Search

Pediatric low-grade gliomas (pLGGs) are the most common CNS tumor of childhood and comprise a heterogenous group of tumors. Children with progressive pLGG often require numerous treatment modalities including surgery, chemotherapy, rarely radiation therapy and, more recently, molecularly targeted therapy. We describe our institutional experience using the MEK inhibitor, trametinib, for recurrent/progressive pLGGs. We performed a retrospective, IRB-approved, chart review of all pediatric patients treated with trametinib for recurrent/progressive pLGGs at Dana-Farber/Boston Children’s Cancer and Blood Disorder Center between 2016 and 2018. Eleven patients were identified, of which 10 were evaluable for response. Median age at commencement of trametinib treatment was 14.7 years (range 7.3–25.9 years). Tumor molecular status included KIAA1549-BRAF fusion (n = 4), NF1 mutation (n = 4), FGFR mutation (n = 1) and CDKN2A loss (n = 1). Median number of prior treatment regimens was 5 (range 1–12). Median duration of treatment with trametinib was 19.2 months (range 3.8–29.8 months). Based on modified RANO criteria, best responses included partial (n = 2), minor response (n = 2) and stable disease (n = 6). Two patients remain on therapy (29.8 and 25.9 months, respectively). The most common toxicities attributable to trametinib were rash, fatigue and gastrointestinal disturbance. Five patients required dose reduction for toxicities. Two patients experienced significant intracranial hemorrhage (ICH) while on trametinib. While it is unclear whether ICH was directly attributable to trametinib, therapy was discontinued. Trametinib appears to be an effective treatment for patients with recurrent/progressive pLGG. The toxicities of this therapy warrant further investigation, with particular attention to the potential risk for intracranial hemorrhage. Early phase multi-institutional clinical trials are underway.

Background Off‐label drug prescribing is common in pediatric clinical medicine, though the extent and impact of this practice in pediatric oncology has not yet been characterized. Methods We completed a retrospective single‐institution cohort study evaluating prevalence, characteristics, and clinical outcomes of off‐label prescribing of 108 FDA‐approved targeted anticancer drugs in patients, Off‐label prescribing patterns and outcomes with targeted anticancer agents have not been previously described in pediatric oncology. In this analysis from a large pediatric cancer center, off‐label prescribing of FDA‐approved targeted therapies was common, increasing, encompassed a broad sample of targeted agents, and was tolerable.

Immune checkpoint inhibition through PD-1 and CTLA-4 blockade has shown efficacy in some adult malignancies and generated interest in pediatrics, including central nervous system (CNS) tumors. We describe our experience with immune checkpoint inhibition in recurrent/refractory pediatric CNS tumors. We performed a retrospective chart review of pediatric patients with recurrent or refractory CNS tumors treated with ipilimumab, nivolumab and/or pembrolizumab at Dana-Farber/Boston Children’s Hospital between 2018 and 2019. Eleven patients were identified. Diagnoses included diffuse intrinsic pontine glioma (DIPG) (n = 2), high-grade glioma (HGG) (n = 5), ependymoma (n = 1), craniopharyngioma (n = 1), high-grade neuroepithelial tumor (n = 1) and non-germinomatous germ cell tumor (NGGCT) (n = 1). Eight patients had recurrent disease, while three had refractory disease. Nine patients received combination therapy (ipilimumab/nivolumab); two patients received either nivolumab or pembrolizumab. Median time from diagnosis-to-treatment was 8 months (range 0.8–156). All patients received prior radiation therapy (RT), with median time from RT-to-immunotherapy was 3.8 years. One patient received concurrent then adjuvant immunotherapy with RT. Median duration of treatment was 6.1 months (range 1–25). Therapy was discontinued in nine patients: seven due to disease progression and two due to toxicity (colitis; transaminitis). Other pertinent toxicities included Type 1 diabetes mellitus, hypothyroidism and skin toxicity. Based on iRANO criteria, best responses included partial response (n = 3), stable disease (n = 7) and progressive disease (n = 1). Durable response was noted in two patients. Immune checkpoint inhibition was relatively well tolerated in a cohort of pediatric patients spanning several CNS tumor diagnoses. Results from prospective clinical trials will be critical to inform clinical decisions.

Pediatric brain tumors are associated with high morbidity and mortality, in part due to insufficient understanding of tumor biology. With limited tissue allocation for research from surgical specimens, a key barrier to improving biological understanding, brain tumor autopsies have become an increasingly valuable resource. This study reviews the brain tumor autopsy practice at our institution and describes specific emerging research utilization patterns beyond the clinical autopsy report. We performed a retrospective analysis of brain tumor autopsies at Boston Children’s Hospital (BCH) between 2007 and 2017 and reviewed their consents, neuropathology reports and final diagnoses. We reviewed the method of tissue triaging for research consented autopsies (bioregistry, frozen and fresh tissue) and documented their specific uses. Ninety-six deaths at BCH were due to brain tumors; 56 autopsies were performed (58.3%), of which 49 (87.5%) were consented for research. Tumor mapping was performed on all cases and tissue was allocated for DNA- and RNA-based sequencing studies (published and ongoing). Three tissue allocations with a postmortem interval of 8 h or less resulted in successful cell lines. Tissue from 14 autopsies was contributed to the National DIPG Registry. Our institutional pediatric brain tumor autopsy clinical experience demonstrates the increased utility and wide utilization of autopsy-derived tissue for multiple types of research. These results support the increased efforts to obtain research consent for brain tumor autopsy and active collection of unfixed autopsy material in the molecular era.

Background Ependymoma is the third most common malignant CNS tumor in children. Despite multimodal therapy, prognosis of relapsed ependymoma remains poor. Approaches to therapy for relapsed ependymoma are varied. We present a single-institution retrospective review of the outcomes after first relapse of intracranial ependymoma in children. Procedure We performed a retrospective, IRB-approved chart review of patients with recurrent intracranial ependymoma treated at Dana-Farber/Boston Children's Cancer and Blood Disorders Center from 1990 to 2019. Results Thirty-four patients with relapsed intracranial ependymoma were identified. At initial diagnosis, 11 patients had supratentorial disease, 22 with posterior fossa disease and one with metastatic disease. Median time-to-first relapse was 14.9 months from initial diagnosis (range 1.4-52.5). Seven patients had metastatic disease at first relapse. Gross total resection (GTR) was associated with improved 5-year progression-free survival (PFS) relative to subtotal resection (STR) and no surgery (p = .005). Localized disease at relapse was associated with improved 5-year overall survival (OS) when compared to metastatic disease (p = .02). Irradiation at first relapse seemed to delay progression but was not associated with statistically prolonged PFS or OS. Tumor location, histology, and chromosomal 1q status did not impact outcome at first relapse, although available molecular data were limited making definitive conclusions difficult. Median time-to-second relapse was 10 months (range 0.7-124). Five-year PFS and OS after first relapse were 19.9% and 45.1%, respectively. Median PFS and OS were 10.0 and 52.5 months after first relapse, respectively. Conclusions Relapsed intracranial ependymoma has a poor prognosis despite multimodal therapy. Novel therapeutic strategies are desperately needed for this disease.

INTRODUCTION The incidence of IDH mutations in pediatric glioma is unclear. Recent publications suggest rates ranging between 0–20%. Therapy poses challenges as it is unclear if the biology and prognosis of pediatric IDH-mutant gliomas are identical to adults. METHODS We performed an IRB approved, systematic retrospective search for IDH-mutant gliomas in the Dana-Farber Cancer Institute/Boston Children’s Hospital database between 2009–2018, analyzing incidence, demographics, histology, co-occurring genetic alterations and outcome. RESULTS We identified 575 patients with glioma, ages 0–21 years. Of these, 394 underwent biopsy/resection (0–9 years:n=204; 10–21 years:n=190), with 294 genetic testing. Fifteen of 294 tumors (5%) were IDH1-mutant. Among patients 0–9 years and 10–21 years, 1/156 (0.6%) and 14/138 (10%) had IDH1-mutant tumors, respectively. Among patients 10–21 year old, 13/115 low-grade gliomas were IDH1-mutant (11%). High-grade gliomas accounted for the remaining 23, with one IDH1-mutant glioma (4%). Most common co-occurring genetic alterations for diffuse astrocytoma (n=12) were TP53 (n=9) and ATRX (n=2). Three patients with IDH1-mutant oligodendrogliomas had 1p/19q deletion. Eleven IDH1-mutant patients were evaluable for outcomes with median follow-up of five years. Five-year radiation-free, progression-free and overall survival for patients with low-grade histology were 76% and 100%, respectively. One patient with high-grade glioma recurred 1.2 years after upfront chemo-radiation and died soon after recurrence. CONCLUSION IDH-mutant gliomas comprise a small proportion of pediatric gliomas. Incidence rate is higher in the second decade of life. Comparative analyses between pediatric IDH-mutant gliomas and adult historical cohorts are currently underway, evaluating outcomes, radiation therapy and frequency of malignant transformation.

INTRODUCTION Immune checkpoint inhibition through PD-1 and CTLA-4 blockade has shown efficacy in some adult malignancies and is being investigated in pediatrics. We describe our institutional experience with immune checkpoint inhibition in pediatric CNS tumors. METHODS We performed a retrospective chart review of patients with recurrent, progressive, or refractory pediatric CNS tumors treated with immunotherapy at Dana-Farber/Boston Children’s Hospital between 2018–2019. RESULTS Eleven patients were identified, with median age of 11 years (range:3–9). Diagnoses included DIPG (n=3), HGG (n=4), ependymoma (n=1), craniopharyngioma (n=1), HGNET (n=1) and NGGCT (n=1). Eight patients had recurrent disease (5 local; 3 disseminated); three had refractory disease (non-recurrent). Nine patients were treated with combination therapy (ipilimumab/nivolumab); two patients received monotherapy with either nivolumab or pembrolizumab. Median time from initial diagnosis-to-treatment was 8 months (range 0.8–156). Ten patients received radiation therapy (RT) prior to immunotherapy, with one receiving concurrent RT. Median duration of treatment was 6.1 months (range:1–19). Therapy was discontinued in nine patients: seven due to disease progression and two due to adverse events (colitis, transaminitis). Other pertinent toxicities included type 1 diabetes, hypothyroidism and skin toxicity. Based on iRANO criteria, best responses included partial (n=4), stable (n=6) and progressive disease (n=1). Durable response (>12months) was noted in two patients (HGG and progressive NGGCT). CONCLUSION Immune checkpoint inhibition appears to have clinical benefit and is relatively well tolerated in this cohort of patients. Results from recently completed prospective clinical trials will be critical to inform clinical decisions.

Despite recent advances in the treatment of medulloblastoma, patients in high-risk categories still face very poor outcomes. Evidence indicates that a subpopulation of cancer stem cells contributes to therapy resistance and tumour relapse in these patients. To prevent resistance and relapse, the development of treatment strategies tailored to target subgroup specific signalling circuits in high-risk medulloblastomas might be similarly important as targeting the cancer stem cell population. We have previously demonstrated potent antineoplastic effects for the PI3Kα selective inhibitor alpelisib in medulloblastoma. Here, we performed studies aimed to enhance the anti-medulloblastoma effects of alpelisib by simultaneous catalytic targeting of the mTOR kinase. Pharmacological mTOR inhibition potently enhanced the suppressive effects of alpelisib on cancer cell proliferation, colony formation and apoptosis and additionally blocked sphere-forming ability of medulloblastoma stem-like cancer cells in vitro. We identified the HH effector GLI1 as a target for dual PI3Kα and mTOR inhibition in SHH-type medulloblastoma and confirmed these results in HH-driven Ewing sarcoma cells. Importantly, pharmacologic mTOR inhibition greatly enhanced the inhibitory effects of alpelisib on medulloblastoma tumour growth in vivo. In summary, these findings highlight a key role for PI3K/mTOR signalling in GLI1 regulation in HH-driven cancers and suggest that combined PI3Kα/mTOR inhibition may be particularly interesting for the development of effective treatment strategies in high-risk medulloblastomas.

INTRODUCTION: Pediatric low-grade gliomas (pLGGs) are the most common CNS tumor of childhood and comprise a heterogenous group of tumors. Children with progressive pLGG typically require numerous regimens of therapy including surgical resection, chemotherapy and rarely, radiation therapy. The presence of frequent genetic alterations activating the mitogen-activated-protein-kinase (MAPK) signaling pathway has recently led to a more targeted therapeutic approach. We describe our institutional experience using trametinib, an FDA-approved MEK inhibitor for recurrent/progressive pLGGs. METHODS: We performed a retrospective, IRB approved, chart review of all patients with recurrent/progressive pLGGs treated with trametinib off-study at Dana-Farber / Boston’s Children’s Cancer and Blood Disorder Center between 2016–2018. RESULTS: Eleven patients were identified, with ten evaluable for response. Median age at commencement of trametinib treatment was 14.7 years old (range 7.3–25.9 years; 6 male: 5 female). Molecular status included KIAA 1549-BRAF fusion (n=3), NF1 (n=4), FGFR mutation and heterozygous CDKN2A loss (n=1 each), unknown mutational status (n=2). Median number of prior treatment regimens was 5 (range 1–7). Median duration on treatment with trametinib was 13.4 months (range 1.5–24.8 months). Based on RANO criteria, responses included partial (n=1), minor response (n=1) and stable disease (n=5) as best response. Four patients remain on therapy with trametinib (range 1.5–24.8 months) with two patients having discontinued therapy secondary to disease progression. The most common toxicities were skin rash, fatigue and gastrointestinal disturbance. Five patients required dose reduction for toxicities. Two patients experienced intracranial hemorrhage (ICH) while on trametinib (at 5 and 498 days). While it is unclear whether ICH was directly attributable to trametinib, therapy was discontinued. CONCLUSION: Trametinib appears to be an effective treatment for patients with recurrent/progressive pLGG. The toxicities of this therapy warrant further investigation, particularly the potential risk for intracranial hemorrhage. Early phase clinical trials are underway.

BACKGROUND: BET-bromodomain proteins bind to H3K27ac enhancers and recruit transcriptional proteins to facilitate the expression of genes. Inhibition of BET-bromodomain proteins shows preclinical promise for MYC-driven tumors such as medulloblastoma (MB), however the mechanism of action and means of resistance are not well described. We hypothesized that genes required for MB cell growth and were sufficient to rescue the effects of BET-bromodomain inhibition (BETi) would signify key downstream effectors of this class of inhibitors. METHODS: We applied an integrative genomics approach: expression profiling of genes suppressed following BETi with JQ1, genome-scale CRISPR/Cas9 screens to determine which of the suppressed genes are also essential for cell viability, and a near-genome scale open reading frame (ORF) rescue screen to determine which of the suppressed genes are also sufficient to drive resistance to BETi. Genes that were nominated by all three data sets were considered BETi transcriptional targets responsible for BET-induced reduced cell viability. RESULTS: BETi in MYC-driven MB cell lines generated widespread changes in gene expression. Genes suppressed by BETi had the tendency to be essential to cell survival. The most frequently included pathways were cell-cycle, DNA replication, and RNA processing/transcription. In each cell line, we identified ORF constructs that significantly rescued cells from BETi which included BCL2 family proteins, which also represented genetic dependencies in treatment naïve cell. Overexpression of BCL2 family member BCL2L1 attenuated both apoptosis and JQ1 mediated cell cycle arrest and expression of the pro-apoptotic genes BAX and BAK were required for BETi response. BCL2 family genes were also validated in drug-tolerant cell lines as having been suppressed by BETi and re-expressed in drug-tolerant cells thus demonstrating their relevance in resistance. CONCLUSION: Integrative genomic analysis identifies BCL2 family members as key mediators of response of MB cells to BET-bromodomain inhibition and also mediate resistance to BETi.

INTRODUCTION: Ependymoma is the third commonest malignant CNS tumor in children. Despite multimodal therapy, prognosis of relapsed ependymoma remains poor with long-term progression-free survival (PFS) of approximately 30%. Approaches to treating relapsed ependymoma are extremely varied. Here, we describe our experience and clinical outcomes of children after first relapse of intracranial ependymoma. METHODS: We performed a retrospective, IRB approved, chart review of patients with recurrent intracranial ependymoma treated at Dana-Farber / Boston Children’s Cancer and Blood Disorder Center from 1990 to 2017. RESULTS: Thirty-four patients with relapsed intracranial ependymoma (12 ST, 21 PF, 1 metastatic) were identified. Median time-to-first-relapse was 13.7 months (range:2.0–53.8). Eight patients had metastatic disease at first relapse. Five-year PFS and overall survival (OS) after first relapse was 23.5% and 60.9%, respectively. Median PFS and OS were 12 months and 66 months. Treatments for first relapse included surgery, radiation and/or chemotherapy/biologic therapy. GTR and localized disease were associated with improved OS on univariate analysis, with GTR being an independent prognostic factor for PFS and OS on multivariate analysis (p=0.02 & 0.013, respectively). Median PFS and OS for patients with GTR was 1.5 and 13 years, respectively (versus 0.4 and 2.3 years for STR / no resection [p=0.023 & 0.001, respectively]). Eighteen patients received radiation at first relapse (14 focal radiation, 8 re-irradiation). Patients who were radiated/re-irradiated at relapse showed longer median PFS and OS, however, these did not meet statistical significance. Chemotherapeutic (n=16) and biologic agents (n=3) used in this cohort were heterogenous and did not significantly impact outcome. Median time-to-second-relapse was 10.2 months (range:2.0–124.8). CONCLUSION: Relapsed intracranial ependymoma has a poor prognosis and unrelenting chronic course despite multimodal therapy. GTR is associated with improved outcomes after first relapse and should be pursued when possible. Novel therapeutic strategies are needed for this disease.

BET-bromodomain inhibition (BETi) has shown pre-clinical promise for MYC-amplified medulloblastoma. However, the mechanisms for its action, and ultimately of resistance, have not been fully defined. Using a combination of expression profiling, genome-scale CRISPR/Cas9-mediated loss of function and ORF/cDNA-driven rescue screens, and cell-based models of spontaneous resistance, we identified bHLH/homeobox transcription factors including NEUROD1, NEUROG1 and NEUROG3, and cell-cycle regulators CCND2 and CCND3 as key gene mediators of BETi’s response. Furthermore, these genes also mediated resistance, and were re-expressed in cells that acquire resistance to BETi through chronic dosing. Cells that acquired drug tolerance exhibit altered cell state, with an increase in neuronally differentiated cell-state and expression of lineage-specific bHLH/homeobox transcription factors. However, they did not terminally differentiate, maintaining the expression of SOX2 and the capacity to cycle through S-phase, re-instating expression of CCND2. Analysis of human medulloblastomas revealed heterogeneous populations of cell states, including those that expressed both neuronal and stem markers, and would be predicted to be less sensitive to BETi. Barcoding techniques to track evolution of cells through treatment indicated the existence of cells that are predetermined to acquire resistance to BETi. Finally, we found inhibition of cell-cycling with CDK4/CKD6 inhibition to delay acquisition of resistance, providing a rationale for combination treatments. These findings provide insights into the mechanisms of action of BETi as a therapeutic strategy for MYC-driven medulloblastoma that can be leveraged to increase efficacy in the clinical setting.

Medulloblastoma is a highly malignant pediatric brain tumor associated with poor outcome. Developing treatments that target the cancer stem cell (CSC) population in medulloblastoma are important to prevent tumor relapse and induce long-lasting clinical responses. We utilized medulloblastoma neurospheres that display CSC characteristics and found activation of the PI3K/AKT pathway in sphere-forming cells. Of all class IA PI3Ks, only the PI3Kα isoform was required for sphere formation by medulloblastoma cells. Knockdown of p110α, but not p110β or p110δ, significantly disrupted cancer stem cell frequencies as determined by extreme limiting dilution analysis (ELDA), indicating an essential role for the PI3Kα catalytic isoform in medulloblastoma CSCs. Importantly, pharmacologic inhibition of the MAPK-interacting kinase (MNK) enhanced the antineoplastic effects of targeted PI3Kα inhibition in medulloblastoma. This indicates that MNK signaling promotes survival in medulloblastoma, suggesting dual PI3Kα and MNK inhibition may provide a novel approach to target and eliminate medulloblastoma CSCs. We also observed a significant reduction in tumor formation in subcutaneous and intracranial mouse xenograft models, which further suggests that this combinatorial approach may represent an efficient therapeutic strategy for medulloblastoma. Implications: These findings raise the possibility of a unique therapeutic approach for medulloblastoma, involving MNK targeting to sensitize medulloblastoma CSCs to PI3Kα inhibition.

Glioblastoma multiforme remains the deadliest malignant brain tumor, with glioma stem cells (GSC) contributing to treatment resistance and tumor recurrence. We have identified MAPK-interacting kinases (MNK) as potential targets for the GSC population in glioblastoma multiforme. Isoform-level subtyping using The Cancer Genome Atlas revealed that both MNK genes (MKNK1 and MKNK2) are upregulated in mesenchymal glioblastoma multiforme as compared with other subtypes. Expression of MKNK1 is associated with increased glioma grade and correlated with the mesenchymal GSC marker, CD44, and coexpression of MKNK1 and CD44 predicts poor survival in glioblastoma multiforme. In established and patient-derived cell lines, pharmacologic MNK inhibition using LY2801653 (merestinib) inhibited phosphorylation of the eukaryotic translation initiation factor 4E, a crucial effector for MNK-induced mRNA translation in cancer cells and a marker of transformation. Importantly, merestinib inhibited growth of GSCs grown as neurospheres as determined by extreme limiting dilution analysis. When the effects of merestinib were assessed in vivo using an intracranial xenograft mouse model, improved overall survival was observed in merestinib-treated mice. Taken together, these data provide strong preclinical evidence that pharmacologic MNK inhibition targets mesenchymal glioblastoma multiforme and its GSC population. Implications: These findings raise the possibility of MNK inhibition as a viable therapeutic approach to target the mesenchymal subtype of glioblastoma multiforme. Mol Cancer Res; 14(10); 984–93. ©2016 AACR.

Brain tumors are the most common solid tumor in children, and unfortunately, they carry the highest mortality rate of any of the pediatric cancers. The tumor itself can pose significant risk to the patient due to local effects on the central nervous system which can be life threatening if not addressed quickly. In addition, treatment can also pose significant risks to the patient as damage to the brain has a much slower recovery than any other organ in the body. This can have long-term ramifications to the patient’s overall functionality for his or her lifetime.

Dramatic advances in the molecular analysis of diffuse intrinsic pontine glioma have occurred over the last decade and resulted in the identification of potential therapeutic targets. In spite of these advances, no significant improvement in the outcome has been achieved and median survival remains approximately 10 months. An understanding of the approaches that have been taken to date, why they failed, and how that information can lead the field forward is critical if we are to change the status quo. In this review, we will discuss the clinical trial landscape in North America with an overview of historical approaches that failed and what might account for this failure. We will then provide a discussion of how our understanding of the genotype of this disease has led to the development of a number of trials targeting the mutations and epigenome of diffuse intrinsic pontine gliomas and the issues related to these trials. Similarly, the introduction of methodologies to address penetration across the blood–brain barrier will be considered in the context of both targeted approaches, epigenetic modification, and immune surveillance of these tumors. The comprehensive analysis of these data, generated through cooperative groups, collaborative clinical trials, and pilot studies in North America will be the focus of the IVth Memorial Alicia Pueyo international symposium in Barcelona on March 12th, 2018 and will be compared and contrasted with a similar comprehensive analysis of the European data with the goal of bringing all of these data together to develop a uniform platform on which new rational trials can be based.

Medulloblastoma is the most common paediatric malignant brain cancer and there is a need for new targeted therapeutic approaches to more effectively treat these malignant tumours, which can be divided into four molecular subtypes. Here, we focus on targeting sonic hedgehog (SHH) subtype medulloblastoma, which accounts for approximately 25% of all cases. The SHH subtype relies upon cholesterol signalling for tumour growth and maintenance of tumour-initiating cancer stem cells (CSCs). To target cholesterol signalling, we employed biomimetic high-density lipoprotein nanoparticles (HDL NPs) which bind to the HDL receptor, scavenger receptor type B-1 (SCARB1), depriving cells of natural HDL and their cholesterol cargo. We demonstrate uptake of HDL NPs in SCARB1 expressing medulloblastoma cells and depletion of cholesterol levels in cancer cells. HDL NPs potently blocked proliferation of medulloblastoma cells, as well as hedgehog-driven Ewing sarcoma cells. Furthermore, HDL NPs disrupted colony formation in medulloblastoma and depleted CSC populations in medulloblastoma and Ewing sarcoma. Altogether, our findings provide proof of principle for the development of a novel targeted approach for the treatment of medulloblastoma using HDL NPs. These findings present HDL-mimetic nanoparticles as a promising therapy for sonic hedgehog (SHH) subtype medulloblastoma and possibly other hedgehog-driven cancers.

Mesenchymal (MES) and proneural (PN) are two distinct glioma stem cell (GSC) populations that drive therapeutic resistance in glioblastoma (GBM). We screened a panel of 650 small molecules against patient-derived GBM cells to discover compounds targeting specific GBM subtypes. Arsenic trioxide (ATO), an FDA-approved drug that crosses the blood–brain barrier, was identified as a potent PN-specific compound in the initial screen and follow-up validation studies. Furthermore, MES and PN GSCs exhibited differential sensitivity to ATO. As ATO has been shown to activate the MAPK-interacting kinase 1 (MNK1)-eukaryotic translation initiation factor 4E (eIF4E) pathway and subsequent mRNA translation in a negative regulatory feedback manner, the mechanistic role of ATO resistance in MES GBM was explored. In GBM cells, ATO-activated translation initiation cellular events via the MNK1–eIF4E signaling axis. Furthermore, resistance to ATO in intracranial PDX tumors correlated with high eIF4E phosphorylation. Polysomal fractionation and microarray analysis of GBM cells were performed to identify ATO's effect on mRNA translation and enrichment of anti-apoptotic mRNAs in the ATO-induced translatome was found. Additionally, it was determined that MNK inhibition sensitized MES GSCs to ATO in neurosphere and apoptosis assays. Finally, examination of the effect of ATO on patients from a phase I/II clinical trial of ATO revealed that PN GBM patients responded better to ATO than other subtypes as demonstrated by longer overall and progression-free survival. Implications: These findings raise the possibility of a unique therapeutic approach for GBM, involving MNK1 targeting to sensitize MES GSCs to drugs like arsenic trioxide. Mol Cancer Res; 16(1); 32–46. ©2017 AACR.

BACKGROUND: Current brain tumor therapies are focused on reducing tumor burden by inducing apoptotic and necrotic tumor cells. However, inflammation in the tumor microenvironment caused by this debris can accelerate tumor progression. Inflammation is endogenously regulated resolvins and protectins, molecules that are biosynthesized and act in the brain and cerebrospinal fluid. They clear cellular debris via local macrophages while reducing localized inflammatory cytokines. We hypothesized that control of inflammation through resolvins and protectins could represent a novel treatment modality by pharmacologically promoting the clearance of tumor cell debris by microglia, thereby depriving the surviving tumor cells of inflammatory stimuli. METHODS AND RESULTS: We show that debris generated by killing of either medulloblastoma or glial tumors (using cisplatin chemotherapy, dabrafenib targeted therapy or the epigenetic inhibitors JQ1) stimulate residual live tumor cells to grow. This effect was inhibited by nanogram levels of resolvins/protectins and without toxicity in a variety of tumor types including orthotopic models. The effect was mediated via enhanced macrophage and microglial phagocytosis of tumor cell debris and counter-regulation the secretion of pro-inflammatory cytokines/chemokines, including CCL5, TNFα, CCL2, CXCL1, and CCL4. CONCLUSIONS: We show that resolvins and protectins stimulate the clearance of tumor cell debris by stimulating macrophage and microglial phagocytosis and counter-regulating pro-inflammatory and pro-tumorigenic cytokines. This approach not only has relevance to our current approaches to cancer therapy (focused on maximal tumor cell kill), but may have an impact on the side effects of current immunotherapies where therapy induced inflammation in the tumor results in significant morbidity.

BACKGROUND: The phosphoinositide 3-kinase (PI3K) and sonic hedgehog (SHH) pathways play important roles in medulloblastoma (MB) and other pediatric cancers. Aberrant activation of the PI3K pathway has been shown to be an important regulator of cancer cell proliferation, metabolism, protein synthesis and apoptosis. Additionally, mTOR activation contributes to therapy resistance likely in part through SMO-independent activation of the transcription factor GLI. Thus we sought to evaluate the effects of combined PI3K/mTOR inhibition in medulloblastoma and Ewing sarcoma. METHODS: Medulloblastoma (DAOY and D556) and Ewing Sarcoma cell lines (TC71, RDES) were grown in 2-D culture to investigate the effects of pharmacologic inhibition of PI3K and mTOR (using BYL-719, a p110α isoform specific PI3K inhibitor and OSI-027, a catalytic inhibitor of mTOR1/2, respectively) on protein kinase signaling, cell proliferation, colony formation, intracellular localization of GLI, and GLI target gene expression. RESULTS: Of all four class I PI3Ks only the p110α isoform is required for MB cell proliferation and colony formation. Pharmacologic targeting of the p110α isoform of PI3K with BYL-719 synergized with the catalytic mTORC1/2 inhibitor OSI-027, resulting in inhibition of effector pathways, and blocked cell proliferation and colony formation in both medulloblastoma and Ewing Sarcoma. Moreover, dual PI3K/mTOR inhibition resulted in decreased nuclear localization of GLI and reduced GLI target gene expression in both pediatric tumor cell lines. CONCLUSIONS: Inhibition of p110α and mTOR exerts potent antineoplastic effects on cancer cell proliferation and transformation. This effect may be due in part to inhibition of GLI nuclear localization. Thus, besides established roles in inhibition of cancer cell proliferation and protein synthesis, dual inhibition of p110α and mTOR is particularly promising for targeting SHH-driven cancers such as medulloblastoma and Ewing sarcoma.

Our recent work suggested that inhibition of mTORC1 activates Mnk in a PI3K-dependent manner, thereby providing a survival mechanism for medulloblastoma cells. The PI3K-Akt axis represents an important survival pathway that also confers therapy resistance to medulloblastoma stem cells, resulting in tumor recurrence. Here, we investigated a role for p110 isoforms of PI3K in medulloblastoma stem-like cancer cell biology and studied the potential of Mnk inhibition for sensitizing medulloblastoma stem-like cancer cells and orthotopic xenograft tumors to PI3K inhibition. We used medulloblastoma cell lines Daoy and D556 grown as conventional 2-D cultures or under stem-cell conditions as 3-D neurospheres to elucidate the roles of PI3K-Akt signaling in medulloblastoma proliferation, colony formation and stem cell function. We employed extreme limiting dilution analysis (ELDA) to ask whether concomitant Mnk inhibition enhances antineoplastic effects of PI3K inhibition on cancer stem cell growth. Additionally, in a preliminary intracereballar xenograft mouse study, we investigated the effects of pharmacologic PI3K and Mnk inhibition. We found that Akt activity greatly increased when 2-D cultures were converted into 3-D neurospheres. This Akt activation coincided with increased expression of CD133 and nestin, suggesting an important role for PI3K-Akt signaling in medulloblastoma stem cells. The p110a specific inhibitor BYL-719 blocked this Akt activation in neurospheres indicating this Akt activation is mediated by p110a. Consistently, of all class I PI3K catalytic isoforms (p110a, p110b, p110d and p110g) only knockdown of p110a disrupted stem cell frequencies in ELDA. We previously reported that mTORC1 inhibition engages Mnk signaling in a negative feedback manner to promote survival. Here we show that Mnk inhibition by CGP57380 sensitized medulloblastoma cells for pharmacologic inhibition and siRNA-mediated knockdown of p110a both in 2-D cancer cells and 3-D stem-like cancer cell cultures. After intracerebellar injection of medulloblastoma cells in nude mice, we found that combined targeting of PI3K-p110a and Mnk results in inhibition of tumor growth and increased survival. In summary, pharmacologic inhibition of PI3K-p110a by BYL-719 showed potent activity against medulloblastoma cells and stem-like cancer cells. Knockdown of p110a disrupted cancer stem cell frequency in ELDA and this effect was greatly enhanced by pharmacologic inhibition of Mnk. Finally, in an orthotopic mouse model we found that concomitant inhibition of p110a and Mnk prolonged survival and reduced tumor size. The striking effects of concomitant Mnk inhibition on stem-like cancer cells, neurospheres and tumors is particularly interesting as it suggests enhanced vulnerability of the therapy-resistant, tumor-initiating cancer stem cell population to p110a inhibition in medulloblastoma. Citation Format: Jonathan B. Bell, Frank Eckerdt, Quanhong Ma, Jessica R. Clymer, Stewart Goldman, Rintaro Hashizume, Leonidas C. Platanias. Mnk targeting enhances vulnerability of medulloblastoma stem-like cancer cells to PI3K-p110alpha inhibition. [abstract]. In: Proceedings of the 107th Annual Meeting of the American Association for Cancer Research; 2016 Apr 16-20; New Orleans, LA. Philadelphia (PA): AACR; Cancer Res 2016;76(14 Suppl):Abstract nr LB-009.

MB-67. DUAL TARGETING OF PI3K AND mTOR SIGNALING IN MEDULLOBLASTOMA Jessica Clymer1,2, Frank Eckerdt2, Jonathan Bell2, Rishi Lulla1,2, Stewart Goldman1,2, and Leonidas Platanias2,3; Division of Hematology/ Oncology/Stem Cell Transplantation, Department of Pediatrics, Ann & Robert H. Lurie Children’s Hospital, Chicago, IL, USA; Robert H. Lurie Comprehensive Cancer Center, Feinberg School of Medicine, Northwestern University, Chicago, IL, USA; Division of Hematology/Oncology, Department of Medicine, Feinberg School of Medicine, Northwestern University, Chicago, IL, USA BACKGROUND: Aberrant activation of the PI3K pathway has been shown to play a role in medulloblastoma cell proliferation, while mTOR activation contributes to therapy resistance. Here we sought to evaluate the effects of combined targeting of the PI3K and mTOR pathways in medulloblastoma. METHODS: Medulloblastoma cell lines Daoy and D556 were grown in 2-D cultures to investigate the effects of pharmacologic inhibition of PI3K and mTORon kinase signaling, medulloblastomacell proliferation, colony formation, and apoptosis. After knockdown of p110 isoforms, cells were also grown under stem cell conditions as 3-D neurospheres and subjected to extreme limiting dilution analysis (ELDA) for assessment of stem-like cancer cell growth. RESULTS: Of all class I PI3K isoforms (p110a, p110b, p110d, and p110g), only p110a was essential for medulloblastoma cell proliferation, colony formation and neurosphere growth. Accordingly, pharmacologic targeting of the p110a isoform of PI3K with BYL-719 synergizedwith the mTOR inhibitor OSI-027, as it resulted in inhibition of effector pathways, blocked cell proliferation and colony formation, and increased malignant cell death by apoptosis. Finally,BYL-719and OSI-027 greatly impairedgrowthof stem-likecancer cells as 3-D neurospheres. CONCLUSIONS: The p110a isoform of PI3K exerts crucial roles in medulloblastoma. Inhibition of p110a and mTOR exerts potent antineoplastic effects on medulloblastoma cells and also blocks neurosphere growth. Dual inhibition of p110a and mTOR might be promising for targeting both malignant medulloblastoma cells and the therapy-resistant cancer stem cell population. Neuro-Oncology 18:iii97–iii122, 2016. doi:10.1093/neuonc/now076.63 #The Author(s) 2016. Published by Oxford University Press on behalf of the Society for Neuro-Oncology. All rights reserved. For permissions, please e-mail: journals.permissions@oup.com.