Work

Update from the 2013 international neurofibromatosis conference

Year: 2014

Type: article

Abstract: The 2013 Neurofibromatosis (NF) Conference took place at the Portola Hotel and Spa, Monterey, CA, from June 8–11, 2013. This international meeting is sponsored annually by the Children's Tumor Foundation (CTF), with the goal of bringing together NF researchers and clinicians from disparate fields of expertise. The conference agenda included a range of preclinical topics with a focus on signaling pathways and mouse models and a number of clinical topics, including a symposium on the interaction of academics, government, and industry in NF clinical trials. Neurofibromatosis is a group of inherited genetic disorders–NF1, NF2, and schwannomatosis – that together affect about 100,000 persons in the US. NF1 is the most common, with an estimated birth prevalence of 1:3,000 [Friedman, 1999; Lammert et al., 2005]; for NF2 and schwannomatosis, the estimated birth prevalence is 1:25,000 and 1:40,000, respectively [Antinheimo et al., 2000; Evans et al., 2005]. NF1, NF2, and schwannomatosis have in common that they predispose affected individuals to develop Schwann cell tumors such as neurofibromas and schwannomas. At a lower frequency, they also predispose to the development of a number of other benign and malignant tumor types, as well as some developmental abnormalities including learning disabilities (in NF1). The disorders arise from mutations in different genes, each of which plays a key role in regulating cellular function. The NF1 gene on human chromosome 17 encodes an intracellular signaling molecule that functions as a GTPase activating protein for Ras proteins, whereas the NF2 gene on human chromosome 22 encodes a cytoskeletal-membrane linking protein with reported roles in the suppression of several different growth-associated signaling pathways. The biology of schwannomatosis remains understudied despite the identification of germline SMARCB1 mutations in schwannomatosis patients in 2007. Approximately 30–70% of individuals with neurofibromatosis type 1 (NF1) have learning disabilities, representing the most significant cause of lifetime morbidity associated with this disease. Moreover, recent findings indicate that signaling proteins within the Ras and the extracellular-signal regulated kinase (ERK) subfamily of mitogen-activated protein kinases (MAPK) pathway are mutated and hyperactivated in a number of human genetic syndromes (known as RASopathies) associated with varying levels of cognitive dysfunction substantiating the view that these pathways are important in regulating brain function [Rauen, 2013]. Yuan Zhu (University of Michigan) presented evidence that bi-allelic, but not monoallelic, inactivation of Nf1 in developing neural stem cells increased the number of glial lineage cells but by distinct mechanisms depending on the region of the brain. In the dorsal forebrain Nf1 loss in neural stem and progenitor cells altered their fate-specification resulting in increased numbers of glial cells in the corpus callosum at the expense of neurogenesis in the olfactory bulb. In contrast, in the ventral forebrain, overproduction of glial cells resulted from increased proliferation of glial-restricted progenitors. Importantly, both of these glial defects appeared to be caused by hyperactivation of Ras/ERK signaling, as a MEK inhibitor rescued the deficits. Yuan Wang (University of Michigan) presented data on the effects of biallelic loss of the Nf1 gene in the cerebellum of mice and provided further evidence for the importance of the ERK pathway showing that hyperactivity of ERK underlies the defects seen in both neurons and glia of these mice, disrupting cerebellum structure and causing motor defects. William Snider (University of North Carolina) performed both loss-of-function and gain-of-function studies on MEK1/2 in developing neural stem cells and post-mitotic neurons and showed that while loss of MEK1 and MEK2 impaired the generation of glial cells, hyperactive MEK/ERK signaling promoted gliogenesis during development. These results provide further genetic evidence that supports the observations of the Zhu group and others that accurate regulation of Ras signaling through the MEK/ERK pathway is critical for normal brain development. Alcino Silva (University of California, Los Angeles) described the possible mechanisms by which loss of one allele of Nf1 can cause deficits in long-term potentiation (LTP) and learning in adult mice. Dr. Silva also presented evidence that mutations in SHP2 (identified in Noonan syndrome) cause similar deficits in LTP and learning. However, he emphasized that what appeared to be a similar phenotype occurred via distinct mechanisms, as the genetic mutations caused their effects in distinct cell types. Freda Miller (University of Toronto, Canada) and her colleagues investigated signaling proteins within the Ras pathway and their activating mutations in embryonic neural stem cells of the murine cortex, and arrived at two major conclusions. First, each protein, including SHP2, B-Raf, and Ras, plays an important role in regulating the genesis of neurons and astrocytes from developing neural precursors. Second, that when these proteins are hyperactivated, as in the Rasopathies, they deregulate cell genesis by distinct mechanisms. Debra Mayes (Cincinnati Children's Hospital Medical Center) presented an analysis of the histological and behavioral changes of mice with hyperactive RAS signaling in oligodendrocytes. She hypothesized that upregulation of nitrous oxide plays a critical role in the phenotypes, which could be reversed with antioxidant treatment. Thomas De Raedt (Brigham and Women's Hospital) presented work characterizing a more severe NF1 phenotype associated with microdeletions of the NF1 region. He identified an additional gene within the deletion and, in a mouse model, recapitulated a more severe NF1 phenotype when this gene was lost in conjunction with the NF1 gene. Nancy Ratner (Cincinnati Children's Hospital Medical Center) presented further work using the DhhCre; NF1fl/fl mouse model to test the requirement of specific signaling pathways for tumor growth. The studies verified the MEK/ERK pathway as a therapeutic target and identified STAT-3 as a potential new target. In studies of NF1-associated low grade gliomas, David Gutmann (Washington University) presented evidence that human tumors are monoclonal and have no other mutations other than loss of NF1 gene expression [Gutmann et al., 2013]. However, in a genetically engineered mouse optic glioma model, loss of Nf1 gene expression in glial cells alone is not sufficient to drive tumorigenesis but requires signals from Nf1+/− microglia in the surrounding microenvironment [Pong et al., 2013]. Luis Parada (UT Southwestern) presented work on a novel glioma mouse model in which stem cells are labeled and can be ablated following ganciclovir treatment. A stem cell population was found to be labeled in glioblastomas in these mice and was relatively quiescent. Following treatment of the tumor however, these stem-like cells differentially proliferated and were responsible for the regrowth of the tumor. These results reinforce the importance of targeting the stem-like cells found in many tumors. Dinorah Friedmann-Morvinski (The Salk Institute) presented a different model of glioblastoma in which tumors were induced from fully differentiated cortical neurons rather than from stem cells, demonstrating that the cell-of–origin may not always be obvious. MPNST is the primary cause of early mortality in NF1 patients [Evans et al., 2011]. Using a Sleeping Beauty genetic screen, Adrienne Watson (University of Minnesota) identified Wnt signaling as a driver of Schwann cell tumorigenesis. Induction of Wnt signaling was sufficient to induce a transformed phenotype in human Schwann cells, while inhibition of both Wnt signaling and mTOR synergistically induced apoptosis of MPNST cell lines. Rebecca Dodd (Duke University) described a mouse model of soft tissue sarcoma in which Nf1 and lnk4a were knocked out by the injection of a Cre-expressing adenovirus into either muscle or sciatic nerve. Treatment with a MEK inhibitor decreased sarcoma growth in these mice. Up to two-thirds of individuals with NF1 demonstrate signs of cognitive dysfunction, ranging from deficits in attention, visual-spatial memory and executive function, to language problems, learning disabilities, and academic underachievement. Nicole Ullrich (Boston Children's Hospital, Harvard Medical School) summarized the major challenges in clinical trial design for neurocognitive interventions, including trial design, selection of ideal outcome measures, selection of the appropriate drug and duration for intervention, dose-finding and appropriate inclusion criteria. Maria Ribeiro (Coimbra University, Portugal) presented data using magnetic resonance spectroscopy to measure brain metabolites such as the inhibitory neurotransmitter GABA in the visual cortex of children and adolescents with NF1. Mouse models have implicated increased release of GABA in the pathophysiology of cognitive issues [Cui et al., 2008]. Children with NF1 have lower measured levels of unbound GABA, which might reflect depletion of GABA from synaptic vesicles [Violante et al., 2013]. Within the NF1-group, higher overall GABA-levels correlated with slower reaction times. Thijs van der Vaart (Erasmus MC, The Netherlands) reported on the final results of the NF1-SIMCODA trial, a 12-month randomized placebo-controlled trial in 84 NF1-children aged 8–16 years. Earlier work had shown positive effects of lovastatin, a cholesterol-lowering drug, in a mouse model of NF1 [Li et al., 2005]. The NF1-SIMCODA trial aimed to detect beneficial effects of simvastatin on cognitive and behavioral problems in children with NF1. The current randomized controlled trial (RCT) was based on a previous 12-week pilot RCT with 61 NF1-children in the same age range, in which only one secondary outcome measure, the IQ-subtest "object assembly", demonstrated a positive effect of simvastatin over placebo [Krab et al., 2008]. The results of NF1-SIMCODA were disappointing, showing no benefit of simvastatin over placebo for treating cognitive and behavioral problems [van der Vaart et al., 2013]. Kristina Hardy (Children's National Medical Center) presented data on the use of a computerized training program, CogMed, in children with NF1. CogMed has been proposed as a neurocognitive intervention in an upcoming clinical trial through the NF Clinical Trials Consortium. Recognizing that clinical trials for NF1 learning disability are currently underway, the Response Evaluation in Neurofibromatosis and Schwannomatosis (REiNS) group has been working to recommend a standardized set of outcome measures in clinical trials for NF. Karin Walsh (Children's National Medical Center) summarized the progress of the REiNS working group on neurocognitive outcome measurements. One concern of commonly utilized neuropsychological measures is that they are not designed to detect change over time. The ideal outcome measures for cognitive interventions should measure constructs of interest that are relevant to individuals with NF1, should detect change over time, and should be generalizable. Peter de Blank (Rainbow Babies & Children's Hospital, Case Western Reserve University) retrospectively evaluated the ophthalmologic records and diffusion tensor imaging measurements of the optic pathway in children with NF1 and optic pathway gliomas. He found that a decrease in fractional anisotropy of the optic radiations was associated with abnormal visual acuity and could be a predictor of visual acuity loss in the following year [de Blank et al., 2013]. Susan Huson (Oxford, UK) presented a study evaluating the prevalence of autistic features in a large population of children 4–16 years of age with NF1. In this population-based study, she found that approximately 30% of subjects met the diagnostic criteria for an autistic spectrum disorder [Garg et al., 2013]. These findings build on previous reports suggesting that autistic symptoms are common in NF1 patients but contrasts with a prior report in adult NF1 patients. [van Eeghen et al., 2013; Walsh et al., 2013]. Amanda Bergner (Johns Hopkins School of Medicine) presented the results of a pilot project evaluating the use of a dedicated adult intake clinic at the Johns Hopkins Comprehensive NF Center in improving the efficiency of care. The intake clinic significantly reduced patient wait time, allowed for more effective time to be spent with NF providers, and generated new revenue for the clinic. Brian Weiss (Cincinnati Children's Hospital Medical Center) presented the results from a 2-stratum phase II clinical trial of the mTOR inhibitor sirolimus for plexiform neurofibroma (PNF) in subjects with NF1. This trial specifically sought to determine whether sirolimus either: (a) increases time to progression (TTP) in progressive PNF (stratum 1) or; (b) results in objective radiographic responses in inoperable PNFs in the absence of documented radiographic progression at trial entry (stratum 2). Disease status was evaluated using volumetric MRI analysis at regular intervals; pain reduction and quality of life (QOL) outcomes were also assessed. No patients on stratum 2 experienced a radiographic response. [Weiss et al., 2013] In contrast, patients in stratum 1 had an estimated median TTP of 15.4 months, compared to 11.9 months in a control group from a completed NCI-sponsored study with near identical eligibility criteria. Brigitte Widemann (National Cancer Institute) reported the preliminary results of an ongoing phase 1 trial of selumetinib, a MEK inhibitor, in children 3–18 years of age with NF1 and inoperable PNFs. Selumetinib is administered orally BID on a continuous dosing schedule. The maximally tolerated dose will be determined based on toxicities observed in the first three courses. Disease status is evaluated using volumetric MRI analysis at regular intervals. Nine subjects had enrolled to date (median age 15 years; range, 5–18 years). No dose limiting toxicities (DLTs) were observed at the 20 mg/m2 dose, but one of three initial subjects developed a DLT at 30 mg/m2 which was reversible and asymptomatic. This dose level has since been expanded to six subjects, with no additional DLTs to date. Decreases in PNF volume have been observed in all subjects who have undergone at least one restaging MRI, and partial response (≥ 20% shrinkage in tumor volume) has been achieved in four of the six subjects. Chronic dosing of selumetinib has been tolerated by children with NF1 and PNF, and enrollment of this study is ongoing. David Viskochil (University of Utah) presented the results of a phase II trial of chemotherapy in sporadic and NFI-associated high-grade malignant peripheral nerve sheath tumors (MPNSTs). Ifosfamide and doxorubicin followed by 2 cycles of ifosfamide and etoposide were used. The objective response rate was lower in NF1 than in sporadic MNST patients, which is consistent with retrospective literature reports. Merlin interacts with the E3 ubiquitin ligase CRL4-DCAF1 in the nucleus and inhibits its function [Li et al., 2010]. Filippo Giancotti (Memorial Sloan-Kettering Cancer Center) described a functional link between Merlin, CRL4-DCAF1, and LATS in human Schwann cells. These results provide a mechanism for Merlin's activation of Lats and the Hippo tumor suppressor pathway in the nucleus of Schwann cells. In addition, they indicate that YAP/TEAD inhibitors, which are undergoing preclinical testing, may be effective for the treatment of NF2. Duojia Pan (Johns Hopkins University) discussed recent efforts in targeting the Hippo pathway as a potential strategy against NF2. He presented findings showing that loss of NF2 mimics inactivation of Hippo signaling in the liver, the lens of the eye, and ependymal cells, suggesting that NF2 is a physiological regulator of Hippo signaling in multiple mammalian tissues. Disruption of the YAP-TEAD complex, the nuclear effector of the Hippo pathway, potently suppresses hepatomegaly/tumorigenesis resulting from Nf2 inactivation in the mouse liver. Screening of a library of drugs approved by the U.S. Food and Drug Administration (FDA) identified verteporfin as a small molecule that inhibits TEAD-YAP association and YAP-induced liver overgrowth, providing proof of principle for YAP inhibitors as molecular-targeted therapeutics for NF2 [Liu-Chittenden et al., 2012]. Joe Kissil (The Scripps Research Institute) detailed the identification of a novel Merlin-interacting protein – angiomotin (Amot) – associated with tight and adherens junctions and an interactor of several small G-protein modulators. Amot directly associates with Yap, a transcriptional activator regulated by the Hippo pathway. Moreover, Amot is required for hepatic "oval cell" response and tumorigenesis in response to toxin-induced injury or Nf2 loss in a liver-specific Amot knockout mouse. In these studies, Amot appears to regulate Yap at multiple levels, both in the cytoplasm and in the nucleus. These studies implicate Amot as a critical downstream effector of Merlin and the Hippo/Yap pathway. Dominique Lallemand (Curie Institute, France) reported the use of a combination of proteomic approaches to evaluate the expression, activity, and link to tumor cell proliferation of several of the signaling pathways regulated by Merlin in a large series of human schwannoma biopsies. They found that Her3, Her2, PDGFR-β, and Axl are the most frequently activated receptor tyrosine kinases (RTKs) in schwannomas and could represent key therapeutic targets. Using Reverse Phase Protein Array, they showed that the expression of the Ki67 proliferation marker is strongly associated with levels of Yap, activated Her3, and PDGFR-β. Also, the expression of PDGFR-β and Her3 can be stimulated by Yap in schwannoma cells in culture suggesting that a signaling network under the control of Yap may be involved in tumor growth. Finally, they observed remarkable heterogeneity in the expression of all the proteins assessed by immunohistochemistry on tumor sections, that the response to will be from patient to (Curie Institute, France) presented data that Merlin and YAP directly with the that Merlin YAP in the from Institute, presented data showing loss of Merlin in to reduced expression of the 1 type They are currently the effects and the for the seen in patients with NF2. Medical School) this by her studies of the of the model of liver progenitor and tumorigenesis. this analysis a function for Merlin during liver development. France) presented an on mouse models of human et al., 2013] showing an important role of for the and malignant progression of associated with NF2 gene recent showed that mutations in and are with NF2 loss in et al., 2013; et al., 2013]. The current strategy to these genetic in mice was School of Medicine) described a genetically engineered NF2 mouse model with of the Nf2 gene by expression under control of a months of of these mice develop and nerve tumors that are identical to human schwannomas. In addition, the development of nerve tumors in these mice may with functional in and as measured by response and an early Schwann cell progenitor tumor cell of (University of discussed the of a library using mouse Schwann cells in to that reduced cell Dr. identified two from an initial that reduced of mouse Schwann cells as compared to normal mouse Schwann cells. In a a number of from the drug as well as in clinical trials for NF2. This demonstrated the of the cell and to drugs for human Research Institute) reported on of inhibitors for NF2. The of was in cell and in human primary schwannoma and in and in two models and in one Nf2 model in et al., 2013]. They found that induced of proteins and proliferation of cells. In addition, expression analysis using a identified of implicated in cell cell and Schwann cell expression was altered by treatment. results showed that in schwannoma multiple pathways for tumorigenesis. Cancer Center) reported that Nf2 mouse cells higher levels of and and that the inhibitor, inhibits cell proliferation in In the treatment of NF2 tumors with alone has shown Using a mouse model, Hospital, Harvard Medical School) showed results using a combination of and and Foundation UK) presented the results of her study of quality of life (QOL) in patients with NF2 using the NF2 on et al., 2012]. This measure was and to and demonstrated and to detect significant changes over time in this patient Hospital) data from the NF2 using new measures by the Response Evaluation in NF and Schwannomatosis (REiNS) schwannomas were evaluated for volumetric progression and was assessed with The median time to tumor progression was A significant number of patients experienced tumor progression and over years et al., (Johns Hopkins University) presented the results of a study of a in subjects years of age with NF2 and schwannomas were for 1 with performed during treatment and of treatment. The primary was a in compared with MRI, MRI, and outcomes were also assessed. subjects (median age 30 years; range, were No subjects had progressive loss while on treatment. of subjects had significant in for three months, and was for subjects. of subjects with an also had significant reduction in tumor volume on These data that is well tolerated in this patient population and that can loss to in patients, of tumor France) presented the results of a trial of the mTOR inhibitor for progressive in NF2 patients. The for the trial was based on preclinical results in a NF2 genetically engineered mouse schwannoma model and and consistent results in an patient et al., no tumor shrinkage has been However, preliminary analysis that inhibition may time to tumor In the year of the trial, patients who tumor growth of will be allowed to the This design is to determine whether on a clinical response to treatment. University) reported the results of a clinical trial using in patients with NF2 and progressive schwannomas. No objective in or imaging response was seen et al., Medical School) described her with and to develop an that has a Schwann injection of the of schwannomas and a reduction in pain in human mice in sciatic nerve. This effect may via a of and has the potential to control growth of tumors that are not directly to the and University) discussed the mechanisms of pain in He described a SMARCB1 knockout mouse with that showed a similar expression and to type cells, but which were to to expression from dorsal of these mice showed suggesting that this pathway could be for therapeutic (University of presented data showing that multiple can be inherited as a distinct genetic caused by mutations in the Medical Netherlands) presented an analysis of SMARCB1 protein expression in schwannomas from schwannomatosis patients, showing that of SMARCB1 can be found in these tumors and that the type of the of protein However, the type of not the of the of in tissue Hospital) presented SMARCB1 analysis and SMARCB1 expression in and sporadic schwannomas. He found a between the rate of SMARCB1 expression in schwannomas schwannomatosis, sporadic schwannomatosis, and low of mutations in the SMARCB1 suggesting other mechanisms may a role in tumor development. In addition, despite a rate of of SMARCB1 expression in NF2 SMARCB1 mutations were not in of the tumors. Foundation UK) presented a report of a patient with schwannomatosis with three and This further the clinical between syndromes and the importance of in an as well as the for genetic and of Amanda Bergner presented the progress of the Schwannomatosis The is a of schwannomatosis patients, and was designed as a to who be for clinical trials or The is to by academic the and a patient entry mechanism was to patients not with an to The 2013 conference novel including small group and an NF1, NF2, and schwannomatosis are by clinical a clinical can be to the in tumor in the and in the clinical with a of other (University of discussed clinical and of in NF1 and in In addition, an of other associated with multiple were including NF2, complex, and She that the diagnostic criteria for NF1 are not as some patients with signs only and

Source: American Journal of Medical Genetics Part A

DOI: 10.1002/ajmg.a.36754