Search

Composite pleomorphic xanthoastrocytoma-ganglioglioma (PXA-GG) is an extremely rare central nervous system neoplasm with 2 distinct but intermingled components. Whether this tumor represents a “collision tumor” of separate neoplasms or a monoclonal neoplasm with divergent evolution is poorly understood. Clinicopathologic studies and capture-based next generation sequencing were performed on extracted DNA from all available PXA-GG at 2 medical centers. Five PXA-GG were diagnosed in 1 male and 4 female patients ranging from 13 to 25 years in age. Four arose within the cerebral hemispheres; 1 presented in the cerebellar vermis. DNA was sufficient for analysis in 4 PXA components and 3 GG components. Four paired PXA and GG components harbored BRAF p.V600E hotspot mutations. The 4 sequenced PXA components demonstrated CDKN2A homozygous deletion by sequencing with loss of p16 (protein product of CDKN2A) expression by immunohistochemistry, which was intact in all assessed GG components. The PXA components also demonstrated more frequent copy number alterations relative to paired GG components. In one PXA-GG, shared chromosomal copy number alterations were identified in both components. Our findings support divergent evolution of the PXA and GG components from a common BRAF p.V600E-mutant precursor lesion, with additional acquisition of CDKN2A homozygous deletion in the PXA component as is typically seen in conventional PXA.

BACKGROUND Alterations in NF2 underlie meningioma tumorigenesis, but tumor suppressor functions of the NF2 gene product, Merlin, are incompletely understood in meningiomas. Here we integrate proteomic proximity-labelling mass spectrometry with CRISPR interference (CRISPRi), RNA sequencing, and biochemical approaches to discover Merlin drives meningioma apoptosis and susceptibility to cytotoxic therapy. METHODS RNA sequencing was performed on triplicate M10G meningioma cells stably expressing CRISPRi machinery and either non-targeting control sgRNAs, sgRNAs suppressing NF2, or sgRNAs suppressing NF2 with Merlin rescue. QPCR in IOMM-Lee and MSC1 meningioma cells expressing non-targeting control shRNAs or shRNAs suppressing NF2 was used for orthogonal validation in vitro. RNA sequencing of euploid meningiomas (n=52) or meningiomas with loss of NF2 as the only copy number variant (n=28) was used for orthogonal validation in vivo. Merlin interactors in meningioma cells were identified using APEX proteomic proximity-labelling mass spectrometry. Mechanistic and functional studies were performed using biochemical, molecular, and cell biology approaches in meningioma cells and CH-157MN meningioma xenografts treated with cytotoxic chemotherapy or ionizing radiation. RESULTS Merlin suppression in meningioma cells and xenografts inhibited pro-apoptotic interferon regulatory factor (IRF) target genes and attenuated meningioma apoptosis. Merlin suppression did not alter IRF stability or subcellular localization in meningioma cells, and proteomic proximity-labelling mass spectrometry revealed a novel interaction between wildtype Merlin and ARHGAP35, a DNA binding factor that inhibits glucocorticoid receptor expression (NR3C1). NR3C1 inhibits IRF activity to prevent apoptosis, and Merlin suppression in meningioma cells induced NR3C1expression, which was inhibited by Merlin rescue. Further, NR3C1 suppression rescued meningioma cell apoptosis in the absence of Merlin, and NR3C1 expression was increased in human meningiomas with loss of NF2 compared to euploid meningiomas. CONCLUSIONS These data shed light on a novel pro-apoptotic tumor suppressor function of Merlin regulating glucocorticoid signalling and susceptibility to cytotoxic therapy in meningiomas.

BACKGROUND Meningioma treatments are limited due to incomplete understanding of meningioma biology. To address this, we performed multiplatform molecular profiling on 565 meningiomas with comprehensive clinical data to define genomic drivers and identify therapeutic vulnerabilities. METHODS DNA methylation profiling was performed on meningiomas from UCSF (n=200, discovery) and Hong Kong University (n=365, validation). Median follow-up was 5.6 years, and there were 388/142/35 WHO grade I/II/III meningiomas. Copy number variants (CNVs) were calculated for all meningiomas, and RNA sequencing was performed on UCSF meningiomas. Cell type deconvolution, metagenomics, CRISPR, and pharmacology were used for mechanistic and functional validation. RESULTS Unsupervised hierarchical clustering of differentially methylated DNA probes revealed that meningiomas were comprised of 3 epigenetic subgroups associated with good, intermediate, and poor outcomes, with representation from all WHO grades in each subgroup. Meningiomas from the subgroup with the best outcomes (52% WHO grade I) were distinguished by recurrent gain of Chr5. Meningiomas from the subgroup with intermediate outcomes (31% WHO grade II) were distinguished by genomic stability, enrichment of innate immune genes, and immune infiltration in the setting of endogenous retroviral gene re-expression, a mechanism of immune recruitment. The most aggressive subgroup of meningiomas (57% WHO grade III) was distinguished by genomic instability, including recurrent loss of Chr22q harboring NF2, and decreased immune infiltration. Consistently, NF2 suppression in primary meningioma cells derived from immunogenic meningiomas decreased expression of innate immune genes critical for immune recruitment, suggesting a novel immunostimulatory function of NF2. The most aggressive subgroup of meningiomas were further distinguished by activation of the mitogenic FOXM1 transcriptional program, and recurrent loss of Chr9p harboring CDKN2A/B, which rendered primary meningioma cells from this subgroup susceptible to CDK4/6 inhibitors. CONCLUSIONS Meningiomas are comprised of 3 epigenetic subgroups defined by genetic mechanisms driving immune infiltration in the tumor microenvironment and meningioma cell proliferation.

BACKGROUND Alterations in the NF2 tumor suppressor gene lead to meningiomas and schwannomas, but the tumor suppressor functions of the NF2 gene product, Merlin, are incompletely understood. To address this problem, we performed a structure-function analysis of Merlin by expressing cancer-associated missense single-nucleotide variants (mSNVs) in primary cancer cells for biochemical and cell biology experiments. METHODS All NF2 mSNVs were assembled from cBioPortal and COSMIC, and modelled on the FERM, a-helical, and C-terminal domains of Merlin (PDB 4ZRJ) using comparative structure prediction on the Robetta server and visually inspected using Pymol. mSNV hotspots were defined from sliding windows with at least 10 mutations within 5 residues in either direction. mSNVs from hotspots in meningiomas, schwannomas, or both, were selected for in vitro mechanistic analyses using immunofluorescence and immunoblotting of whole cell, plasma membrane, cytoskeletal, cytoplasmic, nuclear, and chromatin subcellular fractions from M10G meningioma cells and HEI-193 schwannoma cells. RESULTS We identified the following cancer-associated hotspot mSNVs in NF2, which were over-expressed for mechanistic studies: L46R, S156N, W191R, A211D, V219M, R418C and R462K. Endogenous Merlin was detected in all subcellular compartments, but was enriched in the nucleus. L46R and A211D mapped to hydrophobic pockets in the FERM domain, destabilized Merlin, and excluded Merlin from all subcellular compartments except the cytoskeleton. S156N, W191R and V219M also mapped to the FERM domain, but did not affect Merlin stability, and V219M attenuated chromatin localization, suggesting this motif may be involved in binding events that regulate subcellular localization. R418C and R463K mapped to the a-helical domain, but only R418C destabilized Merlin. CONCLUSION Our results suggest that cancer-associated mSNVs inactive the tumor suppressor functions of NF2 by altering the stability, subcellular localization, or binding partners of Merlin. Further work is required to identify and understand the impact of binding partners and subcellular localization on Merlin function.

Background Malignant peripheral nerve sheath tumor (MPNST) is an aggressive and poorly understood malignant neoplasm. Even in the setting of multimodal therapy, the clinical course of MPNST is frequently marked by metastatic conversion and poor overall prognosis, with optimal treatment paradigms for this rare tumor unknown. Methods We reviewed the medical records and histopathology of 54 consecutive patients who were treated at University of California San Francisco between 1990 and 2018. Results Our cohort consisted of 24 male and 30 female patients (median age 38 years). Fédération Nationale des Centres de Lutte Contre Le Cancer (FNCLCC) sarcoma grading criteria segregated patients into groups with differences in overall survival (OS) (P = .02). Increasing Ki-67 labeling index was associated with poor OS (hazard ratio [HR] 1.36 per 10%, P = .0002). Unsupervised hierarchical clustering-based immunohistochemical staining patterns identified 2 subgroups of tumors with differences in H3K27me3, Neurofibromin, S100, SOX10, p16, and EGFR immunoreactivity. In our cohort, cluster status was associated with improved locoregional failure-free rate (P = .004) in response to radiation. Conclusions Our results lend support to the FNCLCC sarcoma grading criteria as a prognostic scheme for MPNST, although few cases of grade 1 were included. Further, we identify increased Ki-67 labeling as a strong predictor of poor OS from MPNST. Finally, we identify a subset of MPNSTs with a predictive immunohistochemical profile that has improved local control with adjuvant radiotherapy. These data provide insights into the grading and therapy for patients with MPNST, although further studies are needed for independent validation.

BACKGROUND Factors associated with meningioma recurrence after postoperative radiotherapy are poorly understood, and the optimal postoperative radiotherapy target delineation for meningioma is unknown. The objective of this study was to identify factors influencing meningioma recurrence after postoperative radiotherapy to inform patient selection and treatment design. METHODS Medical records were retrospectively reviewed for patients who underwent meningioma resection at a single institution between 1991 and 2015. Patients with sufficient tumor tissue for histologic classification and who received postoperative radiation therapy with external beam radiotherapy (EBRT), stereotactic radiosurgery (SRS) or brachytherapy, were included. Local freedom from recurrence (LFFR) was analyzed according to tumor and treatment characteristics using the Kaplan Meier method. RESULTS We identified 86 patients with 96 meningiomas who met inclusion criteria. Nineteen meningiomas (20%) were WHO grade I, 56 (58%) were grade II and 21 (22%) were grade III. Forty-one meningiomas (43%) were recurrent, and 55 (57%) were de novo. The postoperative radiotherapy modality was EBRT for 58 patients (60%), SRS for 20 (21%) patients and brachytherapy for 18 (19%) patients. With a median follow up of 4.3 years (IQR 2.1–8.8 years), there were 48 (50%) local failures that occurred a median of 17 months after immediate prior resection (IQR 9–33 months). WHO grade II/III and recurrent meningiomas had worse LFFR (p< 0.001). The 5-year LFFR was 53% after EBRT (95% CI 41–69%), 53% after SRS (95% CI 34–84%) and 15% after brachytherapy (95% CI 3–74%), although meningiomas that were treated with brachytherapy were significantly more likely to have received prior EBRT or SRS (86% versus 29%, p< 0.001). CONCLUSIONS These data provide a foundation for understanding patterns of meningioma recurrence after postoperative radiotherapy. Ongoing analyses aim to quantify the relationships between postoperative radiotherapy dose, target delineation and local control of meningioma.

INTRODUCTION Little is known about the mechanisms by which gliomas integrate into functional neural networks and influence complex cognitive processes such as language. Glioma-neuron interactions are bidirectional, with increased neuronal activity promoting tumor growth and the latter in turn influencing neuronal excitability and synaptic connections. It remains unknown whether glioma-neuron interactions play a role in maintaining long-range neural networks subserving cognition in humans. We test the hypothesis that glioma-neuron interactions (“synaptogenic glioma cells”) are enriched within intratumoral high functional connectivity (FC) network hubs, thereby influencing language processing via release of synaptogenic factors into the tumor microenvironment. METHODS We employed magnetoencephalography imaginary coherence measures to identify intratumoral high (HFC) and low (LFC) functional connectivity network hubs in newly diagnosed glioblastoma patients. Primary patient samples and cultures from HFC and LFC sites were assessed for pre and post-synaptic marker expression (IF), cocultured with murine hippocampal neurons, and induced neuron organoids. ECOG Field recordings were performed on HFC/LFC tumors. Secreted proteins were measured from patient serum and LFC/HFC culture supernatant. Language assessments were performed to correlate task performance with FC measures. RESULTS Primary patient samples from HFC regions are enriched for glioblastoma cells with a synaptogenic profile as characterized by pre- and post-synaptic marker expression at both tissue and cellular level (coculture with mouse hippocampal neuron and organoid models). RNA sequencing and proteomic analyses from HFC samples revealed a neurogenic signature including thrombospondin 1 (TSP1). Overexpression of TSP1 in LFC primary patient cultures rescues the synaptogenic and proliferative phenotype. Importantly, we found a linear relationship between intratumoral HFC with patient serum TSP1 (ELISA) with a further correlation with language task performance. CONCLUSION An enriched population of synaptogenic glioma cells are organized within intratumoral high network connectivity regions. Glioma-induced neuronal synaptogenesis contributes to the microenvironment in support of network connectivity through secretion of TSP1.

Long noncoding RNAs (lncRNAs) exhibit highly cell type-specific expression and function, making this class of transcript attractive for targeted cancer therapy. However, the vast majority of lncRNAs have not been tested as potential therapeutic targets, particularly in the context of currently used cancer therapy. Malignant glioma is rapidly fatal, and ionizing radiation is part of the current standard-of-care used to slow tumor growth in both adult and pediatric patients. Here, we used a CRISPR interference (CRISPRi) platform to systematically screen 5689 lncRNA loci in human glioblastoma (GBM) cells, identifying 467 hits that modified cell growth in the presence of clinically relevant doses of fractionated radiation. 34 of these lncRNA hits sensitized cells to radiation, and based on their expression in adult and pediatric glioma, nine of these hits were prioritized as lncRNA Glioma Radiation Sensitizers (lncGRS). Knockdown of lncGRS-1 – a primate-conserved, nuclear-enriched lncRNA – inhibited the growth and proliferation of primary adult and pediatric glioma cells, but not the viability of normal brain cells. Using human brain organoids comprised of mature neural cell types as a 3-dimensional tissue substrate to model the invasive growth of glioma, we found that antisense oligonucleotides (ASOs) targeting lncGRS-1 selectively decreased tumor growth synergistically with radiation therapy. These studies identify lncGRS-1 as a glioma-specific therapeutic target and establish a generalizable approach to rapidly identify novel therapeutic targets in the vast non-coding genome that can enhance the standard-of-care for cancer therapy.

OBJECTIVES: Ependymomas are associated with distinct anatomic, genomic and clinical characteristics, but the diversity of molecular features within individual ependymomas is unknown. Here, we perform multiplatform molecular profiling of 6 spatially- and radiographically-distinct regions of an ependymoma with C11orf95-RELA fusion to elucidate intratumor heterogeneity. METHODS: Immunohistochemistry, whole exome sequencing, 850k DNA-methylation profiling, RNA-seq, and quantitative MRI were performed to define the genomic and radiographic characteristics of stereotactically-collected regions of an ependymoma. Molecular phylogenetic analysis was performed using methylation- and DNA-based variants, and functional assays were conducted in cell culture. RESULTS: All ependymoma regions were diffusely positive for L1CAM and contained the pathogenic C11orf95-RELA fusion, but not YAP1 fusion, along with chromosome 11 chromothripsis. Principal component analysis of transcriptomic data identified three molecularly distinct regions corresponding to stem cell/proliferation (OLIG2), neuronal differentiation (LBX1, HDAC9), and immune/stress response (CXCL2, HSPA1A), each of which reflected transcriptional programs of distinct cell types in neural development. Furthermore, transcriptional heterogeneity correlated with differences in blood perfusion quantified by MRI. Phylogenetic analysis of methylation-derived CNVs and whole exome-derived SNVs revealed early chromosomal structural alterations followed by point mutations that distinguish various regions. A novel mutation in SETD2 was discovered in all regions, and overexpression of this novel mutant (SETD2(K2R)) led to increased cell proliferation and viability, suggesting that epigenetic misregulation could be a driver of ependymoma heterogeneity. CONCLUSION: Anaplastic ependymoma is associated with a previously unappreciated molecular heterogeneity that may influence tumorigenesis and design of molecular therapeutics.

OBJECTIVES: Quantitative radiologic and radiomic features can identify brain tumors at risk for poor outcomes. Here, we investigate prognostic models for meningioma grade, local failure (LF) and overall survival (OS) based on demographic, radiologic, radiomic and therapeutic data. METHODS: We developed a database that was enriched for high grade meningiomas from 219 patients who underwent surgery for 229 meningiomas from 1990 to 2015 who had comprehensive clinical, pathologic and radiologic information available for retrospective review. The median imaging follow up was 4.3 years, and there were 112 WHO grade I (49%), 93 grade II (41%) and 24 grade III (10%) meningiomas. Two neuro-radiologists independently annotated 17 radiologic features, and 154 radiomic features were extracted from preoperative post-contrast 3D SPGR MR images for each meningioma. Random forest models were trained using nested resampling, and the performance of each model was assessed by calculating feature importance, mean balanced accuracy (BA) and area under the curve (AUC). RESULTS: Models restricted to preoperative demographic information and quantitative imaging features had superior BA (0.60–0.67) and AUC (0.60–0.76) for LF or OS as compared to models based on meningioma grade and extent of resection (BA 0.65, AUC 0.64). Integrated models incorporating all available data provided the most accurate estimates of LF and OS (BA 0.67, AUC 0.76). Radiomic features alone or in combination with other variables showed moderate and marginal predictive value for grade (BA 0.63, AUC 0.72) and LF (BA 0.61, AUC 0.65), respectively. Among radiologic features, meningioma diffusion characteristics significantly strengthened prognostication of grade and LF (RR 25.6, P = 0.001). Recursive partitioning analysis identified tumor size, primary versus recurrent presentation, grade, sphericity, apparent diffusion coefficient, location, extent of resection and T2 signal as the most important features for LF. CONCLUSIONS: Models using clinical and quantitative imaging data can accurately predict meningioma outcomes.

OBJECTIVES: Pediatric embryonal brain tumors include medulloblastoma, supratentorial primitive neuroectodermal tumor, and pineoblastoma. The purpose of this study was to investigate the prognostic utility of quantitative three-dimensional (3D) magnetic resonance imaging (MRI) radiomic analysis for primary pediatric embryonal brain tumors. METHODS: Thirty-four pediatric embryonal brain tumor patients with concurrent pre-operative T1-weighted post contrast (T1PG) and T2-weighted fluid-attenuated inversion recovery (FLAIR) MR images were identified from an institutional database. The median imaging follow-up was 5.2 years. Radiomic features were extracted from axial T1PG and FLAIR contours using MATLAB, and 15 features were selected for analysis based on qualitative radiographic features with known prognostic significance for pediatric embryonal brain tumors. Logistic regression, linear regression, receiver operating characteristic curve, Harrells C index and Somers D index were used to test the relationships between radiomic features, demographic variables and clinical outcomes. RESULTS: We found that pediatric embryonal brain tumors in older patients had increased normalized mean tumor intensity (P=0.05, T1PG), decreased tumor volume (P=0.02, T1PG) and increased markers of heterogeneity (P0.01, T1PG and FLAIR) relative to younger patients. We identified 10 quantitative radiomic features that delineated between medulloblastoma, pineoblastoma and supratentorial primitive neuroectodermal tumor, including size and heterogeneity (P0.05, T1PG and FLAIR). Decreased markers of tumor heterogeneity were predictive of neuraxis metastases and trended towards significance (P=0.1, FLAIR). Tumors with increased size (AUC=0.7, FLAIR) and decreased heterogeneity (AUC=0.7, FLAIR) at diagnosis were more likely to recur. CONCLUSIONS: Quantitative radiomic features are associated with pediatric embryonal brain tumor patient age, histology, neuraxis metastases and recurrence, and could be used for risk stratification.

Long non-coding RNAs (lncRNAs) are transcripts >200 nucleotides long with essentially no protein coding potential. While certain lncRNAs play key roles in cancer and therefore represent an important class of therapeutic targets, very few lncRNAs have been studied in glioma. Here, we show that knockdown of a primate-specific lncRNA – referred to as GTT1 – selectively inhibits the growth of both adult and pediatric glioma but does not adversely affect the viability of normal human glia and neurons. In a recent genome-scale CRISPR interference (CRISPRi)-based screen, we identified GTT1 as one of 65 lncRNA genes that modify the growth of U87 glioblastoma cells. To prioritize lncRNAs for further study, we next performed a CRISPRi-based screen to identify lncRNA targets that also sensitize tumor cells to radiation. Both CRISPRi-mediated and antisense oligonucleotide (ASO)-mediated knockdown of GTT1 inhibited the propagation of U87 cells in culture, and this effect synergized with radiation treatment. Furthermore, GTT1 knockdown inhibited the growth of patient-derived glioma cells including adult glioblastoma and pediatric diffuse intrinsic pontine glioma (DIPG) in culture. Although GTT1 is expressed in normal human brain, knockdown of this lncRNA in normal human astrocytes, cortical neurons and fetal forebrain tissues did not reduce cell viability. As GTT1 is primate-specific, knockdown of this lncRNA in mouse xenograft models would not fully assess potential adverse effects. We therefore developed a human brain organoid model comprised of mature astrocytes and functional neurons. Patient-derived glioma cells engrafted into these organoids, and tumors grew in an infiltrative manner. GTT1 knockdown in this brain organoid model strongly impaired tumor growth but did not reduce organoid viability. These studies identify GTT1 as a glioma-specific therapeutic target and illustrate how this human brain organoid tumor model can be used to rapidly evaluate the tumor-specificity of novel therapeutics.

OBJECTIVES: Chordoid meningiomas are a rare histologic variant that follow an aggressive clinical course. Here, we examine histopathologic and radiologic features of chordoid meningiomas to identify risk factors for recurrence. METHODS: Retrospective chart reviews were performed on 11 patients with chordoid meningioma and 15 patients with meningioma with focal chordoid features (

Meningeal solitary fibrous tumor (SFT)/hemangiopericytoma (HPC) is a rare tumor with propensity for recurrence and metastasis. We compared the WHO 2016 CNS tumor classification (CNS-G), a 3-tier system based on histopathologic phenotype and mitotic count, to the WHO 2013 Soft Tissue classification (ST-G), a 2-tier system based on mitotic count alone, in a cohort of 133 patients (59 female, 74 male; mean age 54 years [range 20–87]) with meningeal SFT/HPC. Tumors were pathologically confirmed through review of the first tumor resection (n=97), local recurrence (n=35) or distant metastasis (n=1). STAT6 immunostain showed nuclear expression in 132 cases. NAB2-STAT6 fusion was detected in 99 (of 111) successfully tested tumors (89%) including the lone STAT6 immunonegative tumor. Tumors were classified as grade 1 (n=43), 2(n=41) or 3 (n=49) using the CNS-G, and SFT (n=84) or malignant SFT (n=49) by using the ST-G. Necrosis was present in 16 cases (12%). On followup, 42 patients had at least 1 subsequent recurrence, including 8 with metastases. 29 patients died. On univariate analysis necrosis (p=0.0018) and CNS-G or ST-G (p-value respectively 0.014, 0.0041) were significantly associated with recurrence-free (RFS) but not overall survival (OS). NAB2-STAT6 fusion type was not associated with RFS or OS. Ten-year RFS was 61, 58 and 34% for CNS-G 1,2 and 3 versus 59 and 34% for ST-G SFT and malignant SFT, respectively. Ten-year RFS was 0% and 56% for tumors with and without necrosis. Our data suggest that SFT/HPC are better stratified using a two-tiered grading scheme. On multivariate analysis, necrosis was an independent predictor of RFS (HR 2.9, p=0.016) while ST-G was not quite significant (HR=1.9, p=0.062), suggesting necrosis should be reintroduced among SFT/HPC grading criteria.

OBJECTIVE: Extracranial meningioma metastases are rare, and patients with meningiomas are not routinely screened for systemic metastases. We report our experience with meningioma metastases and screening for metastases in select patients with recurrent meningiomas. METHODS: All patients undergoing surgical resection or stereotactic radiosurgery for primary or recurrent meningioma from 2009 to 2017 at a single center were retrospectively reviewed. Indications for metastasis screening were patients with recurrent meningioma after 2 prior resections and radiation therapy or symptoms concerning for metastasis. Screening was performed with CT chest/abdomen/pelvis, FDG PET-CT, or somatostatin receptor specific (DOTATATE) PET-CT. RESULTS: Of 1203 patients treated, 929 (77.2%) had WHO grade 1 meningioma by surgical pathology or imaging, 234 (19.5%) had grade 2 and 35 (2.9%) had grade 3 meningioma. A total of 298 meningiomas (24.8%) recurred, with a mean of 1.6 recurrences per patient. Screening for metastases was performed in 28 patients; one (3.6%) had a grade 1 tumor, 16 (57.1%) were grade 2, and 11 (39.3%) were grade 3. Five patients (17.9%) were screened because of systemic symptoms. Of the patients screened, 27 (96.4%) had recurrent meningioma, with a mean 3.2 recurrences. Ten patients (35.7%) had suspicious extracranial lesions by imaging. On biopsy, 8 were meningioma metastases, 1 was a non-meningioma malignancy, and 1 was lost to follow-up. Biopsy-confirmed metastases occurred in: liver (5), lung (3), mediastinum (1) and bone (1). The overall incidence of metastases was 0.67% (n=8). Incidence increased to 2% of WHO grade II tumors and 8.6% of grade 3. Using our indications, the number needed to screen to identify one patient with biopsy-confirmed malignancy was 3.1. CONCLUSIONS: Screening of patients with multiply recurrent meningioma or symptoms concerning for metastasis may identify extracranial metastases in a significant proportion of patients and can inform decision making for additional treatments.

BACKGROUND Medulloblastoma is an aggressive pediatric brain tumor that is associated with misactivation of the Hedgehog (HH) pathway. Our lab has shown that CDK6, a critical activator of the cell cycle, is a direct transcriptional target of oncogenic HH signaling, and that inhibiting CDK6 blocks the growth of HH-associated medulloblastoma in mice. A clinical trial exploring the efficacy of CDK6 inhibition in medulloblastoma patients is underway, but prior attempts to target the HH pathway in medulloblastoma have been encumbered by resistance to molecular monotherapy. Thus, we sought to identify mechanisms of resistance to CDK6 inhibition in HH-associated medulloblastoma. METHODS We performed orthogonal CRISPR and CRISPR interference screens in HH-associated medulloblastoma cells treated with pharmacologic inhibitors of CDK6 in vitro, and RNA-sequencing of HH-associated medulloblastomas with genetic deletion of CDK6 in vivo. Mechanistic and functional validation of resistance pathways was performed using CRISPR interference, immunoblotting, immunofluorescence, genetics, and pharmacology. Lipid quantification was carried out by ultra-high performance liquid chromatography-tandem mass spectrometry. RESULTS Our results reveal that decreased ribosomal protein expression underlies resistance to CDK6 inhibition in HH-associated medulloblastoma, leading to endoplasmic reticular (ER) stress and activation of the unfolded protein response (UPR). We show that ER stress and the UPR increase the activity of enzymes producing Smoothened-activating sterol lipids that sustain oncogenic HH signaling in medulloblastoma despite CDK6 inhibition. These discoveries suggest that combination molecular therapy against CDK6 and HSD11ß2, an enzyme producing Smoothened-activating lipids, may be an effective treatment for HH-associated medulloblastoma. In support of this hypothesis, we demonstrate that concurrent genetic deletion or pharmacological inhibition of CDK6 and HSD11ß2 additively blocks the growth of multiple models of HH-associated medulloblastoma in mice. CONCLUSIONS Smoothened-activating lipid biosynthesis underlies resistance to CDK6 inhibition in HH-associated medulloblastoma, revealing a novel combination therapy to treat the most common malignant brain tumor in children.

BACKGROUND: Recent evidence supports hippocampal sparing during whole brain radiotherapy (HS-WBRT) to improve neurocognitive outcomes in patients with brain metastases (BM). This study sought to quantify the hippocampal dosimetry and treatment efficacy of stereotactic radiosurgery (SRS) to 10 or greater BM to clarify the roles of SRS and WBRT. METHODS: Patients at a single institution treated with SRS to 10 or more BM without WBRT from 1999 to 2016 were retrospectively reviewed. Treatment-related outcomes including overall survival (OS), freedom from progression (FFP), freedom from new metastases (FFNM), and adverse radiation effect (ARE) were quantified. Hippocampal volumes were retrospectively delineated and dosimetry was evaluated in patients treated with upfront SRS. RESULTS: 143 patients with a total of 2198 lesions met criteria for inclusion with 75 patients treated with upfront SRS and 68 treated as salvage from prior WBRT. Median age was 57 (IQR: 46–65) and median KPS 80 (IQR: 70–90). Histologies included breast (n=52), lung (n=49), melanoma (n=30), and other (n=12). Median number of lesions per patient was 13 (IQR 11–17) with median total volume of treatment of 4.1 cc (IQR 2.0–9.9). 12-month FFP per lesion for upfront and salvage treatment was 96.8% (95% CI: 95.5–98.1) and 83.6% (95% CI: 79.9–87.5) respectively (p < 0.001). 12-month FFNM for upfront and salvage FFSRS was 18.8% (95% CI: 10.9–32.3) versus 19.2% (95% CI: 9.7–37.8) respectively (p = 0.90). Mean hippocampal dose was 150 cGy (IQR 100–202). Symptomatic ARE was observed in 2% of patients or 1% of treated lesions. CONCLUSIONS: High rates of local control can be achieved when treating patients with greater than 10 BM with hippocampal doses that are dramatically lower than for HS-WBRT. Hippocampal sparing is readily achievable with expected rates of new metastatic lesions developing in treated patients with low rates of symptomatic ARE.