, in a combinatorial fashion, tailored to a patient, is going to offer that, you know, tailored to a patient is going to offer that, you know, solution long-term. We have time for one more question. What kind of efficacy are we seeing with selumetinib in refractory NF1 clinically, and is surgical resection to reduce tumor burden recommended as an adjunct to the use of selumetinib? Well, Katie could probably answer that as well as I can. You know, the responses were encouraging but modest. So, a third of patients in roughly in either stratum, the non-NF1 and the NF1 stratum showed response. The durability of the response obviously is also not clear. I think reduction of tumor burden is important when technically feasible. I think the problem is that in a lot of patients, meaningful reduction of tumor volume is a challenge, simply because they're in eloquent areas, hypothalamus, deep parts of the brain. So, you know, we can get tissue, but a meaningful reduction in tumor burden is not achievable. It's a good question though, and I think that the literature from the adult data of low grades suggests that, you know, an extensive resection is helpful, but I think it's really less well known for the infiltrating grade one tumors, particularly those associated with NF1. All right, we'll actually start moving on, but please, at any point, especially after Q&A portions are over, don't hesitate to join the conversation lounge that will bring you into the Remo platform, where our speakers will all be present to continue on this conversation, and it'll be your chance to really get in your questions with these experts. And then when you're done chatting in those conversation lounges, just join the panel right back. But at this moment, we will move on with the next section, which will present some prerecorded abstracts, and then we'll hear from an expert to give an opinion on these abstracts. So, we'll start off with Natasha Karas, an MD-PhD student, and then it'll follow with Michelle Kamita-Smith, a PGY-5 resident, and then Amanda Mu-Sorazis, attending at Lurie Children's, followed by Erica Power, also a PhD student, and they'll show us all of their great research. I'm Natasha Karas, an MD-PhD student at McGovern Medical School. The title of my talk is High Dose MTX-110 Soluble Panobinostat Safely Administered into the Fourth Ventricle in a Non-Human Primate Model. This study was conducted under Dr. David Sandberg, a pediatric neurosurgeon. As we know, new strategies are needed to treat medulloblastoma and other malignant tumors that originate in the posterior fossa in children. Systemic chemotherapy that we currently use causes considerable toxicity and is often ineffective in a recurrent setting. Thus, we believe that local delivery of chemotherapeutic agents directly into the fourth ventricle could be a novel treatment approach that may play a role in treating children with these tumors. In our study, we used panobinostat, which is an HDAC inhibitor. Thus, the objective of our current study was to evaluate the safety and pharmacokinetics of MTX-110 soluble panobinostat infusion directly into the fourth ventricle of non-human primates. For this study, we used four rhesus macaques that underwent posterior fossa craniectomy and catheter insertion into the fourth ventricle. We had two groups, a short-term group and a long-term group. In the short-term group, the fourth ventricle catheters were externalized and a lumbar vein catheter was placed simultaneously for short-term infusions. This group received five consecutive days of MTX-110 infusions directly into the fourth ventricle. In the second group, the fourth ventricle catheters were connected to a subcutaneously placed port for long-term infusions. These animals received four cycles of MTX-110, each cycle consisting of five daily infusions, such that we had one week of infusions and one week off that occurred over the course of eight weeks. After the infusions, we measured serial CSF and serum panobinostat levels. Animals also underwent a detailed neurological exam, MRI scans, and post-mortem histological analyses. These MRI scans were obtained after MTX-110 infusions. Images A and B are T2-weighted MRI images, and the white arrow here points to the catheter and shows that the catheter was placed within the fourth ventricle. Panel B, again, is a T2-weighted MRI image showing that the catheter was appropriately placed in the fourth ventricle. Lastly, panel C is an MRI flare image, and you can note that there are no obvious signal changes. We also performed histological analyses. On the left is a picture of the gross brain with an arrow pointing to the catheter placed within the fourth ventricle. On the right is an H&E stain of the brain showing the cerebellum, the pons, and the fourth ventricle. The zoomed-in inset shows that there were some inflammatory changes noted near where the catheter was placed. However, there were no other cytoarchitectural changes noted in the brain. We also evaluated the other organs throughout the body and did not note any cytoarchitectural changes. After panobinostat infusions, we measured the levels of panobinostat in the fourth ventricle and the lumbar cistern. As you can note by the red trace, right after the infusion, we were able to achieve supertherapeutic levels of panobinostat in the fourth ventricle that declined over the course of the next 12 hours. Interestingly, we were also able to detect panobinostat in the lumbar cistern up to an hour to two hours after infusion. However, the levels of panobinostat in fourth ventricle were 600-fold higher than the panobinostat levels in the lumbar cistern. We also evaluated panobinostat levels in the long-term group, which received four cycles of panobinostat infusions. As you can note, directly after the infusions, we were able to achieve supertherapeutic levels of panobinostat, and 24 hours after each infusions, the levels were undetectable to about nine nanograms per milliliters. So to summarize, we showed that soluble panobinostat can be safely infused into the fourth ventricle in non-human primates at high supertherapeutic doses, as indicated by our MRI scans, histological analyses, and neurological exams. Post-infusion CSF panobinostat peaks immediately in the fourth ventricle and then rapidly decreases over 24 hours. The panobinostat level in the fourth ventricle is 600-fold higher than that in the lumbar cistern. These results will provide background data for a pilot clinical trial in patients with medulloblastoma. In fact, we have already recruited our first patient for a pilot clinical trial. Thank you for your attention. Good afternoon. My name is Michelle Kameda-Smith, a PGY-5 in Neurosurgery at McMaster University in Canada. I'd like to thank the organizers for the opportunity to present my thesis work titled Integrative Multiomic Analysis of MSI-1 Function in Brain Cancer Reveals Context-Specific Downstream Targets for Drug Discovery. I have no conflicts of interest to disclose. Pediatric medulloblastoma is the most frequent malignant CNS tumor and represents 20% of all childhood brain cancers. High-throughput transcriptomic analysis substratified these tumors into distinct subgroups with a consensus statement in 2012 identifying four molecular subgroups, with group 3 representing the children with the poorest prognosis and frequently presenting with metastatic disease. Analysis of dysregulated pathways in medulloblastoma further subdivided group 3 and B into three subtypes with an over-representation of dysregulated RNA processes and translation. Key biological players in these processes are RNA-binding proteins, and specific to neural system, the neural RNA-binding protein SASHI-1 became a prime candidate for further investigation. Repositive transcriptomic data identified high expression of MSI-1 in group 3 medulloblastoma and patient's Xenograft models identify a similar pattern in the protein level in group 3 and B as compared to non-group 3 subgroup cell lines, with high expression of MSI-1 associated with cancer stem cell function. To answer this, short hairpin knockdown experiments were carried out and we observed that even with modest inhibition, cells were phenotypically smaller, with reduced ability to form secondary spheres and proliferate. When these cells were injected into the cerebellum of immunocompromised mice, we observed large tumors in the control mice as compared to the MSI-1 inhibited counterparts, which translated into significant survival benefit in multiple group 3 and B patient-derived cell lines. Next, I wanted to determine how MSI-1 was regulating the cancer stem cell phenotype in group 3 and B, and undertook an unbiased large-scale screen to identify MSI-1 binding targets in normal and cancer neural stem cells using an enhanced cross-linking and immunoprecipitation and sequencing technique called eCLIP. In group 3 and B and NSC, MSI-1 were bound to the 3' UTR and to its known consentence sequence, and we find that MSI-1 shared and had unique binding targets between group 3 and B and neural stem cells. Papua's analysis identified in purple the processes uniquely dysregulated in group 3 and B as compared to neural stem cells in yellow. Here we observed genes associated with chromatin remodeling, transcription, and translation to be dysregulated, and also dysregulation of known pathways, namely the JAK-STAT and TGF-BETA pathways. Because MSI-1 was previously shown to be an essential gene in neural development, I set out to identify targetable downstream MSI effector genes for therapeutic discovery. While DNA is transcribed faithfully to mRNA, not all mRNA is translated into protein. And is that because binding of an RNA-binding protein is degrading, stabilizing the transcript? And if so, is there a protein product? And if not, is that because of translational dysregulation? To answer these questions, I designed three high-throughput experiments to quantify these changes after MSI-1 knockdown at the transcriptomic, transectomic, and proteomic milieus. After MSI-1 knockdown, our RNA sequencing experiment found that there was thousands of differentially expressed genes with downregulation of multiple genes associated with known group 3 and B pathways, such as the TGF-BETA and GABA receptor activation pathways. Most were not MSI-1 bound, but those that were and downregulated, were they degraded? And if upregulated, were they stabilized for translation? We then pursued polysome profiling experiments to identify those mRNA associated with the polysome fraction that are being actively translated. We identified a number of cancer-annotated genes in pink that were downregulated after MSI-1 knockdown and those that were upregulated. And if they were upregulated, were they promoted or merely stalled in translation? To answer this, we subsequently carried out our tandem mass spectrometric experiment to quantify the changes at the proteomic level after MSI-1 knockdown. We observed downregulated and upregulated proteins. Protein enrichment analysis identified proteins associated with tumor aggressiveness to be downregulated and those associated with differentiation to be upregulated after MSI-1 knockdown. To put all this data together, we performed multi-platform analysis to identify a group of genes, one of which was HIPK1, a homeodomain-interacting protein kinase 1, as a group 3 medulloblastoma E-bound downstream gene for targeted discovery. The transcriptomic expression of MSI-1 highly correlated with that of HIPK1 expression in primary medulloblastoma samples. This protein in the literature seems to have a context-specific role, where in certain organ systems it's identified as an oncogene and in others a tumor suppressor gene. What is functionally known about HIPK1 is that it induces phosphorylation of a gene called CREB. Interestingly, microdeletions of CREB is associated with Rubens-Tavi syndrome, a syndrome suggested to predispose to medulloblastoma, specifically group 3 medulloblastoma. In conclusion, the RNA-binding protein Musashi1 is essential in tumor propagation in human models of group 3 medulloblastoma, and our large-scale unbiased comparative multi-omics study of MSI-1 function identified a group of novel candidate genes, one of which was HIPK1, which may be clinically relevant in group 3 medulloblastoma genesis and is a potential target for therapeutic drug discovery that may also spare normal neural developmental processes. Thank you. Today I'll be speaking on the epigenomic landscape and 3D genome structure in pediatric high-grade glioma. I have no disclosures. Pediatric high-grade gliomas, including GBM and DIPG, are highly morbid brain tumors. Up to 80% of DIPGs and 50% of GBMs harbor a mutation in genes encoding histone H3. We sought to investigate whether these mutations are associated with distinct chromatin structure that alters transcription regulation. To do so, we utilized a cohort of cell lines and frozen tissue specimens for integrated RNA-seq and ChIP-seq, as well as the first-of-its-kind Hi-C and Atac-seq analyses. Principal component analysis of transcriptome data shows clustering of technical replicates and diagnoses, while sets of differentially expressed genes were identified in DIPG, GBM, and normal cell lines. Unsupervised hierarchical clustering of RNA-seq data on the left and genome-wide H3K27 acetyl signals on the right revealed tumor-specific clustering across sample types and similarity of the transcriptome and H3K27 acetyl landscapes. H3K27 acetyl ChIP-seq revealed DIPG and GBM wild-type and H3G34V mutant-specific distal enhancers. H3K27M protein also co-enriched with these enhancer signals in DIPG cell lines. Representative genome browser views shown here demonstrate these distinct patterns of H3K27 acetyl enrichment at DIPG and GBM-specific enhancers. These data also revealed enrichment of tumor-type-specific transcription factor motifs on the left and distinct biological processes on the right. Through Hi-C analyses, we found greater DNA interaction frequency and intensity, as well as subtads around the MycN region in DIPG as compared to GBM in normal cells seen on this contact map. Further, aggregated signals from DIPG and GBM-specific loops revealed predominantly promoter-promoter interactions in both tumor types. We also found that DIPG-specific DNA loops linking the Olig2 promoter with a distal enhancer within the distinct gene body were not observed in GBM or normal cell lines. Tissue attack C signals of the same region revealed higher chromatin accessibility in Olig2 than DIPG compared to normal brain tissue from the same patient. Lastly, we identified multiple novel enhancer hijacking events in DIPG cells. For example, an inversion and deletion event as well as copy number gain were all observed over the SYBU region, while an inversion and interchromosomal translocation was observed at the A2M gene. In conclusion, we have shown genome structural variations and enhancer-promoter interactions that impact gene expression in pediatric high-grade glioma and may thereby contribute to gliomagenesis. As such, further studies examining the impact of these alterations are now underway. I'd like to thank you for your attention as well as to thank my lab, collaborators, funding sources, and most importantly the patients and families that supported this work. Thank you. Hi, my name is Erica Power and I'm a third-year PhD student at the Mayo Clinic and today I'll be sharing my new xenograft model for DIPG called the Chicken Embryo Chorioallantoic Membrane or CAM model. I have no disclosures. DIPG or Diffuse Intrinsic Pontine Glioma is a deadly pediatric brain tumor with a median survival of 9 to 12 months after diagnosis. Nearly all DIPG tumors harbor the signature H3K27M mutation, which is believed to be a driver of tumorogenesis. Despite nearly 100 clinical trials, there is no treatment or cure for DIPG. Nowadays, more and more patients with suspected DIPG tumors are being biopsied, which has led to an increase of tissue and thus patient derived cell lines for study. Preliminary studies are done in cell culture and then moved into rodent models, but these rodent models are extremely challenging, requiring significant amounts of time, money, and other additional resources. Therefore, there is a need for an intermediate model that could be used to help bridge the gap between in vitro and in vivo studies. The CAM model is one such intermediate model that has previously been reported for use in other cancers. The CAM is a vascular-rich membrane encasing the chicken embryo that we can implant patient-derived tumor cells on. Within 48 hours, visible tumors are formed. Through IHC staining, we can see that these tumors maintain the genetic and epigenetic landscape of their native tumors, such as here where you see the tumors are positive for H3K27M. These tumors can be subjected to new therapies and their tumor volume and vasculature assessed by 3D ultrasound. The entire experiment takes 14 to 16 days. CAM tumors can be subjected to topical drug treatments. Here we have treated two different DIPG CAM tumors with one of four drugs every 48 hours for six days. In the following day, the tumors underwent 3D ultrasound where their volume was quantified. Each of these drugs were toxic to cells and culture and are being considered for rodent models in clinical trial. We found that aliceratib, panatnib, and MSN1 lute were potent to DIPG13 CAM tumors based on a significant decrease in tumor volume following treatment. The ultrasound can also be used to assess tumor vascularity and any changes we might observe following drug treatment. Radiologic evidence would suggest that DIPG is not a very vascular tumor and our results are indicative of that. Majority of CAM tumors show a tumor vascularity of 10% or less. To compare, pancreatic cancers are between 40 and 50%. While drug testing has been done in other CAM cancer models, no one has explored radiation therapy. Since most patients with DIPG undergo radiation, we were keenly interested in being able to use the CAM model to assess radiation. Using the rodent XRAD irradiator, the CAM tumors were subjected to a single targeted dose of two-grade radiation, although we did not see any significant decreases in tumor volume. In conclusion, our data suggests that the DIPG CAM model may be useful for bridging the gap between in vitro and in vivo studies. Here are my acknowledgments. Thank you very much. All right, I was just going to introduce you real briefly. Dr. Sui Dan is here to just give us his opinion on these talks. And also just a quick housekeeping, the speakers of these abstracts, please join the conversation lounge so that after hearing your thoughts and opinion, everybody can go into those lounge to get those questions answered directly with the speakers. And you can find that on your agenda, click on the link, and then you'll sign up in the remote lounge and then just come back to this session. All right, take it away, sorry. That's quite all right. Thank you, Annie. I first want to start with a huge congratulations to all the presenters in this session. I've watched these now probably four or five times since last week. Overall, excellent presentations. And I think it represents an amazingly robust cross-section of translational science and the world of neuro-oncology and really high caliber science. I want to compliment all of them also for their effort in focusing on a clinical space that warrants a need for further advancement in therapeutic direction, EIPG, high-risk, or excuse me, high-grade three ailment kids and high-risk medullos. So fantastic clinical interfaces. And I also want, and I think it's critical to acknowledge the senior authors in all these abstracts really because they're driving efforts, which is really a testimony to our profession in pediatric neurosurgery. And I look at this with admiration and continue to applaud their efforts to mentor physician scientists moving forward. So thank you. In the first abstract by Kras et al, they're capitalizing on a local regional drug delivery platform, intrathecal delivery. Nothing new, but certainly new in the realm of molecular stratification of tumors. And in the world of pediatric brain tumors, really championed by the senior author, Dr. David Sandberg already into a clinical space. You know, they've done a really good job of, I think for the first time, demonstrating some differential concentrations with regard to the preference for fourth ventricular administration over those in the CSF lumbar system. So I compliment that result. The rationale for a PNHDAC inhibitor in medulloblastoma I may have missed in the abstract, but the epigenetic dysregulation in medulloblastoma and maybe not as strong as some of the other tumor subtypes that have been championed with the HDAC inhibitors. And then the other, I think, important point is the short-lived therapeutic concentrations in their chronic administration arm begs the question of how to schedule in a translatable format. The abstract by Kamita Smith et al, you know, wonderful effort in drug discovery in poor risk category of group three medullo. Again, a place where there is space to improve. The expression of MSI-1 seemed to be higher in the group four versus the group three. This was acknowledged, but it was unclear why the so much emphasis on the group three translational effort. But this merges really well with an expanding domain of transcriptional dysregulation potential therapeutic targeting. So I expect that this will probably even crosstalk with a lot of efforts in the DMG world. I always love listening to Amanda Saratsos' talks on the epigenetic profiling of high-risk DMGs. Really a very unique space to go with 3D chromatin structure. And I think it's great that she's developing and looking at other efforts for therapeutic targeting versus some of the other more conventional means that have been broadcast of late. So I compliment her on this effort, which is novel, and look forward to that moving forward. And then lastly, the group from Mayo, Dave Daniels and Power, who presented this really lovely abstract on utilizing an assay that was unbeknownst to me with regard to a translational platform and in vitro or bio in vitro platform, I think was really intriguing. Listening to this, I was very engaged in the idea of incorporating it in my own laboratory, but there's a lot of unanswered questions regarding tumor microenvironment and immunologic integrity as we look at some therapeutic platforms that are fastly becoming of interest. Then the last comment with regard to that abstract, the idea of using a topical administration for drug screening in this CAM assay, again, I think would beg the question of how that translates into a clinical realm with a systemic versus local or local regional drug delivery. So across the board, wonderful abstracts. I really enjoyed listening to them. I congratulate all the authors. Excellent points. And then I'll just read, I guess, two of the questions that were brought up here. One was actually a question that I also had, and then maybe you can comment on that, and then we'll encourage people to go into the conversation lounge and ask these questions directly to the abstract speakers. So the first question was, was there any, for the first abstract by Natasha Karas, was there any assessment done to confirm the presence of penobinostat in the brain parenchyma or to show whether the drug reaches the tumor tissue? So I agree, I think, you know, I don't know if they didn't have time to present it in this study, but I'm sure that they're thinking about that and they anticipate this question of saying, you know, how can we measure what assay to use and have they measured it in the brain parenchyma? And then I would also kind of want to know in a tumor model as well in the tumor tissue to think of the blood brain barrier. And then same as just see in the blood, you know, how much of that drug makes it to the vascular system. So thinking of systemic side effects. And then the other question that I'll just mention real quick, and then we'll, you know, hopefully not use too much time of the conversation lounge, but we'll hear from you, is for Dr. Powers, is there a controlled study quality assurance at the end of the CAM model that suggested the IPG CAM model's genetic background is the same as the beginning of the study? How do you know there are no other driving mutations? So one in terms of tissue samples and measures, and then the other changes in the CAM model. So I don't know if we want to have more discussion over here or send people over to the conversation lounge. We'll send you all over to the remote conversation lounge if you don't mind. And after that, there will be a quick break of 15 minutes, and then we'll pick up right back up with a very interesting case presentation by Dr. Clemo. And so that is planned for around 2.20, 2.22 PM. So if you want, please go ahead, join those conversation lounges, talk with the speakers and we'll see you soon. Thank you. All right, welcome back, everyone. I hope you were able to chat a little bit in the conversation lounge. So next up, we have a very interesting case presentation by Dr. Clemo from Memphis. And in the hot seat to kind of give our thoughts on the management of the case as he presents it, you'll have myself, Dr. Sheshire, and Dr. Sandberg to provide some input on the case. So take it away. All right, thank you, Annie. So I wanted to present a recent international patient of ours, it's a nine-year-old female who came from India. Back in March, she started complaining of some headaches and neck pain. And over the next four months, her symptoms persisted. She started having some vomiting. Her parents noticed that she was often squinting, complaining that she couldn't see small print. She was holding her electronic devices much closer to her face than she would normally do. As some of these kids, unfortunately, what happens with some of these kids, unfortunately, they kind of get bounced around. So she was seen by her pediatrician several times. She was thought to have a UTI. So they gave her some antibiotics for that and then reflux for another 14-day course of some medicine. She was eventually referred to a vision hospital close to her home because of these visual complaints. And I was unfortunately diagnosed with vision loss and pathoedema. The eye doctor ordered an MRI. I have some notes from when she initially presented there. And in India, she was awake and alert. Her vitals were stable. She didn't have any concerning vital signs, no bradycardia. It was very difficult for her to track because of her poor vision. Her pupils were poorly reactive. She was quite ataxic, but that also was difficult for them to assess because of her compromised vision. This is several shots of her MRI. This is a T2-weighted images, which shows a large fourth ventricular mass measuring about five centimeters. These are T1 with contrast. The left and middle images are the T1 with contrast and the far right is a flare image. So she has these kind of bands of enhancement. I guess you could describe it within the tumor, but for the most part, the tumor is non-enhancing. And of course, she's got pretty significant hydrocephalus with a lot of transpendymal flow. Spinal imaging was done. She had a clean spine imaging, no evidence of any drop mets. So this will be the first break point for this case presentation. And I thought the first thing we could talk about is how to manage her hydrocephalus. So the first poll that I would present is the treatment options for her hydrocephalus. The first one would be to place an EVD now. Second would be to wait until you're in the OR with her. Third is to put it in the OR only if needed. Fourth would be a shunt and fifth would be an ETV. So I think the poll could be sent out. And while we're waiting for the results on the poll, I will ask Jenny and others, and anybody can chime in, I guess, on how they would manage this girl's hydrocephalus. Perfect. So you guys should be seeing the poll pop up. So feel free to answer that question. I mean, I think to me, the main things that decide the way I've managed her hydrocephalus and the way I've managed these patients coming with severe hydrocephalus and a posterior plossa mass is to stabilize them clinically first. And then, I mean, if they are having significant vital sign changes or altered mental status, I would place the EVD ahead of time. The next thing that kind of drives me to decide the timing of EVD on that day that they're admitted versus in the OR is how quickly will I get the OR and also a little bit of time of examining them. So in terms of my answers, they kind of go between one or two, depending on clinically how that person shows up and when the OR is available. Personally, I wouldn't go with EVD just if needed. I think she has pretty severe hydrocephalus and trying to manage and putting in an EVD real quickly while you're trying to open that posterior fossa, you know it's going to herniate in your face when you're trying to open. So I would say probably a little bit more conservative of putting an EVD either before doing the craniotomy in the OR or the day before or whenever they come in, depending on how the patient is clinically. The other options, I kind of want to be more conservative of saying I want to give them the chance to not have a permanent shunt and would want to have the EVD in place to be able to wean them afterwards and assess if they really need a shunt permanently. So that's why I would go probably with an EVD at first and see afterwards. Yeah, I in general agree with you. At least here, we don't tend to put EVDs up front for this particular patient with her imaging, with her presentation. I would probably plan for an EVD in the OR. So I would get her positioned for the tumor resection, have a burr hole ready to go if I needed. And then I would see how the dura felt. If it was very full, then yeah, I would just put one at the time of the resection but ETV, I think is a reasonable option. But unless someone is really compromised in terms of their vitals and their clinical presentation, I don't tend to put an EVD up front. I agree. Yeah, and then I guess for the sake of time of moving on with the rest of your interesting case, the poll results I would say right now as the results are coming in, we have at 38% external ventricular drain in the OR, 36% EVD in OR if needed. So it kind of goes along with what we're talking. And then 23% would have done the EVD now at time of presentation, 2% ETV and 1% shunt. So I think we'll close those results. Interesting. Okay, I will continue. So I just wanted to bring up this really nice paper from Jay's group. It's the, I'll have to remember, it's the Canadian PRE-OP prediction rule for hydrocephalus. So for this particular patient with her papilledema and also transependent flow, moderate to severe hydrocephalus and her diagnosis of medullo, it would give her a six month probability of persistent hydrocephalus of about 42%. I thought this was a really nice tool to use in order to try to guesstimate what the chance of still needing a shunt or some kind of CSF diversion after surgery. So what was done was a shunt and this is not unusual. We've had a number of kids that come to us from India and that seems to be their management strategy. They put the shunt in first on 7-27 and then a few days later, they take the kid to the OR for the surgery. They put in a Chhabra, I'm not sure if that's the correct pronunciation, but they put their shunt in, no post-operative complications. So which brings us to the second break point and this is something that's always been something I've wondered is whether or not as surgeons, we can try to mitigate or lower the risk of the child developing a posterior fossa syndrome. So again, we'll do another poll. So the question is, does surgical technique or experience impact the risk of a child developing posterior fossa syndrome? So the options are yes, for sure. No, definitely not. And then number three is a maybe. So again, as people are doing the poll, we'll switch over to the great Dr. Sandberg to give us his thoughts on this question. So the question is, does surgical experience impact whether you have posterior fossa syndrome? I don't know that I can answer that question, honestly. I mean, I think all of us would agree that surgical experience matters for these challenging posterior fossa tumor cases. And I think we all have different strategies. It's controversial even what causes the posterior fossa syndrome. There are some who would argue that it's retraction of the brain and one should do a telovailor approach in order to mitigate that. And there are others who insist that it has to do with the extent of resection of the vermis, actually, and you should resect as little of the vermis as possible. I don't think there's definitive proof either way to be truthful. I mean, my personal preference is always to do a telovailor approach to resect no vermis if possible. And that's what I do. Yeah, I would agree with that. I wanted to present this paper from Bob Keating's group. I thought this was a really, really neat paper, and I fully agree with their premise that there appears to be a two-hit phenomenon for these kids. I think they do come in with certain risk factors, a large midline tumor, medullopatients seem to be at higher risk than other types of tumor, whether there's a lot of splaying of the peduncles, infiltration into the brainstem. I think those are kind of risk factors that are not necessarily modifiable, but I do think that there is something to be said about surgical technique and experience because this posterior fossa syndrome is really a surgical problem in the sense it only occurs with surgery. And so they found that minimizing retraction and minimizing the incision or an absent of, or avoiding an incision in the vermis really helped them minimize their risk of a posterior fossa syndrome. And I would agree with that. I think I never use fixed retractors. I really try to work around the tumor as much as I can, devascularize it once I feel like I've come around it as much as possible. I'll do an internal debulking. I try to move as quickly as possible. I think speed of surgery does to a certain degree help. Once I start to get out to the boundaries, then I really slow it down. I really try not to extend any kind of resection, retraction beyond the ependymal borders. And I think it's worked, that kind of strategy has worked fairly well here. Our risk of a significant posterior fossa syndrome for these kinds of tumors, I think is less than 10%. So I do think that there's, like I said, there are risk factors that are not really modifiable. A lot of it depends on imaging, but also probably age of the patient and other demographics. And then there are things that I think we can control at the time of surgery. Do we have the poll results back? The poll results kind of go along with what we just discussed too. 50% said yes, 40% said maybe, and 10% said no. Very interesting, okay. Great, So I'll keep moving along here. So she finally comes to us about a month after her initial resection. All of her wounds are healing up well. Unfortunately, her vision is no better. She's quite impaired. She can see a few things. She has to hold the phone really close to her eyes. Fortunately, she did not develop a posterior fossa syndrome after her resection. Her pathology, of course, it's histologically a classic variant. On a molecular level, it's a non-WNT, non-sonic hedgehog, p53 wild type. There's no MYC or MYC-N gain or amplification. She's got negative CSF cytology, and she has new imaging done. So the new imaging shows that they were very good at resecting the portion of the tumor within the fourth ventricle, but there is a sizable residual in the right cerebellar medullary angle. It measures about 2 by 1.4 centimeters. The left is the T2. This is the T1 with contrast, but subtracted image, and here's a regular T1 with contrast on the right. Again, you can see these veins or streaks of enhancement within the tumor, and I will also draw your attention to the side of the medulla and the appearance of the tumor and its attachment to the side of the medulla. It has almost a little bit of an irregular fuzzy border there. So the natural next question would be for this child with a non-WNT, non-sonic hedgehog tumor is, do we take her back to the OR for that 2 centimeter residual tumor? And we'll have another poll. Question is, do we reoperate or do we just have her start her adjuvant therapy? So option number one would be to reoperate. Number two, proceed with adjuvant therapy, and number three is not sure. I'm sorry to interrupt, and I know we're going to hear from Dr. Shashir. We only have about 2-3 minutes before we get back to the abstract, so hopefully we can get to the end of the case. Okay, so while we wait for the third poll, Sam, do you want to give us some thoughts? So just briefly that I think we should all try to achieve safe surgical resection and weigh the risks of doing the reoperation versus the benefits of doing the operation. I think the best study done thus far in this topic, it was in 2016 by Dr. Mike Taylor in Toronto, published in Lancet Neuro-Oncology. They did 700 patients. There was close to 800 patients. They molecularly phenotyped them, and they showed that there was a progression-free survival with gross total or near total resection compared to subtotal resection, which this would be because it's at or greater than 1.5 centimeters squared. And in the group four group, there was an actual survival benefit for gross or near total versus subtotal. So this patient would get benefit by achieving at least a near total resection. And I even think progression-free survival is also worthy in her considering she may go back to India and having a longer time period without progression is good for her as well. Okay, wonderful. Do we have the poll results? Yes. Oh, Sam, do you want to go ahead with that one? So poll results for this. 46% would say start adjuvant therapy, 45% will reoperate, and 9% are not sure. Wow, that's a little surprising to me. Yeah, I definitely feel that she should be taken back to the OR. What we always try to achieve is no measurable disease prior to starting adjuvant therapy. I'll just briefly mention that the whole notion of the 1.5 centimeter square was from a 1996 paper, so it's almost been 25 years since we've been dealing with this question of whether to take kids back or not. So in the interest of time, I'll kind of go through the rest of it. So we did take her for a second look operation, but unfortunately, her tumor was very adherent to the side of the medulla. I could not really do much with it. So she actually went on to get some chemo, and after her chemo, she had a new scan which showed that her residual tumor had pretty much been taken care of by the chemo. So the chemo was able to achieve what I couldn't surgically for her. I'm not sure we have time for this, but what do we do with the shunt? So I'll do another. Do we have time for another poll, or should I just go through? Just show us the rest of the case, and then if people have questions later on, we can direct them to the Conversation Lounge. All right, so what I did with the shunt, I thought that she had a reasonable chance of not needing a shunt, so I externalized her shunt at the time of her second look surgery. After a few days, we clamped it, but unfortunately, she started to get headaches and a pseudomeningocele, so I put a new shunt back in. And as a follow-up, she has just completed her radiation therapy. She'll have a six-week treatment break, and then she'll start four rounds of chemotherapy and complete her adjuvant therapy sometime in the spring, and then hopefully, she will be back home in India later in the spring and summer. And that's it. All right, well, thank you. That was a great, interesting case, especially coming from elsewhere, so it exposes us to different realities. Thank you for that. So again, to everybody in the crowd, please, at the later breaks, join the Conversation Lounge, and hopefully, Dr. Clemo will maybe pop in there to answer some questions if people have some for your case later on, and I will give it away to Sam to move on as well. Thank you very much. That was a very interesting discussion. Now, for our next abstract presentation, we have three abstracts. The first one is by Raul Kumar, Patient-Matched Molecular Comparison of Primary and Relapsed Medulloblastoma. The second one is by Ricardo Serra, Disulfame and Copper Combination of Prolonging Group 3 Metablastoma through NPL-4 Inhibition, and a paper, an abstract on subgroup divergence at medulloblastoma relapse by Raul Kumar. I hope this message finds you well. My name is Raul Kumar, and it's a privilege to present our work describing comparisons of the molecular landscape between primary and relapsed medulloblastoma. A third of medulloblastoma patients will experience relapse, which confers an abysmal prognosis with only around 10% of patients surviving five years later. Previous studies exploring the genomic landscape of relapse disease have uncovered key themes in our understanding of the molecular changes at the time of relapse. However, many of these studies have been limited to modest cohorts and have employed a variety of molecular techniques. Additionally, novel descriptions of intertumoral heterogeneity in the form of molecular subtypes have not yet been explored in the context of medulloblastoma relapse. In this study, we profiled 127 patient matched diagnostic medulloblastoma and relapse tumors by DNA methylation array and next generation sequencing. Using robust bioinformatics classifiers, we were able to confidently segregate secondary malignancies from true relapse. While we identify subgroup conservation in 96% of cases, the conservation rate of novel group 3,4 and sonic hedgehog subtypes at relapse was lower at approximately 80%. No specific pattern of subtype divergence was noted. However, cases with group 3 or 4 subtype divergence at relapse seemed to be enriched for the presence of MYC or chromosome 2p amplification in relapse tumors. Additionally, we annotated the landscape of driver gene and chromosome arm copy number variations for our entire cohort. Synthesizing these data, we uncover subgroup-specific patterns of driver gene alterations and chromosome arm copy number variations. For driver gene alterations, sonic hedgehog tumors had a greater proportion of conserved alterations. For chromosome arm copy number variation, a greater degree of conservation was noted amongst group 3 and group 4 tumors. Overall though, the proportion of primary specific alterations was low. Finally, we wish to highlight that secondary malignancies, namely high-grade gliomas, may masquerade as relapse medulloblastoma. An exemplary case is shown where a patient with a gross totally resected primary group 4 tumor experienced a new lesion adjacent to the local tumor bed. The sighted genetic landscape of the secondary malignancy clearly demonstrates meta-amplification and homozygous CDKN2AB loss characteristic of radiation-induced glioma. These results highlight the utility of molecular confirmation in putative relapse disease. In conclusion, we largely recapitulate the previously described phenomenon as subgroup conservation at relapse, with rare exceptions. Additionally, we identify 80% conservation rate for novel group 3, 4, and sonic hedgehog subtypes. Additionally, we uncover distinct patterns of molecular alterations within each molecular subgroup, with variations in both driver gene and chromosome arm copy number alterations. Finally, we wish to emphasize the utility of tissue sampling at suspected medulloblastoma relapse, particularly given the appreciable incidence of occult secondary malignancies. Additionally, relapse tissue will facilitate a deeper understanding of the molecular changes that may be driving relapse or therapy resistance. Good afternoon everyone, I'm Ricardo Serra, neurosurgery resident at the University of Maryland Medical Center. Today I'm going to present 12-day sulfur and copper as a new combination therapy in prolonged survival of group 3 and group 2 SHH medulloblastomas through NPL4 inhibition. I have no disclosures. First, I would like to give a brief introduction to two of the most aggressive subtypes, group 3 and SHH driven medulloblastomas, especially those with p53 mutation. As you all know, we don't have a lot of good options for treatment of these subgroups, and although they don't constitute the biggest subgroups of medulloblastomas in terms of just sheer numbers, the survival of these patients is often lower than that of other medulloblastoma patients. So the goals of our study were basically to test a new combination of safe and FDA-approved therapeutics such as disulfur and copper to assess their toxicity in vitro and then apply that in vivo into xenograft models of medulloblastoma, and finally to evaluate ex vivo the markers of cell stemness, DNA damage, and NPL4 inhibition. Here we can see the cytotoxic effect of disulfur and copper at nanomolar and micromolar concentrations. You can see that the IC50 is usually around 100-150 nanomolar. We then turn to flow cytometry, and we want to look at the fraction of double positive and nexin-5 NPI cells that represent the necrotic fraction after 12, 24, and 48 hours, and you can see that most of these lines, cells tend to die fairly early, and after treatment, usually about 12-24 hours, they already show some signs of necrosis. We then turn to markers of cell stemness, in particular CD133 and nestin. We also tested aldehyde hydrogenase, and we showed with flow cytometry that there's a decrease in double positive cells after combination treatment, as well as in cellular expression of aldehyde hydrogenase. With western blotting, we wanted to demonstrate that certain patterns of apoptosis are enhanced after treatment, and you can see here the glyph bar fraction that's increased, as well as the AIF after combination, and then the main targets of disulfur and copper, NPO4, that activates the shock response is increased, as well as NRF1 that is cleaved in different subforms that are not usually present in normal cells. DNA damage is also thought to be involved in disulfur and copper cytotoxicity. You can see here the inhibition of CHK1, then also immunofluorescence in the central slide. You can see how H2AX tends to increase and segregate in the nucleus. There's a lot more foci that are positive for this protein after treatment, as well as AIF1 that segregates into the nucleus, and NPO4 that's targeted and expressed at higher levels after treatment. With this slide, we also wanted to show that disulfur and copper is a safety combination, and you can see that there's no decrease in weight after three weeks of treatment at different concentrations, even subtherapeutic. And then the efficacy. So we used two cell lines, in this case D425 and D341 with intracranial xenografts, and we treated them with low concentrations of disulfur and copper for three weeks, given orally. You can see that both survivals are significant. We wanted to show that also the patterns and pathways of disulfur and copper action are increased, and you can see here that both AIF and NPO4 and phospho-2H2AX are increased after treatment. And with immunohistochemistry, we're also able to show that cas67 is decreased after treatment, as well as caspase-3 tends to increase after these two therapeutics are administered in vivo. NPO4 increases, even if not significantly, in these slides. But importantly, NUAN and GFAP do not change after treatment, meaning that most likely there's no toxicity on neurons and astrocytes. In summary, we're able to demonstrate in vitro the efficacy of these therapeutics in cell lines, and then to confirm these findings in vivo using two different models, and then confirming these findings with different techniques such as immunohistochemistry and Western blotting. There are some limitations to our models. First, again, using implanted xenografts is one, and then the need for additional therapeutic options to be tested both in vitro and in vivo in combination with disulfur and copper. So in the future, we'd like to test more cell lines, as well as lines derived directly from patients. And also, I think it's important to test these therapeutics with current chemotherapeutic and radiation treatment options. With this, I would like to thank Dr. Bram and Dr. Tyler, and all the members of the lab that helped me in this project. And thank you for your attention. I hope this message finds you well. My name is Rahul Kumar, and it's a privilege to present this work highlighting rare instances of subgroup divergence at medulloblastoma relapse. It is well appreciated that medulloblastoma is comprised of four clinically and molecularly distinct subgroups known as Wnt, Sonic Hedgehog, Group 3, and Group 4. Previous studies have suggested that subgroup is conserved absolutely between primary and relapse tumors. In this study, we leveraged DNA methylation arrays and robust bioinformatic classifiers to definitively annotate subgroup and other molecular features of patient-matched diagnostic and relapse tumors. Of 118 patients with relapsed medulloblastoma, we observed five cases in which the subgroup affiliation at relapse was discordant with that of the primary tumor. Notably, all of these cases occurred between Group 3 and Group 4 tumors. Given that clear discrimination of Group 3 and Group 4 tumors can sometimes be gray, we utilized probabilistic confidence scores generated from our subgroup classifier. As shown in the scatterplot, four of these primary tumors were classified as high-confidence Group 4 tumors that then diverged at relapse to high-confidence Group 3 tumors. The converse case involving a high-confidence primary Group 3 tumor, which then relapsed as an intermediate score in Group 4 tumor, suggests a borderline case in terms of confidence in relapse subgrouping call. When we looked at the molecular alterations in the four cases of Group 4 to Group 3 divergence, we identified an enrichment of relapse-specific alterations in the MYC-MYCN pathway, including deleterious alterations in FBXW7, which targets MYC for proteasomal degradation. Additionally, all four of these cases exhibited discordant Group 3-4 subtype classifications. An exemplary case is depicted here, where the high-confidence Group 4 primary tumor exhibited a low-level gain of MYC. This patient subsequently experienced extraactional relapse in the femur, with copy number profiling demonstrating a significantly higher amplitude of the MYC amplification at relapse. Such an instance not only highlights the aggressive nature of Group 3 tumors, but also raises interesting questions regarding the cellular origins of Group 3-4 tumors, as well as the role of MYC and other master transcription factors in determining subgroup affiliation. In conclusion, we describe, to our knowledge, the first cases of subgroup divergence in medulloblastoma. While a rare phenomenon that occurs from Group 4 primary tumors to Group 3 relapse tumors, we observe relapse-specific enrichment of MYC-MYCN pathway alterations in such cases. These results spark interesting questions related to the potential interplay between developmental origins and the methylation signatures underlying subgroup classification. Future studies are needed to recapitulate these findings and to explore mechanistic underpinnings. Nevertheless, these results challenge the notion of absolute subgroup stability amongst Group 3-4 medulloblastoma. Welcome back. I enjoyed those presentations. We will now have expert commentary by Dr. David Sandberg. He is a professor of neurosurgery and director of pediatric neurosurgery at Memorial Hermann Hospital at University of Texas. Thank you, Dr. Cheshire. It's a privilege to comment on these outstanding papers. And I'd first like to just congratulate the authors, both Dr. Kumar and Dr. Sarah, and their senior authors on really outstanding work. I learned something from all three of these papers. I'm going to comment briefly, and I'm going to comment on the first and third paper together first because they're by the same group of authors and with similar themes. As we learn more and more about the molecular subcategorization of medulloblastomas, we know that it affects prognosis, but increasingly it's going to affect treatments as there are specific treatment algorithms developed over the course of the coming years for these different subcategories. And I think intuitively, we would all think that if it's a Group 3 tumor at presentation, it's likely going to be a Group 3 tumor at recurrence. As it turns out, what these papers have taught us is that this is true most of the time, but it is not true all of the time. And I think what this is going to lead to from a standpoint of the neurosurgeons in the audience is I think it's going to lead to a reconsideration about, particularly as new treatments are developed, whether tissue is needed at the time of recurrence in order to make treatment decisions. You know, obviously, if it's a large mass, you're going to go back in and re-resect it to try to get down to minimal disease in the setting of a recurrence. But for a smaller mass, the question remains, should you go back? And, you know, I think for oncologic purposes, it may not be necessary from a surgical standpoint, but if there's a different treatment, if that tumor has changed from a sonic hedgehog to a Group 3 or Group 4 or something else, I think those are going to be the decisions that we're faced with with our oncology teams in the future. Outstanding papers. The second paper by Dr. Serra talks about combination therapy of two agents, disulfiram and copper, that were tested first in vitro in high-risk medulloblastoma cell lines, and then followed through in vivo with xenografts and looking at markers of DNA damage in NPL4 inhibition, and found that there was in vitro correlation with in vitro results. The in vivo results were that the tumor shrunk, and in vitro there was induction of apoptosis that corresponded with NPL4 aggregation in tumors. This is a great study. I think we would obviously need further safety tests in larger animals and, you know, you know, more detailed studies to optimize the dosing before proceeding with human pilot trials, but, you know, another promising avenue of research. So thank you and congratulations to these outstanding papers. All right, welcome back everybody. So we are unfortunately at the last section of abstracts for this tumor session. Such great, interesting topics. I hope you all went to the conversation lounge. Please go there as well later, and don't forget to pop on to Twitter and all the social media as well. There's a lot of action going on there right now. And so now we will hear from Dr. Mohamed Fouda and Dr. Michael McDowell and Trinka Vismaisi and Eric Prince about all of their work. And afterwards, we'll hear from Dr. Storm and his thoughts on these great, interesting abstracts. So go ahead. Good afternoon, everyone. My name is Mohamed Fouda. I'm currently a postdoc fellow at the Department of Neurosurgery at Johns Hopkins University. And today I'm going to present one of my projects that I've done before during my time at Boston Children's. In this study, we looked at the numbers of postoperative MR imaging studies that have been obtained for patients with pediatric craniopharyngioma at Boston Children's over the past 30 years. The aim of this study is to propose new standardized postoperative MRI surveillance protocols that yield efficient detection rates of recurrence as early as possible, using fewer imaging studies compared to the current practice. This is a retrospective core study in which we included all pediatric craniopharyngioma patients who have been diagnosed and treated at Boston Children's between 1990 and 2017. Preoperative CT brain, pre and postoperative MRI brain with and without contrast were available for all patients. MR angiography and MR ventricular check were excluded from the count. 80 patients met the inclusion criteria. The median age at time of initial diagnosis was about nine years old. And all patients were followed for at least one year with a median follow up period of 11 years. In order to determine the independent risk factors for tumor recurrence or progression, we ran the univariate and multivariable logistic regression analysis, and these turned out to be fine calcification and subtotal resection alone without any adjuvant radiotherapy. In this graph, we can see that more than 75 percent of the recurrent cases took place in the first three years postoperatively. In the Kaplan-Meier curves, we can see that the subtotal resection only without any adjuvant radiotherapy associated with the highest risk of recurrence. We can also see that the recurrent cases clustered in the first three years postoperatively. There was also no statistically significant difference between gross total resection and subtotal resection with adjuvant radiotherapy in term of tumor control. 1020 scans in total, the median was 12 scans per patient. Indications included intraoperative 2.5 percent, postoperative 5 percent, postoperative 5 percent, clinical deterioration 2.5 percent, routine follow-up 90 percent, and positive radiographic findings were only found in 8 percent of the scans. In this graph, we can see that the number of MRI scans that were done for the recurrent progressed cases was almost twice that of stable ones. Based on the previous Kaplan-Meier curves, in patients who underwent gross total resection, the recurrence took place within the first five postoperative years. In patients who underwent subtotal resection alone without any adjuvant radiotherapy, the recurrent cases clustered within the first two postoperative years. And in patients who underwent subtotal resection followed by adjuvant radiotherapy, the recurrence took place later between 8 and 10 years postoperatively. Therefore, we proposed three different postoperative MRI surveillance protocols based on the management strategies. 0, 9, 15, 36, 48, and 60 months for patients who underwent gross total resection. 0, 3, 6, 12, 18, and 24 months for patients who underwent subtotal resection alone. 0, 3, 12, 72, 96, and 120 months for patients who underwent subtotal resection followed by adjuvant radiotherapy. In this graph, we can see the substantial decrease in the number of MR scans necessary to safely follow up these patients and detect radiographic recurrence as early as possible. In conclusion, patients with pediatric craniopharyngioma may experience high rate of recurrence or progression even in the setting of extensive resection. The vast majority of the recurrent cases are asymptomatic. Our recommended postoperative MR imaging surveillance protocols would allow us to substantially decrease the number of MRI scans needed to safely detect radiographic recurrence as early as possible. These protocols would also provide a potential 50% decrement of healthcare costs, and they may also minify the psychological burden of frequent MRI scanning and visits for these patients and their families. This paper has been published in Child's Nervous System, and I hope you can find the time to read it, and please feel free to reach out to me if you have any questions or concerns. I'd like to thank Emily, Steve, Sarah Henderson, and the whole research team at Boston Children's, Dr. Scott, Dr. Marks, and finally Dr. Lisa Bird for her invaluable mentorship of this work. Thank you. Afternoon, my name is Michael McDowell. I'm a PGY-7 at University of Pittsburgh. Thank you for the opportunity to present our experience in early endoscopic endonasal surgery. Here are our disclosures. Endoscopic endonasal approach has gained enormous popularity as a result of its less disruptive and direct approach to address diverse pathologies affecting the skull base. It has begun to find favor in pediatric neurosurgery. However, early childhood cases remain sporadically reported. To illustrate, here's a case of a six-year-old girl presented with diplopia and claviclerdoma. Traditional concerns include potential anatomic restrictions such as nasal aperture size and disruption of normal growth plates. In addition, younger children tend to have higher rates of pre-cellular and conchal pneumatization patterns, as we see there, and also smaller intracranial distances. As you can see, the obstruction of traditional landmarks theoretically increases the risk of neurovascular injury. Nonetheless, with careful assessment of anatomy at appropriate intervals and generous use of the drill, safe access can be achieved. Large-core clinical validation, therefore, is needed to substantiate the safety and efficacy of early endonasal surgery. In this study, we report our experience. We hypothesize that endoscopic endonasal approaches could be performed effectively and safely in this age group. To investigate this, we performed a retrospective observational study on children less than the age of seven from 2002 to 2019 who underwent endoscopic endonasal surgery. Univariate statistics were performed as appropriate. A total of 36 patients qualified for the study. The mean age was 4.5 years old at the time of surgery, and the youngest child was about a year and a half old. 11 different pathologies were treated, with the two most common being encephalocele and craniopharyngioma. While one might assume that the majority of pathologies treated would be limited to the cellular and supercellular regions, anterior fossil locations were the most common, and we accessed all classical regions, including the coronal plane. Of the 21 patients, 11 had gross total resection based on radiographic comparison. 10 had near total resection, which was defined as 95 percent or greater resection of tumor. Here's an example of a chordoma with a small amount of inferolateral residual. Investigating our rates of gross total resection for the two most common tumor pathologies, chordoma and craniopharyngioma, we compared our data in this cohort to the past data from our institution regarding gross total resection. For chordoma, we found a very similar rate of resection compared to older children. Similarly, for craniopharyngioma, gross total resection rates were similar to older children, and both turned out to be better than adults, which may be related to the number of recurrent tumors in adults compared to children. Nine patients exhibited recurrence of pathology, about 25 percent total, of which six were craniopharyngiomas. Two patients died of disease progression, one being a chordoma and one being a rhabdomyosarcoma. The mean follow-up time was 64 months. Again, looking at specific tumor subsets, chordomas, there was a trend towards a higher recurrence rate in children, 33 percent versus 18 percent, but this was not statistically significant as the numbers were very small. For craniopharyngiomas, there was again a non-significant trend towards recurrence in younger children, 54 percent versus 41 percent, and 32 percent in adults, which was also trended towards but was not significant. Post-operative complications were fairly typical, the two most common being panhypopituitarism in craniopharyngioma patients and CSF leaks. There was an overall 8 percent CSF leak rate. Of note, there were no intraoperative vascular injuries. Looking at our CSF leak rate over time, we see that it strongly trends towards early cases in the cohort. As our experience and use of vascularized flaps and selective CSF drainage have increased over the last decade, we also see that this leak rate has fallen to around 4 percent, which is comparable to our other patient population. In conclusion, endoscopic endonasal surgery is possible for a wide array of pathologies in the sagittal and coronal planes. Resection rates are fairly comparable to older patients with similar rates of neurovascular complication, which are very low. However, recurrence rates may be higher, possibly due to a difference in tumor or patient biology at this age. Finally, standard reconstruction techniques, including nasal septal flaps, can be effectively implemented, but there was a steep learning curve. Thank you very much. I look forward to answering any questions at the appropriate time. Hello, my name is Srinkar Bajmasi. In this presentation, I would like to talk about our work on pediatric craniopharyngioma in Dr. Todd Hankinson's lab in the Department of Neurosurgery at the University of Colorado Denver and the Children's Hospital Colorado. Adamantinomatous craniopharyngioma, or ACP for short, is a heterogeneous tumor with solid and cystic components. The solid component consists of distinct histological features, such as the palisading epithelium and the epithelial walls at the tumor brain interface. The cystic spaces are filled with the characteristic machine oil-like cyst fluid. The only known genetic aberration is a point mutation in beta-catenin gene that makes beta-catenin resistant to proteolytic degradation. As a result, beta-catenin accumulates in the cytoplasm and the nucleus. The main objectives of this project were to study the role of cyst fluid in disease pathology and specifically to study the effect of cyst fluid on normal human epithelial cells. We conducted the experiments by culturing normal human epithelial cells with or without ACP cyst fluid at 10% concentration for different durations. We observed the morphological changes of the cells We observed the morphological changes of the cells by real-time live cell imaging and investigated cytoskeletal remodeling by labeling actin filaments with phalloidin. We determined the concentration of secreted cytokines using a multiplexed magnetic bead-based cytokine assay. We determined the relative gene expression of selected genes related to the epithelial to mesenchymal transition process using RT-qPCR. When normal epithelial cells were treated with ACP cyst fluid, they undergo profound morphological changes. The top panel shows epithelial cells grown in controlled conditions. The bottom panel shows epithelial cells cultured with ACP cyst fluid. Cyst fluid treated cells progressively became large, irregular, elongated, and spindly. This led us into thinking whether these cells were undergoing epithelial to mesenchymal transition. We investigated this further by labeling actin filaments using FITC-conjugated phalloidin. In cells grown under controlled conditions, actin filaments were uniformly observed to have a ring-like structure around the cytoplasm. In cyst fluid treated cells, this cytoskeletal organization was disrupted and actin filament remodeling appeared to have occurred. The transformation from regular ovoid-shaped epithelial cells to spindly mesenchymal-like cells can be appreciated here. Multiplexed cytokine assay of cultured supernatants revealed that normal epithelial cells responded to cyst fluid treatment by secreting copious amounts of pro-inflammatory and immunomodulatory cytokines. Some key cytokines such as IL-6 and IL-8 were upregulated by up to 20 to 54 times. The EMT-regulating transcription factors ZIP2 and SNAIL1 were upregulated following cyst fluid treatment. Bimentin is a prototypical marker of the mesenchymal cell lineage and is considered a late-stage marker of the EMT process. In cyst fluid treated cells, bimentin upregulation was observed only after eight days of treatment but not after shorter durations of treatment. In conclusion, we say that ACP cyst fluid induces profound morphological changes in normal epithelial cells and this change resembles the EMT process. ACP cyst fluid elicits a strong inflammatory response in normal epithelial cells. The EMT-regulating transcription factors ZIP2 and SNAIL1 and the mesenchymal marker bimentin are upregulated in epithelial cells following cyst fluid treatment. Our hope is that appreciating the role of cyst fluid in the context of EMT-like processes will help us to understand the pathology of ACP and to develop mechanistically targeted drugs. I'd like to extend my special thanks to all members of the Foreman Lab, the Wibrocker Lab, and the Hankinson Lab at the University of Colorado for their invaluable help with my work. Finally, I'd like to thank my mentor, Dr. Todd Hankinson, for his constant support and guidance and for the opportunity to work on this topic in his lab. Thank you. Hello, my name is Eric Prince. I'm from the University of Colorado and I would like to thank you for the opportunity to speak to you today about our work at clinical radiographic metrics and how we can use them to subgroup pediatric adamantinomenous craniopharyngeal patients. ACP is a rare benign epithelial tumor arising in the supercellar region of the brain. It is highly heterogeneous and presents with epithelial and reactive glial It is highly heterogeneous and presents with epithelial and reactive glial phenotypic signatures as well as calcification and cystic components as visualized by the radiographic images below. It presents for the bimodal age distribution with peak incidence in children happening around nine years old and peak incidence in adult happening around late 60s. Our data set is comprised of 50 patients that were obtained through the Advancing Treatment for Pediatric Craniopharyngioma Consortium over about a three-year period where our mean age reflects that of what we would expect for a pediatric population and we have about an even distribution of male and female patients. Our measurements that we're considering are all radiographic and there are four location measurements, 12 involvement invasion measurements, seven tumor characteristics, four cyst characteristics, and five ventral characteristics. All of these measurements were obtained by board-certified neuroradiologists. In order to make sense of all of these variables we had to reduce those to three dimensions which we did using multiple factor analysis which is similar to principal components analysis which you might be familiar with. The variance explained across these three dimensions is 67 percent which indicates that our model is well fit and we're able to represent the data in a reliable way. If you look we are able to see that each dimension has a unique contributing characteristic to it. For example dimension one is most contributed to by the size of the tumor, dimension two is most contributed to by the ventricle size, dimension three is most contributed to by the cyst size. Using these three dimensions and the Bayesian information criterion we are able to cluster and identify that there are three optimal clusters within this 50 patients. By using matched transcriptome data we're able to use differential expression analysis for each cluster to determine unique expression signatures for each one. For example in cluster one we see the canonical oncogene pathways MYC and p53 were enriched using gene set enrichment analysis. In cluster two we see sonic hedgehog signaling is enriched and in cluster three we see TGF beta signaling and epithelial mesenchymal transition signaling are enriched. Each one of these subgroups are uniquely interesting in the context of ACP. Thank you. I want to thank the presenters and the organizers for this cluster of presentations. I tell my patients that craniopharyngiomas are benign tumors and malignant locations with significant morbidities regardless of management. These are complex, rare, heterogeneous, and controversial entities. Do you go for a gross total resection or a subtotal resection and radiate? Do you approach through a craniotomy or endoscopic endonasal surgery? What is the best post-operative scanning regimen? How can we find better or less morbid management strategies? Today's presenters address some of these, but it's important to remember that with small numbers it's hard to draw meaningful conclusions. That's why if we're going to make advances, collaborative research is imperative, like the one that Todd Hankinson is leading called the Advancing Treatment for Pietra Craniopharyngiomas and another one is the Children's Brain Tumor Network. I think it's only through these kind of collaborative efforts where we make meaningful advances and accelerate treatments. At this point, I'll turn it over to our moderators and our presenters and thank you for the presentations. Maybe you can just hang on in here before you move on to the conversation lounge. We have one question in the Q&A directed to the speaker who shared their experience with EEA in children. The question is how do you reconstruct your celluloflora, mostly nasoceptal flap versus fascia lata? I'd be curious of hearing what do you guys do at CHOP? So at CHOP we use the nasoceptal flap. We also use fat and fascia. We do not use lumbar drains and so I'd say our leak rate right now is about three to four percent. We stopped using lumbar drains years ago but we have not stopped. We've talked about that we do lumbar drain and ovulate using fascia and fat fascia lata but we've decided we've got a good thing going so we're just going to no drain fat and fascia lata and fat from the thigh. Yeah, I agree. We generally will even save the nasoceptal flap only for high flow leaks. Small leaks we'll just do a mucosal flap or you know reconstruction with durogen and things like that and rarely use a lumbar drain. Rarely use a fascia lata as well. Sam, what do you guys do as well? We're pretty much the same. Rarely use the lumbar drain in these situations. Perfect. Well, all right. Well, guys, I'm going to save it for the last time. Please go to the conversation lounge. Dr. Surm is going to try to go over there. You can ask him questions there. Hopefully the other speakers are there as well. I'm going to try to pop over there as well and hopefully, Sam, you join us too. And this wraps up the tumor session. Such great talks on a variety of different histopathologies and from the bench work to clinical work. Sam, anything else you want to say to the crowd? No, I very much enjoyed all the presentations and I'd like to give a special thanks to all our speakers who are here live as well as who are here by video only. So we appreciate your time and we look forward to seeing you in Salt Lake City for real next year. All right. And I hope you guys join the next session which will be our fellows work and showcase which is an important part for them and their future careers that we can all support. So see you all guys there later.

pediatric brain tumors

diagnosis

treatment

vision loss

pathoedema

multidisciplinary care

molecular comparisons

pediatric medulloblastoma

surgical experience

posterior fossa syndrome

CSF diversion

combination therapy

group three medulloblastoma

cyst fluid

pediatric craniopharyngioma