|Home | About | Journals | Submit | Contact Us | Français|
The quiescent state is thought to be a characteristic property for the maintenance of hematopoietic stem cells (HSCs). Interaction of HSCs with their particular microenvironments, known as the stem cell niches, is critical for adult hematopoiesis in the bone marrow (BM). We demonstrate that quiescent HSCs adhere to osteoblasts (OBs) in the niche through the Tie2/angiopoietin-1 and/or mpl/thrombopoietin. Since Ang-1 enhanced the expression of N-cadherin in HSCs and induced the cell adhesion to bone, we examined the role of N-cadherin in vivo. When HSC cells transfected with dominant negative N-cadherin, which lacks extracellular domain, HSC cells can reach bone marrow but not enter the osteoblastic niche, resulting in the defect in hematopoietic reconstitution. When the BM is ablated during BM transplantation or after treatment with myelosuppressive agents, the quiescent HSCs enter the cell cycle and proliferate to supply progenitors of committed hematopoietic cells. We found that reactive oxygen species (ROS) induce the exit of HSCs from the niche after 5FU injection through downregulation of N-cadherin. Similarly, ROS was elevated after serial BM transplantation, and upregulation of MAPK p38 and INK4A was detected only in HSCs. In vivo treatment of recipients with antioxidant or p38 MAPK inhibitor restored exhaustion of HSCs in the repeated BM transplantation. We are now studying the cellular metabolic changes of HSCs. I will show that the metabolism of quiescent stem cells is glycolysis-dependent in the hypoxic niche, while that of proliferating progenitors are mitochondria-dependent, using HIF-1α/VHL-deficient mice. Moreover, I will discuss the therapeutic strategy targeting the niche of cancer cells.
Gliolan (ALA) has been developed and approved in Europe for intra-operative identification of malignant gliomas by inducing fluorescent porphyrins in these tumors. Data from the pivotal approval study have consequently increased the understanding for the principle value of cytoreductive surgery in malignant gliomas. Since then, novel data have been collected that underlines the value of cytoreductive surgery in the context of recently approved adjuvant therapies such as BCNU-wafers or concomitant radiochemotherapy with temozolomide. On the other hand, ALA-derived porphyrins are potent photosensitizers with efficacy for photodynamic therapy and demonstrate preliminary efficacy for stereotactic phototherapy of malignant gliomas. Preliminary experimental evidence indicates immunostimulation to play a role in mediating this efficacy. Finally, ALA-induced tumor porphyrins may be of value in other tumors relevant to neurosurgeons, such as pituitary adenomas, medulloblastomas, meningiomas, or spinal chord tumors.
Adoptive cell transfer immunotherapy is based on the infusion of tumor reactive T cells into patients with cancer. This approach has multiple advantages compared to other forms of immunotherapy. It is possible to administer large numbers of highly selected cells with high avidity for tumor specific antigens that are required for CNS tumors, and the cells can be activated ex vivo to exhibit antitumor effector function and thus do not depend on activating stimuli at the tumor site. The exact cell sub-populations and effective functions that are required for cancer regression in vivo can be identified and it is possible to manipulate the host prior to cell transfer to provide an altered environment for the transferred cells. We are exploring the use of engineered T cells bearing chimeric receptors. As a model for nonimmunogenic tumors, we are targeting the surface membrane glycoprotein mesothelin is a promising target for the immunotherapy of mesothelioma, ovarian, and pancreatic tumors due to the uniform overexpression of mesothelin and the benign phenotype of mesothelin null mice. We hypothesize that previous trials of adoptive immunotherapy for cancer that have used CTL have failed due to poor trafficking to sites of tumor, and insufficient effector functions to self antigens. Our preclinical data indicates that use of lentiviral engineered T cells with chimeric receptors that incorporate a “tumor resistance genotype” should have improved function for cancer immunotherapy. We have tested mesothelin redirected T cells in humanized mouse models bearing tumor xenografts. The T cells are able to eradicate large, well established tumors at an E:T ratio of at least 1:70 in vivo. Three routes of administration with redirected T cells have been tested in mice bearing xenografted flank tumors. The intraperitoneal route was found to be inferior to intravenous or direct intratumoral injection of the redirected T cells. Long-term engraftment of the redirected T cells correlates with antitumor efficacy. This strategy could be adapted to CNS tumors, where the engineered T cells could prove to have resistance to the highly immunosuppressive tumor microenvironment that is characteristic of glioblastoma.
Human brain tumors appear to have a hierarchical cellular organization, suggestive of a stem cell foundation. We have recently derived adherent cell lines with high efficiency from human malignant glioma that display stem cell properties and initiate high-grade gliomas following xenotransplantation. Significantly, these cell lines from different tumors exhibit divergent gene expression signatures and differentiation behavior that correlate with specific neural progenitor subtypes. The diversity of gliomas may therefore reflect distinct cancer stem cell phenotypes. Adherent brain tumor stem cell lines offer significant experimental advantages compared to “sphere”-based cultures, and provide a simplified model enabling refined studies of cancer stem cell behavior. In addition, it remains unclear if genetically engineered mouse models of cancer recapitulate the functional heterogeneity observed in their human counterparts. We demonstrate medulloblastomas arising from Patched1 (Ptc1±)deficient mice contain a subpopulation of cells that demonstrate a neural precursor phenotype, show clonogenic and multilineage differentiation capacity, wild-type Patched-1 expression, and the ability to initiate tumors following allogeneic orthotopic transplantation. The normal neural stem cell surface antigen CD15 enriches for the in vitro clonogenic and in vivo tumorigenic potential from uncultured medulloblastomas supporting the existence of a cancer stem cell hierarchy in this clinically relevant mouse model of cancer. Interestingly, the cell that drives the growth of these tumors has a stem cell phenotype, lending further support to a potential cell of origin of these tumors from normal cerebellar neural stem cells, or acquisition of a stem cell phenotype as part of the neoplastic transformation process from lineage restricted progenitors. I will present these recent data along with my current perspectives on cancer stem cell research, particularly the challenges the field faces moving forward.
Up to now there still has been no therapeutic standard for the heterogenic groups of low-grade gliomas (LGG). Our tumor data bank demonstrates an incidence of 5–10% LGG among averaged 180 primary surgically treated brain tumors per year with an increasing tendency. In this manuscript we describe our operative procedure and its results on patients with symptomatic LGG in critical brain area.
The analysis revealed 30 patients with LGG who were operated on in the last 3 years. Most patients had diffuse astrocytoma WHO grade 2 (15 patients). Rare tumors such as ganglioglioma (4), central neurocytoma (2), and sub-ependymoma (2) were also diagnosed during this period. The tumor was mostly localized perisylvian (15) less in ventricle (5), brainstem (5), central region (4), or basal ganglia (1). All patients were operated on with navigation support, endoscopic- microsurgical technique and intraoperative neurophysiologic monitoring (IOM) of different modalities. Since 2007 the morphologic data ha been fused with the functional one (fMRI, DTI, and PET) for navigation setting.
Total excision could be achieved in 17 patients. Subtotal (11) or part resection (2) has had to be performed due to the functional data and the IOM results. Temporary worsening of the neurological finding occurred in 18 patients for several days. The examination 3 months after surgery and later on demonstrates, however, an improvement status compared to that before surgery in 25 patients. During this period no recurrence surgery was needed. The seizures improved in 10/13 patients with or without antiepileptic drugs but in decreased dosage.
Our results show that LGG in critical area can be operated on safely with good outcome thus improvement of the quality of life. This therapeutic option using modern morphologic and functional data should be offered to patients with symptomatic LGG.
MicroRNAs (miRs) are recently described small non-protein-coding RNA molecules that are involved in the modulation of protein synthesis, via binding to complementary mRNA. They are also known to modify the activity of oncogenes and tumor suppressor genes of tumors, such as glioblastoma multiforme (GBM), the most common and deadly form of brain cancer. Using microarray analysis and quantitative RT-PCR, we have noted that miR-128, which is normally abundant in normal brain tissue, is significantly underexpressed in GBM tumor samples (18.75-fold reduction). Ectopic expression of miR-128 using oligonucleotide precursor or lentiviral vector reduced glioma cells growth considerably—both in vitro (U87, U251) in vivo (flank), as well as in models that enrich for glioma “stem-like” cells. Direct effects on several target genes functionally linked to regulation of oncogenic growth related processes, as well as impact exerted on stem cell-like self-renewal will be discussed. Another characteristic of gliomas is invasion into normal tissues. miR-451 was noted to have the most profound change in expression during the invasion process (relative signal strength drop from 2.753 to 0.000) when we compared invading versus noninvading glioma cells. This downregulation during invasion was corroborated in a number of glioma cell lines. In overexpression experiments, cells treated with pre-miR-451 or stably expressing pri-miR-451 demonstrated significantly reduced migratory behavior during invasion assay experiments. Potential targets of miR-451, involved in migration/invasion processes, will be discussed and validated. In conclusion, miRs are important modulators of GBM proliferation and invasion, and have the potential to be used as diagnostic or therapeutic agents.
Advances in understanding the genetic basis of cancer have led to the identification of molecules that are specifically expressed or mutated in tumors. Such molecules represent targets for therapy whose targeting might be expected to have increased selectivity for tumor cells compared to normal tissue with resultant enhancements in effect and reductions in the nonspecific toxicities that lead to chemotherapeutic side effects. In order to provide a large-scale comprehensive catalogue of such targets, the National Institutes of Health have initiated The Cancer Genome Atlas project and the first cancer to be subjected to this multi-center analysis is glioblastoma multiforme. The purpose of this lecture is to review the progress this effort has made.
The purpose of the TCGA is to catalogue and discover major cancer-causing genome alterations in large cohorts of human tumors through integrated multidimensional analyses. To this end, a committee screened retrospective biospecimen repositories for possible inclusion and each of these was then reviewed at the Biospecimen Core Resource. This resulted in approximately 35% of candidates being included in the initial studies (n=206). These samples were considered to be of sufficient quality and size to be entered into the analysis stage. Various parameters were then examined: DNA copy number by high-resolution comparative genomic hybridization, gene and microRNA expression was determined on several platforms, loss of heterozygosity was determined using single nucleotide polymorphism arrays, cancer-specific DNA methylation in CpG islands and promoters of 2,305 genes was determined using Golden Gate technologies, and 91 matched normal-tumor pairs were subjected to Sanger sequencing of 601 selected genes. All of this information was then entered into a large database for statistical analysis in the Data Coordinating Center (http://cancergenome.nih.gov/) that has various levels of public accessibility. The major findings from this effort are (1) it is possible to perform a comprehensive, multiinstitution, multiinvestigator effort and the structure to accomplish this has now been established; (2) genes that were previously found to be targets of the oncogenic process such as EGFR, PDGFRA, MDM2, CDK4/6, CDKN2A/B, PTEN, RAS, and AKT were found to have significant levels of alterations in this larger data set; (3) new potential targets such as NF1, ERBB2, MET, and PIK3R1 were discovered; (4) an interaction between MGMT promoter methylation and a hypermutator phenotype was uncovered with clinical significance for the application of alkylating therapeutics.
These results demonstrate the value of a large comprehensive genetic analysis of tumors, particularly glioblastoma multiforme. The alterations that were uncovered in this way suggest a limited number of signaling/metabolic pathways that are likely to be defective in tumors and furthermore that targeting pathways rather than specific molecules may be a reasonable approach in future.
The common practice of care for anaplastic astrocytoma (AA) is to use postoperative radiotherapy alone or to embark on regimens developed for the treatment of malignant gliomas in general. Can and should the increase in survival gained by radiochemotherapy with temozolomide (TMZ) in glioblastoma be extrapolated to patients with AA? Taking the evidence on nitrosourea-based therapy from the glioma meta-analysis, it would seem reasonable to suggest that a similar survival benefit might be seen amongst AA patients. However, there are concerns regarding late neurotoxicity related to the administration of concurrent radiochemotherapy, especially for patients who are living long enough to experience these unwanted sequelae. In the early 1990s, physicians began treating most AO patients with procarbazine, lomustine (CCNU), and vincristine (PCV) chemotherapy not only at recurrence but also at diagnosis in combination with or without radiotherapy, despite the lack of convincing data supporting this practice. In the United States, the prospective randomized phase III Radiation Therapy Oncology Group (RTOG) study (R9402) compared radiotherapy with and without preceding PCV. The contemporaneous study of the European Organization for Research and Treatment of Cancer (EORTC) (26951) used the same comparators but completed radiotherapy before commencing PCV. Both trials showed increased progression-free survival (PFS) but not overall survival (OS) in patients receiving radiochemotherapy, gained at the cost of relevant toxicity with the PCV regimen. Deletion of 1p and 19q may predict a superior outcome in AO and anaplastic oligoastrocytoma (AOA) in response to genotoxic treatments but not in patients with low-grade oligodendroglial tumors treated with surgery alone. To date, it has still remained unclear whether the combined 1p/19q deletions simply represent a molecular signature in anaplastic gliomas, which reflects a favorable natural biological behavior, or whether these markers are mechanistically related to response to therapy. The NOA-04 phase III, multicenter, open-label trial compared the efficacy and safety of radiotherapy versus chemotherapy (PCV or temozolomide) in 318 patients with newly diagnosed, supratentorial anaplastic gliomas of WHO grade III. In addition, the clinical significance of 1p/19q deletion and MGMT promoter methylation in these tumors was assessed in a translational study. Median time to treatment failure (TTF), PFS, and OS were not different between arms. Patients with an astrocytic tumor had a worse TTF. Oligoastrocytic tumors share the same favorable risks of pure oligodendroglioma. Combined 1p/19q deletion and hypermethylation of the MGMT gene promoter provided a large risk reduction for TTF and PFS irrespective of histology and treatment. In conclusion, there is an identical clinical course of anaplastic oligodendroglioma and oligoastrocytoma. The favorable impact of an oligodendroglial component was as strong as detecting combined 1p/19q deletion in the tumor tissue. MGMT promoter methylation was associated with prolonged PFS also in the radio-therapy arm A. Initial therapy in all anaplastic gliomas could possibly be either temozolomide or radiotherapy. The next challenges include intensifying therapy for anaplastic gliomas with unfavorable prognosis (e.g., no MGMT methylation, no combined 1p/19q deletion) and optimizing treatment for anaplastic gliomas with favorable prognosis without increasing toxicity. Trials basically testing radiochemotherapy with temozolomide separate for these entities are ongoing. Further, it will be important to test novel agents also in anaplastic gliomas, especially if influence on differentiation, angiogenesis or invasiveness can be postulated with a potential of influencing malignant progression.
MR imaging is an important technology that is of interest for diagnosing, planning focal therapy, evaluating treatment effects and assessing prognosis for patients with glioma. Despite the widespread application of conventional anatomic imaging techniques, there is a critical need for more advanced methods that are able to characterize biological properties of the lesion and surrounding normal brain. This is particularly important for understanding the effects of combination therapies, which influence proliferation and angiogenesis, as opposed to having a direct effect on cell viability. A number of advanced in vivo MR imaging methodologies are being evaluated in the clinic to see which combination of parameters are the most relevant for evaluating patients with gliomas. Anatomic Parameters: In addition to optimizing the acquisition times and resolution of anatomic images with standard contrast mechanisms, high-resolution T2*weighted magnitude and phase images are being acquired with 3-T and 7-T whole-body scanners. These images visualize heterogeneity in the region of T2 hyperintensity caused by local changes in susceptibility due to hemorrhage and other treatment effects. Using either the phase image alone or by the applying postprocessing methods to generate susceptibility weighted data it is possible to visualize small veins and hence examine differences between the tumor and surrounding tissue. Vascular Parameters: To further measures changes in vascular properties, data acquisition and analysis techniques have been developed to measure arterial spin labeling, dynamic contrast enhanced (DCE) and perfusion weighted (PW) images. The latter two methods utilize changes in T1-weighted or T2*-weighted images during and after an injection of gadolinium to estimate parametric maps of with vascular density and permeability. This is particularly important for the evaluation and quantification of measures of response for patients receiving antiangiogenic therapies. Structural Parameters: Degradation in the normal architecture of the brain can be evaluated using Diffusion Tensor Imaging. This gives information about the change in the magnitude of the apparent diffusion coefficient of water (ADC), as well as utilizing the directional properties of the diffusion tensor to mapping of connectivity using tractography. The ADC measurements have been reported as being linked to tumor cellularity and as being an early predictor of treatment effectiveness. The tractography is being used routinely for preoperative analysis of the disruption in normal tissue structure caused by the tumor and for surgical planning. Metabolic Parameters: The use of MR spectroscopy methods has been a major focus of our brain tumor research. 1H spectroscopy has been shown to highlight areas of metabolically abnormal, nonenhancing tumor that are important for assessing disease burden, treatment planning and assessing treatment response. Of particular interest are the spatial extent of areas where both the spectral peak corresponding to choline containing compounds and the peaks from lactate and/or lipid are elevated. Quantitative measures of these regions have been shown to be predictive of poor outcome in patients with GBM. More recent studies using very rapid spectroscopy acquisitions following the injection of hyperpolarized C-13 agents have shown promise for monitoring metabolically active lactate.
The molecular events that lead to the cancer-initiating cell involve critical mutations in genes regulating normal cell growth and differentiation. Cancer stem cells, or cancer initiating cells, have been described in the context of acute myeloid leukemia, breast, brain, bone, lung, melanoma, and prostate. Yet, there is emerging evidence that tumor initiation may occur also from tumor cells that are not defined as cancer stem cells. The origin of the cancer-initiating cells and their mechanisms of progression are not clear. At present there are two main models of cancer development: (1) the hierarchical model that suggests that tumors are generated and maintained by a small subset of undifferentiated cancer stem cells that are able to self renew and differentiate into the bulk tumor population, and (2) the stochastic model implying that tumor progression is the result of acquired genetic variability within the original clone allowing sequential selection of more aggressive subclones. The strongest evidence for a hierarchical model comes from studies of leukemia. For other tumors, however, the evidence is less clear. It has often been a problem to identify a clear-cut marker that defines the cancer stem cell. For brain tumors, recent evidence show that also CD133 cells can be tumorigenic indicating that CD133 is not a unique marker for cancer stem cells in the CNS. Therefore, it is not clear if brain tumor development follows a hierarchical or stochastic model of progression. What is clear is that within tumors, there are tumor cells that express and behave as stem cells. These cells can, as normal stem cells, adapt to new microenvironments and may well represent the most resistant cells to therapy. Based on their adaptability and differentiation capacities they can most likely not be defined by a specific marker. To what extent the environment and epigenetic factors influence the cancer stem-like cells deserves further studies. In conclusion, increased knowledge of developmental aspects in relation to self-renewal and differentiation, both under normal and deregulated conditions, will probably shed more light on the mechanisms that lead to tumor initiation and progression.
About 200,000 primary brain tumors are diagnosed worldwide each year, 70% in low-income countries. The cure rates for several cancers of both children and adults have improved markedly in high-income countries in recent decades. However, substantial improvements in the cure of patients with primary brain tumors remain an elusive goal even in high-income countries, except for those with “pediatric” brain tumors such as medulloblastomas, ependymomas, pilocytic astrocytomas, and germ cell tumors. Quality of life has improved for many brain tumor patients in high-income countries, even in the absence of improved cure rates, through introduction of advanced imaging, neurosurgical and more focused radiotherapeutic technologies. Concerted efforts in recent years in several low-income countries throughout the world, mainly focused on children with leukemias and lymphomas, have produced remarkably speedy and dramatic improvements in cure rates. However, such tumors do not require neurosurgical, imaging and radiotherapeutic skills and technology to improve outcome. No single approach can be employed in tackling the problem of patients with brain tumors in low-income countries. First, there is tremendous heterogeneity of resources amongst low-income countries: one cannot compare Paraguay with Brazil, the Congo with the Republic of South Africa, or Cambodia with Thailand. Any and all efforts must be placed within the context of the major prevailing health issues of the particular country: malnutrition, sanitation and water supply, infectious disease (including but not confined to HIV-AIDS), immunization. Efforts must recognize that in the majority of low-income countries, some 35–60% of the population will be less than 21 years of age—necessitating a substantial “pediatric” focus. Finally, for several countries, one cannot even begin to consider definitive treatment of brain cancer when there is a total absence of potent analgesics, anticonvulsants or even dexamethasone. One cannot address improving neurosurgical skills in so many countries who have fewer than one neurosurgeon for 12 million people (70% of African countries, at least) when WHO recommends 1 per 100,000 people. One cannot address radiotherapeutic techniques when 90% of the African continent has less than one machine for 30 million people and where expertise lags even further behind availability of equipment.
(a) The “big picture”—creation of a global patchwork quilt: triaging of efforts, avoidance of duplication, maximizing economy of effort, serving to connect interested/enthusiastic individuals/programs with emerging countries’ programs. (b) The details—palliative care/quality of life strategies for malignant glial tumors; neurosurgical/neurooncological training in and establishment of regional centers; establishment of satellite centers affiliated with regional centers to facilitate decentralization of supportive and acute care; development and adaptation of radiotherapy avoiding treatment strategies for “pediatric” brain cancers. The WFNO Mini-Symposium on Developing Strategies for Brain Tumor management will endeavor to address several aspects of this “solution” already in operation in different parts of the world.
An endothelial cell-specific growth factor VEGF (VEGF-A) is deeply involved in physiological as well as most of the pathological angiogenesis detected in cancer and arthritis. VEGF-A binds and activates two tyrosine kinase (TK) receptors, VEGFR-1 (Flt-1) and -2 (KDR/Flk-1). VEGFR-2 has a 10-fold higher kinase activity compared to that of VEGFR-1, whereas VEGFR-1 has a strong affinity to VEGF-A. The major signaling pathway from VEGFR-2 has recently been clarified. However, the possible role of VEGFR-1 and its signaling are not fully understood yet.
A unique signaling of VEGFR-2 for angiogenesis. First, I will briefly summarize the signaling of VEGFR-2, the major positive signal transducer for angiogenesis. We found that VEGFR-2 activates a unique signaling, PLCγ-PKC (C-kinase)-Raf-MAP kinase pathway for DNA synthesis in endothelial cells. Ras-activation was only minor for this downstream signaling from VEGFR-2. Also, we have shown that a single autophosphorylation site 1175PY (phosphotyrosine, 1173 position in mice) is critical for this pathway. 1173F/F mutant mice died at E8.5–E9.0 with a severe defect on vasculogenesis, indicating that this pathway is important for angiogenesis in vivo, and a good target for suppression of tumor angiogenesis.
An important role of VEGFR-1 in tumor growth and metastasis. VEGFR-1 gene knockout mice are embryonic lethal due to an overgrowth of blood vessels at E8.5, suggesting a negative role of VEGFR-1 in early embryogenesis. We showed that VEGFR-1 ligand-binding domain (VEGFR-1 TK−/−) rescues this lethality. Since VEGFR-1 TK−/− mice develop essentially normal angiogenesis, these mice are useful to study the role of VEGFR-1 signals at adult stages in disease models such as cancer. We found that in various tumors, tumor growth and angiogenesis are significantly lower in VEGFR-1 TK−/− mice compared with that in the wild-type mice mostly due to a defect of macrophage recruitment from bone marrow. In collaboration with us, Kerber et al. also showed a slower growth of gliomas in VEGFR-1 TK−/− mice. VEGFR-1 is expressed not only in vascular endothelial cells but in macrophages, suggesting that VEGFR-1 plays an important role in the promotion of tumors including glioma via activation of macrophage-lineage cells. Finally I will discuss on a possible increase in tumor aggressiveness after low-oxygen-nutrition stress.
Glioblastoma, the most common malignant brain tumor, is amongst the most lethal and difficult cancers to treat. Although EGFR-mutations are frequent in glioblastoma, their clinical relevance is poorly understood. Unexpectedly, studies of tumors from patients treated with the EGFR-inhibitor lapatinib revealed that EGFR induces the cleavage and nuclear translocation of the master transcriptional regulator of fatty acid synthesis, SREBP1. We show that this response is mediated by Akt, but surprisingly, clinical data from rapamycin-treated patients shows that SREBP1 activation is independent of the mTORC1, potentially explaining its poor efficacy in the treatment of such tumors. The importance of these observations is underscored by the finding that non-EGFR-activated glioblastomas are resistant to inhibition of fatty acid synthesis whereas introduction of an activated EGFR-allele sensitizes tumor xenografts in mice to cell death, which is augmented by the HMG-CoA reductase inhibitor atorvastatin. These results identify a novel EGFR-mediated prosurvival metabolic pathway, suggesting a promising approach for treating EGFR-activated glioblastomas.
To improve the survival and quality of life for young children with newly diagnosed primary malignant CNS embryonal tumors through a strategy of relatively brief and intensive chemotherapy (<5 months duration), followed by myeloablative chemotherapy in an effort to avoid CNS irradiation.
Since 1991 three sequential prospective nonrandomized multicenter feasibility studies have been conducted. In the first study, “Head Start” I (1991–1997) eligible children received a single treatment strategy: Induction chemotherapy of vincristine, cisplatin, cyclophosphamide, and etoposide administered every 21–28 days for five cycles (regimen A). This was followed by a single cycle of myeloablative chemotherapy (thiotepa, carboplatin, and etoposide) and rescue with autologous hematopoietic progenitor cells. In the subsequent “Head Start” II (1997–2003) study, patients with localized medulloblastoma/PNET received the “Head Start” I induction regimen A, while patients with AT/RT or metastatic tumor received a modified induction regimen A2 intensified with high-dose systemic methotrexate (HD-MTX, 400 mg/kg). In the current “Head Start” III study (2003 to present), all patients receive the same new Induction Regimen D, which includes three cycles of regimen A2 (cycles 1, 3, and 5), alternating with two new cycles (cycles 2 and 4), in which HD-MTX and cisplatin are deleted, and replaced by oral etoposide and oral temozolomide.
In children with medulloblastoma, gross total resection of localized primary tumor, absence of dissemination and desmoplastic histology conferred highly favorable irradiation-free 5-year overall survivals (OS) of 73% (classical) and 88% (desmoplastic). The use of HD-MTX in the “Head Start” II protocol produced markedly improved 5-year EFS (45%) and OS (53%) for children with disseminated medulloblastoma, exceeding even the 5-year EFS of 29% for children with localized but incompletely resected medulloblastoma who did not receive HD-MTX. For this reason, the “Head Start” III regimen incorporates HD-MTX for all children. For children with supratentorial PNET (sPNET), the only statistically significant prognostic factor was location; children with pineal region PNET (pineoblastoma) experienced poorer outcome than those with nonpineal sPNET (EFS of 15% vs. 44% and OS of 25% vs. 62%). The small group of patients with brainstem PNET represents the first reported cures (two of six patients) of this rare entity. For children with AT/RT, the addition of HD-MTX in “Head Start” II resulted in an overall significant improvement in 5-year EFS (43%) and OS (54%), but only for patients with localized disease at diagnosis. Prospective longitudinal follow-up, including formal quality of life, neuropsychological, endocrine, and audiometric assessments of the surviving patients is ongoing, and has now been completed for “Head Start” I and II. Neuropsychological studies document preserved quality of life and intellectual functioning for those children who avoided irradiation, including children who received HD-MTX. Hearing loss continues to be the major long-term morbidity of treatment, and seems magnified by the use of ototoxic antibiotics during treatment. Endocrine studies have identified only rare isolated examples of growth or thyroid dysfunction requiring hormonal replacement therapy. Second malignancies have not been identified in patients receiving only protocol chemotherapy without CNS irradiation.
The “Head Start” protocols have demonstrated significant improvements in both the survival and quality of life of young children with CNS embryonal tumors, compared with previously published treatment regimens.
Low-grade gliomas (LGG) account for the largest group of brain tumors in the pediatric age, representing approximately one-third of all childhood brain tumors. These tumors can arise in any part of the nervous system. Subtypes of LGG are distinguished by their histological features, but all LGGs are classified as World Health Organization (WHO) grade I or II. LGG are usually slow growing although for reasons that are not completely clear, some of these tumors may show aggressive behavior and/or have erratic growth rates, particularly in young children with hypothalamic/chiasmatic gliomas. Surgery is the treatment of choice of resectable tumors and is often curative in these instances. However, a large number of pediatric LGG arise in areas where the role of surgery is limited, due to the risk of morbidity associated with attempts at aggressive resection. Radiation therapy to such tumors has demonstrated evidence of benefit, with objective radiographic responses often for long periods of time, although the use of radiation therapy in young children results in serious long-term sequelae, including permanent neurocognitive and neuroendocrine deficits, as well as an increased risk of vasculopathy and second malignancy. It is now clear that chemotherapy can delay the need for radiation therapy. Over the past decade, the upfront use of chemotherapy has become standard therapy for children and infants with unresectable tumors. Various regimens have been tested, both in chemotherapy naïve patients and in patient with recurrence after previous radiation and/or chemotherapy. Carboplatin is one of the most effective agents against low-grade gliomas. The regimen of weekly carboplatin and vincristine has been extensively piloted and is considered by many to be the standard chemotherapeutic regimen for young children with unresectable low-grade gliomas. Other regimens have been suggested, and the choice between different options is often a challenge for pediatric oncologists. Beside the efficacy, the decision should take into account various factors such as short- and long-term toxicity, the route and schedule of administration and the cost. Recent prospective studies have shown that a majority of children (up to 70%) will require additional treatments after a first line of chemotherapy. The management of these children who do not respond to upfront chemotherapy or show progression after a first line of treatment is challenging. Traditionally, radiation has been offered as a “salvage treatment,” regardless of the age of the patient. Increasingly, chemotherapy is being been considered in this context, particularly when young age predicts significant neurointellectual deficits with the use of radiation. Several studies are ongoing, looking at new agents or combinations in children with recurrent LGG. How many lines of chemotherapy can be used without compromising long-term outcome is still unclear. It appears, however, that the use of multiple lines of chemotherapy is possible and that the need for radiation can be reduced to a small proportion of children. Recently, biological studies have suggested that it might be possible to predict the behavior of pediatric LGG by evaluating the proliferation index (mib-1 or Ki-67), microvessel density, or telomere length. Evaluation of these parameters in prospective protocols is ongoing and the results of these studies may potentially influence the chemotherapeutic management of patients with low-risk and high-risk features.
Survival rates for those diagnosed with brain tumors have generally improved with advances in surgery, radiation therapy, and chemotherapy. As a result, understanding the long-term impact of therapy including the neurocognitive and psychological effects of cancer treatment are important. Most brain tumor survivors who achieve durable survival often present with various deficits. Significant sequelae, including intellectual decline, as well as variable attention-concentration, slow processing speed, and motor deficits, have been observed as a result of treatment, with specific risk factors including young age at the time of treatment, higher dose and larger brain volume involved in radiotherapy, and longer follow-up. Cranial irradiation has been associated with white matter changes, particularly in the frontal cortex. Consequently, a young child treated for medulloblastoma with craniospinal radiation will likely experience more adverse late effects compared with an adolescent treated for a CNS germinoma with focal radiotherapy. In an effort to preserve intellectual functioning and quality of life, many current treatment studies are utilizing reduced doses and volumes of radiotherapy, while others are employing postoperative chemotherapy to delay or avoid radiation therapy. Despite “benign” histology and lower toxicity of treatments for low-grade tumors that are commonly limited to maximal surgical resection, these tumors are associated with increased risk for late effects in both cerebellar and supratentorial locations given their slow-growing and infiltrative presentation. In contrast, high-grade gliomas are a group of CNS tumors characterized by heterogeneity and a typically poor treatment response to surgery, chemotherapy, and radiotherapy. Consequently, newly emerging treatment regimens often involve significant toxicities that necessitate the sequential measurement of quality of life by both patient and caregiver report as a primary outcome in conjunction with length of survival. Two prominent pediatric multicenter intervention studies in the United States involved the use of methylphenidate (Ritalin) medication, and the other utilized a Cognitive Remediation Training Program in order to address symptoms such as attention, memory, and academic achievement difficulties following treatment for pediatric cancer, and the results have demonstrated modest effect sizes. Similarly, recent adult studies have reported encouraging pilot data in response to patients using substances such as donepezil (Aricept), modafinil (Provigil), and Ginkgo biloba. Emotional effects from the treatment of a brain tumor vary depending on the patient, diagnosis, location of disease, treatment type, family dynamics, and preexisting psychological conditions. Some studies indicate that patients with brain tumor survivors are at no greater risk for long-term emotional sequelae than would normally be expected in a healthy population with lower levels of psychological distress, better psychological health, lower levels of aggressiveness, antisocial behavior, and substance abuse than case controls. However, other survivor studies have identified social difficulties in withdrawal, isolation, and poor social competence attributed to physical appearance, functional impairments, and reduced time with peers. Thus, while information about brain tumor patients’ social-emotional and behavioral functioning continues to emerge, there does appear to be an increased risk for deficit in certain domains of physical and psychosocial functioning. Consequently, there is currently a change under way in research orientation from cancer treatment as a necessarily pathology-inducing experience but rather focusing upon the skills and resilience of patients and families who were struggling with a difficult situation. As such, it is important to identify risk factors in patients and families in order to tailor interventions to minimize the psychological sequelae in brain tumor survivors. In sum, neuropsychological, social-emotional, behavioral, and quality-of-life evaluations and interventions provide a range of benefits to the field of neurooncology including the assessment of functioning prior to treatment, as well as measuring the neurocognitive impact between treatment regimens and facilitating the psychological transition through treatment and beyond into survivorship.
Chemotherapy, with or without radiotherapy, is the mainstay of treatment for primary CNS lymphoma (PCNSL). High-dose methotrexate (MTX) is the most effective drug, and is used in doses of 1–8 g/m2 either as a single agent or in combination with other drugs such as corticosteroids, cytarabine, procarbazine, vincristine, carmustine, lomustine, thiotepa, cyclophosphamide, temozolomide, and rituximab. To date, an overwhelming number of different regimens utilizing high-dose MTX have been reported. However, given the lack of randomized trials, the optimal treatment remains controversial. Varying methodology makes the comparison of available studies extremely difficult; however, some common themes can be found throughout literature. Treatment paradigms vary considerably according to the patient’s age. Most studies support the use of chemotherapy-only treatments for elderly patients (>60 years), given the high risks of neurotoxicity associated with radiotherapy. However, the prognosis remains poor independent of the chemotherapy chosen, and less toxic regimens might be preferable for such elderly patients. Conversely, the goal of treatment in younger patients (<60 years) is a cure, and a more aggressive approach seems warranted. Growing evidence suggests that commonly used chemotherapy-only regimens for younger PCNSL patients are associated with increased relapse rates that may not justify deferral of radiotherapy. Thus, a significant focus of research has been the development of intensified chemotherapy regimens that could replace radiotherapy. Several ongoing studies are evaluating the role of high-dose chemotherapy with stem cell rescue utilizing drugs that penetrate the blood-brain barrier. An important feature of these studies is the inclusion of thorough neuropsychological evaluation in order to evaluate the long-term cognitive outcomes. Another important venue of investigation is trying to identify molecular predictors of aggressiveness, as well as markers of biological predisposition for neurotoxicity, which would allow more judicious therapeutic decisions.
NG2 expressing cells (NG2+ cells) are the largest group of neural cycling precursors in the adult brain and exhibit wide anatomical distribution and proliferative activity. Based on these characteristics, we hypothesized that NG2+ve cells are potential candidates as cancer stem cells (CSCs) that maintain the glioblastoma (GBM) growth.
GBM tumor-initiating cell (TIC) lines were derived from fresh clinical samples under serum-free (SF) conditions according to our Cambridge Protocol (Fael Al-Mayhani et al., 2009). Cells were analyzed and sorted using FACS. Comparative studies on NG2+ and NG2 cells were conducted in vitro and in vivo. Molecular studies were performed using microarray and comparative genomic hybridization (CGH).
Microarray data showed that genes associated with neural precursors (NG2, Olig2, PDGFRa, and NKX2.2) were expressed by all GBM TIC lines and the majority (90%) of GBM tumors tested (n=147). In vitro data show that NG2+ cells isolated from TIC lines exhibit high level of growth, proliferation and clonogenic potentiality compared to NG2 cells. Only NG2+ cells were able to form tumors in vivo. Array CGH demonstrate that both NG2 populations have typical GBM cytogenetic profile (e.g., Ch7 gain and Ch10 loss), however, NG2+ cells exhibit significant additional abnormalities.
Genes associated with neural precursors are widely expressed by GBM tumors and GBM TIC lines. NG2+ cells isolated from TIC lines appear to conserve the precursor status of their normal counterparts and have tumorigenic characteristics in vivo. In addition, NG2+ cells exhibit significant structural chromosomal abnormalities that differ from NG2 cells and among different samples reflecting GBM heterogeneity. These data indicate that NG2 can be used as CSCs marker in GBM and draw attention to possible clinical implications.
Mounting experimental evidence suggests that malignant brain tumors contain and may arise from cancer stem cells (CSCs). CSCs are believed to play important roles in tumorigenesis and recurrence. CD133, a pentaspan transmembrane glycoprotein, is a putative marker of glioma-derived CSCs based on the ability of CD133-positive cells to recapitulate the tumor in animal models. Glioma tumor tissues are known to be populated by a heterogeneous mix of cells with distinct genotypes and phenotypes, and different proliferative capacities. The cellular origin of such karyotypic heterogeneity is unclear. Whether the observed population diversity within a tumor has mono- or polyclonal origin has important implications for cancer progression.
To investigate the cellular origin of population diversity, we established 23 subclones from a single glioma CSCs positive for CD133. These subclones were subsequently propagated and analyzed.
The self-renewal and proliferative capacities of the subclones were highly distinct. Sensitivity to chemotherapeutic agents also differed widely. FACS analysis for expression of stem and differentiation markers (e.g., CD133, A2B5, CD44, CXCR4, and CD24) showed that each subclone was composed of a distinct population of cells. Immunoblot showed different expression levels of cell cycle regulation proteins such as Bmi-1, p53, p27, and p21. CD133-negative cells failed to establish subclones.
Our observations suggest that a single CD133-positive tumorigenic CSC is capable of producing phenotypically heterogeneous self-renewing progeny in an in vitro setting. Whether such observed diversity is regulated by genomic alteration or gene expression remains to be seen.
Tumors are thought to contain a small population of cells, designated cancer stem cells (CSCs), that possess the ability both to self-renew and to differentiate and that play a key role in the initiation and maintenance of cancer. However, whether CSCs arise from multipotent immature cells or differentiating cells has remained unclear. Moreover, the minimal repertoire of genetic changes required for the generation of CSCs is unknown. The establishment of CSCs from cells of various differentiation stages by genetic manipulation might be expected to provide insight into these questions. We have now established mouse malignant brain tumor models by overexpressing c-MYC or RasV12 in neural stem cells/multipotent progenitor cells derived from subventricular zone of mouse with a homozygous deletion of the Ink4a/Arf locus (Ink4aKO mice). Tumors induced by c-MYC are histologically different from those induced by RasV12, suggesting that pathological phenotype of tumor is dependent on the gene(s) responsible for tumorigenesis. Tumor induced by RasV12 is highly heterogeneous and invasive, and a fraction of the induced-tumors cells was found to exhibit properties of CSCs. Invading cells at the tumor margin expressed high levels of CD44, which plays a role in both cell migration in extracellular matrix and stemness of cancer cells. Furthermore, tumor cells formed cuffs around the preexisting microvessels, and those invading the area contained numerous microglial cells. Our genetically induced cancer systems provide ideal models for the characterization of CSCs and the development of novel therapeutic approaches to malignant brain tumors.
It has been reported recently that endothelial cells function as a stem cell niche to promote CD133-positive cancer stem cells (CSCs) self-renewal in glioblastoma (GBM). However, the mechanism that endothelial cells function as a CSC niche is largely unknown. It has been shown that endothelial cells promote neural stem cell self-renewal by providing Notch signaling and that Notch ligands are expressed in GBMs. To test if endothelial cells and differentiated tumor cells could function as a CSC niche by providing Notch ligands to CSCs, first we cocultured GBM neurospheres with primary human endothelial cells (PHECs) which express high levels of Notch ligands Jagged/Dalta. We found the PHECs promoted GBM neurosphere growth in vitro. Then, we found the CD133-negative population expressed a higher level of Jagged/Dalta and glial differentiation marker GFAP, compared with the CD133-positive population. When GBM neurospheres forced differentiation to grow as a monolayer culture, the CD133-positive population and CD133 expression were reduced, whereas expression of GFAP (glia marker), GalC (oligodendrocyte marker), and Notch ligands were significantly induced. Furthermore, treatment of GBM neurospheres with a Jagged peptide increases GBM neurosphere propagation in a dosage-dependent pattern, indicating that Notch ligands expressed in both differentiated tumor cells and endothelial cells can function as a niche to promote GBM CSC self-renewal in vitro. Finally, we found that Notch ligands were highly expressed in the blood vessels of GBM primary tumor and intracranial xenograft, and that CSCs accumulated around the blood vessels within the tumor. In summary, our studies demonstrate that Notch activation is driven by juxtacrine signaling between tumor cells replicating the classical process of lateral inhibition and by stromal niche signals, suggesting that targeting both CSC and its niche may provide a profound strategy to deplete CSCs in GBM. This may also apply to other cancers.
C-MET, preferentially expressed in stem/progenitor cells, plays critical roles in the regulation of cell growth and motility during embryogenesis and throughout the life. Deregulated MET signaling has been implicated in various types of cancers including glioblastomas (GBMs). Increasing experimental evidences support the hypothesis that a subpopulation of tumors, namely, cancer stem cell/tumor-initiating cells (CSC/TICs), are responsible for tumor initiation/propagation and share many developmental pathways with their normal counterparts. However, the precise roles of MET signaling pathway operated in CSCs/TICs are largely unknown. Here we show evidence that some of Methigh GBM cells are indeed CSC/TICs. Methigh populations but not Metlow/− populations are highly tumorigenic and give rise to both Methigh and Metlow/− cells, thereby establishing tumor hierarchy. Methigh cell populations significantly overlap with cells positive for expression of CD133, a putative CSC/TIC marker, and these cells are highly resistant to radiation. Furthermore, we found that suppression of MET pathway in these GBM CSC/TICs achieved by siRNA or shRNA mediated knockdown of c-met gene results in significantly reduced cell survival in vitro and impaired tumorigenic potential in vivo, suggesting that C-Met activation is required to maintain cancer stem cell phenotype in these cells. Interestingly, suppression of MET activation also significantly reduces migration/invasion property of GBM TSC/TICs, in part by derepressing expression of E-cadherin. Taken together, our data demonstrate that MET activation in GBM is a functional requisite for “stemness” phenotypes in CSC/TICs. In summary, we provide experimental evidence supporting an essential role of Met signaling pathway in the survival and invasion of glioblastoma CSC/TICs. Not only will such insights pave the way for a more thorough understanding of CSC biology, but also they further implicate Met as a promising therapeutic target.
Gliomas are among the most frequent tumors prevalent in the adult and pediatric populations. Unfortunately, despite current therapies implemented for patients with high-grade gliomas, the prognosis is extremely poor. There is a current sparsity of in vitro preclinical models of gliomas that recapitulate the pathological and treatment responses reminiscent in human gliomas. Current in vitro models such as glioma stem cells are very problematic due to having low clonality and variable therapeutic responses. In an effort to explore alternative robust in vitro models of gliomas, we have developed the first nonneural stem cell-based model of gliomas from murine embryonic stem cells. Specifically, we have noted that synthetic astrocytes derived from murine embryonic stem cells are more representative in the origin of gliomas, even at high clonality, compared to terminally differentiated somatic astrocytes. Similar findings are also documented with synthetic astrocytes differentiated from murine neural stem cells. With the use of molecular transcriptome fingerprinting, these stem cell derived synthetic astrocytes are most similar (R-value=0.8, p-value<0.01) to murine somatic astrocytes from the embryonic cortex (E13.5) and newborn hippocampus (P4). Furthermore, these synthetic astrocytes have an intrinsic property of enhanced proliferation that may potentiate an ease toward transformation. We have noted that the genetic alterations of these synthetic astrocytes with candidate gliomagenesis genes including MDM2, AKT, and H-RAS can potentiate intracranial high-grade gliomas in Nod-Mice. Furthermore, we also undertook functional random mutagenesis screens on these synthetic astrocytes using gene trapping. We obtained a genome wide functional map of genes that are stringently correlated with a pathological spectrum of astrocytomas. This body of work augments the use of stem-cell-based preclinical models to decipher the genetic and pathological basis of gliomagenesis. Moreover, these models can also be utilized for therapeutic targeting applications.
Mesenchymal stem cells (MSCs) have been shown to migrate toward tumors, but their distribution pattern in glioma is not completely portrayed. The primary purpose of the study was to assay the tropism capacity of MSCs to glioma, to delineate the pattern of MSCs distribution in glioma after systematical injection, and to track the migration and incorporation of magnetically labeled MSCs with clinical 1.5-T MRI.
Fisher 344 rats MSCs were colabeled with superparamagnetic iron oxide nanoparticles (SPIO) and enhanced green fluorescence protein (EGFP). The tropism capacity of MSCs was quantitatively assayed by Transwell system in vitro. To in vivo track the migration of MSCs, MRI was performed at 7 days and 14 days after systematic administration of labeled MSCs. After scanning, the distribution patterns of MSCs in glioma burdened rats were examined by Prussian blue and fluorescence staining.
In vitro study showed MSCs possessed significantly greater migratory capacity than fibroblast cells, and lysis of F98 glioma/cultured F98 cells showed more capacity to induce migration of cells than other stimuli. At 7 days after MSCs transplantation, the SPIO/EGFP-colabeled cells distributed throughout the tumor, where a well-defined dark hypointense region represented on gradient echo (GE) sequences. While at 14 days, most of labeled MSCs were found at the border between tumor and normal parenchyma, which represented on GE sequence as diluted amorphous dark areas at the tumor’s edge.
This result demonstrated systemically transplanted MSCs home to glioma with high specificity with a temporal-spatial pattern, which can be tracked by MRI.
Treatment and cure of medulloblastoma continues to be disappointing. Approximately 30% of cases recur, and even when a cure is achieved, the treatment is highly morbid to the developing brain and results in long-term sequelae. Recent evidence suggests that many solid tumors are sustained by a population of cells with stem cell-like properties that have extensive capacity for self-renewal. Tumor stem cells (TSC) are also thought to have increased radio- and chemoresistance and may be responsible for treatment failures. Mechanisms of DNA damage response have not been explored previously in neural stem cells derived from medulloblastoma. Identifying novel therapeutic targets that spare normal neural stem cells while eradicating the tumor-derived stem cell population is a long-term goal. TSCs were isolated from medulloblastomas that spontaneously arose in mice haploinsufficient for Patched (Ptc), a component of the Sonic Hedgehog receptor, and cultured in serum-free media as free-floating spheres. Neural stem cells were derived from the early postnatal hippocampus and cerebellum of wild-type and Ptc± mice. Cultures were exposed to 2 Gy of ionizing radiation and cell cycle progression was characterized with flow cytometry. We found that normal neural stem cells had a similar phenotype with sustained accumulation in G1 within 2 h of irradiation. In contrast, we found that TSCs failed to arrest in G1 and showed a marked but unsustained G2/M arrest 8 h after radiation. By 48 h after irradiation TSCs had reentered the cell cycle. The level of apoptosis was unchanged compared to controls. We repeated these experiments at 10 Gy of radiation and in the TSCs found a large increase in apoptosis over time. Accumulation of TSCs in G2/M was again noted at 8 h, however, the G2/M arrest was sustained at 24 and 48 h after irradiation. Sequencing of the open reading frame demonstrated that p53 is wild type in TSCs, appropriately phosphorylated on serine 15, p21 is induced and Rb phosphorylated, suggesting that the abrogated G1 arrest is not due to abrogation or mutation of p53 pathway. We have characterized the cell-cycle and DNA damage response of murine NSCs and found that this population of undifferentiated cells undergoes cell cycle arrest in G1 after DNA damage. This is in contrast to embryonic stem cells that lack a G1 checkpoint, do not initiate cell cycle arrest and undergo apoptosis after exposure to ionizing radiation. Additionally, we found that TSCs have escaped the G1 checkpoint in a p53-independent manner and proceed to replicate their DNA without adequate repair that likely contributes to genomic instability and propagation of new therapy-resistant clones through aberrant phosphorylation of Retinoblastoma gene. These data suggest that TSC and normal neural stem cells have distinct responses to ionizing radiation and DNA damage. Studies are ongoing to determine the molecular basis for the distinct phenotype observed in response to radiation in normal and neoplastic stem cells and to specifically increase the sensitivity of TSCs to apoptosis after DNA damage.
Supported by NREF (A.B.F.), Sontag Foundation (C.W.), Waterman Foundation for Cancer Genetics (C.W.), Brain SPORE (C.W.), Fraternal Order of Eagles (C.W.).
Brainstem gliomas are common in children. Despite recent attempts to optimize or combine radiotherapy and chemotherapy, the poor prognosis of these tumors has not changed in the past two decades. This poor outcome is due to relatively inoperable location and infiltration into the brain. Precisely targeting these tumor satellites may prove critical for the success of any potential therapeutic strategy. Based on the selective killing of malignant cells by TRAIL and targeted migration toward tumor of MSCs, it is logical to hypothesize that membrane-spanned TRAIL-engineered MSCs may provide dual-targeted antitumor effects against brainstem glioma. We have demonstrated systemically transplanted MSCs immigrate to brainstem glioma with high specificity. MSCs can penetrate the vessels surrounding tumors, then stream as a chain pattern toward gliomas, and eventually circumscribe the tumor. Membrane-spanned TRAIL engineered MSCs not only express full-length TRAIL in surface, but secrete some sTRAIL in medium. After infected with rAAV-hTRAIL, hMSCs show no increase in apoptosis, which suggests it is feasible to engineer MSCs with TRAIL for glioma therapy. After cocultured hTRAIL engineered MSCs with U87MG cells, the apoptosis of U87MG cells significantly increase than sTRAIL treated U87MG cells. Systematic delivery of hTRAIL engineered MSCs into established human brainstem glioma xenografts results in the potent induction of apoptosis in gliomas, but not in normal brain. This confirms that hTRAIL engineered MSCs are dramatically effective at killing glioma cells, but no toxic to normal brain tissue.
Growing evidence suggests that the growth of brain tumors is maintained by cancer stem cells. Thus, normal neural precursor cells (NPCs) are attractive candidates for the origin of human gliomas.
To explore the relationship between gliomas and stem cell properties, by comparing the microRNA (miRNAs) expression signature of glial tumors, embryonic stem cells (ESCs), NPCs and normal adult brain tissues of both human and mouse.
We compared the expression profiles of 186 mature microRNAs in ESCs, NPCs, gliomas and normal brains. Our panel contained RNA samples derived from glial tumors of various grades, ESCs, NPCs, and normal adult brains from both humans and mice. The miRNA expression was quantified by RT real-time PCR.
Human gliomas, regardless of their grade, displayed a miRNA expression profile that was reminiscent of that in NPCs. Of special note is the 80% identity observed between the miRNA expression signature of human gliomas and the methylcholanthrene-induced mouse glioma, GL-261. Interestingly, about half of the miRNAs in this shared profile were clustered in seven genomic regions susceptible to genetic/epigenetic alternations in different types of cancers. These clusters comprised the miR-17 family of miRNA clusters, mir183–182, and the stem cell-specific clusters mir367–302 and mir371–373, which are upregulated in gliomas, ESCs and NPCs. We also identified the bipartite cluster of 7+46 miRNAs on chromosome 14q32.31, which was downregulated; therefore, it might represent the largest tumor suppressor miRNA cluster.
Our finding that all gliomas displayed an NPC-like microRNA signature might imply that normal NPCs that sustain cancer-promoting mutations are the cells of origin of gliomas. It might also suggest that cells within this differentiation window are particularly susceptible to gliomagenesis. Alternatively, a more differentiated progeny, that acquire aberrations in specific genomic regions, could have given rise to this characteristic “NPCs microRNA signature.”
The treatment of malignant brain tumors comprises surgery, radiotherapy and chemotherapy. Especially in childhood, patients with malignant brain tumors are enrolled in clinical studies. This allows uniform treatment protocols, comparable results and close follow-ups.
Children with malignant supratentorial tumors, for example, PNETs, glioblastomas, malignant ependymomas, plexus carcinomas, and rarely occurring tumors, for example, chondrosarcomas and anaplastic meningiomas, were admitted to our department. Surgical resection was performed in all cases. Whenever possible, even recurrent tumors or metastases were resected. In highly vascularized lesions, preoperative embolization was considered. Besides surgical resection, the children were all enrolled in clinical studies, being treated according to study protocols. All children survived the surgical procedures without major neurological deficits. Particularly in tumors of the ventricle (plexus carcinomas and a chondrosarcoma), intraoperative blood loss could be high. The surgical resection of unusually large neoplasms is possible in children without causing major neurological deficits. Especially at a very young age, the plasticity of the brain compensates the resection of large intraparenchymal tumor masses, not compromising the normal development of the children. Besides well-established radio- and chemotherapeutical protocols, multiple resections seem to be justified. They are well tolerated and help to prevent neurological deficits and prolong the survival time of the young patients.
Surgical resection is well tolerated by most of the infants and children, independent of age. The inclusion into clinical studies should be mandatory, leading to optimal radiotherapeutical and chemotherapeutical management and a close follow-up. Reoperations should be considered, whenever possible.
There is still no general consensus in the literature regarding the role of extent of glioma resection in improving patient outcome. Although the importance of resection in obtaining tissue diagnosis and alleviating symptoms is clear, a lack of class I evidence prevents similar certainty in assessing the influence of extent of resection. We reviewed every major clinical publication since 1990 on the role of extent of resection in glioma outcome. Twenty-eight high-grade glioma articles and 10 low-grade glioma articles were examined in terms of quality of evidence, expected extent of resection, and survival benefit. Despite persistent limitations in the quality of data, mounting evidence suggests that more extensive surgical resection is associated with longer life expectancy for both low- and high-grade gliomas. The prognostic role of extent of resection (EOR) of low-grade gliomas (LGGs) is a major controversy. We designed a retrospective study to assess the influence of EOR on long-term outcomes of LGGs. The study population (n=216) included adults undergoing initial resection of hemispheric LGG. Region-of-interest analysis was performed to measure tumor volumes based on fluid-attenuated inversion-recovery (FLAIR) imaging. Median preoperative and postoperative tumor volumes and EOR were 36.6 cm3 (range, 0.7–246.1 cm3), 3.7 cm3 (range, 0–197.8 cm3), and 88.0% (range, 5–100%), respectively. There was no operative mortality. New postoperative deficits were noted in 36 patients (17%); however, all but four had complete recovery. There were 34 deaths (16%; median follow-up, 4.4 years). Progression and malignant progression were identified in 95 (44%) and 44 (20%) cases, respectively. Patients with at least 90% EOR had 5- and 8-year overall survival (OS) rates of 97% and 91%, respectively, whereas patients with less than 90% EOR had 5- and 8-year OS rates of 76% and 60%, respectively. After adjusting each measure of tumor burden for age, KPS, tumor location, and tumor subtype, OS was predicted by EOR (hazard ratio [HR]=0.972; 95% CI, 0.960–0.983; p<0.001), log preoperative tumor volume (HR=4.442; 95% CI, 1.601–12.320; p=0.004), and postoperative tumor volume (HR=1.010; 95% CI, 1.001–1.019; p=0.03), progression-free survival was predicted by log preoperative tumor volume (HR=2.711; 95% CI, 1.590–4.623; p ≤0.001) and postoperative tumor volume (HR=1.007; 95% CI, 1.001–1.014; p=0.035), and malignant progression-free survival was predicted by EOR (HR=0.983; 95% CI, 0.972–0.995; p=0.005) and log preoperative tumor volume (HR=3.826; 95% CI, 1.632–8.969; p=0.002). Improved outcome among adult patients with hemispheric LGG is predicted by greater EOR. Language sites in the cortex of the brain vary among patients. Language mapping while the patient is awake is an intraoperative technique designed to minimize language deficits associated with brain-tumor resection. To study language function after brain-tumor resection with language mapping, we examined 250 consecutive patients with gliomas. Positive language sites (i.e., language regions in the cortex of the brain, 1×1 cm, which were temporarily inactivated by means of a bipolar electrode) were identified and categorized into cortical language maps. The tumors were resected up to 1 cm from the cortical areas where intraoperative stimulation produced a disturbance in language. Our resection strategy did not require identification of the stimulation-induced language sites within the field of exposure. A total of 145 of the 250 patients (58.0%) had at least one site with an intraoperative stimulation-induced speech arrest, 82 patients had anomia, and 23 patients had alexia. Overall, 3,094 of 3,281 cortical sites (94.3%) were not associated with stimulation-induced language deficits. A total of 159 patients (63.6%) had intact speech preoperatively. One week after surgery, baseline language function remained in 194 patients (77.6%), it worsened in 21 patients (8.4%), and 35 patients (14.0%) had new speech deficits. However, 6 months after surgery, only 4 of 243 surviving patients (1.6%) had a persistent language deficit. Cortical maps generated with intraoperative language data also showed surprising variability in language localization within the dominant hemisphere. Craniotomies tailored to limit cortical exposure, even without localization of positive language sites, permit most gliomas to be aggressively resected without language deficits. The composite language maps generated in our study suggest that our current models of human language organization insufficiently account for observed language function.
It is well accepted that surgical management of malignant gliomas has dramatically changed by using new advanced modalities. Preoperative precise anatomical and functional information enabled safe and sure surgical resection of malignant gliomas around the eloquent areas. It could be mentioned that surgery contributed to get good QOL of glioma patients. However, improvement of survival duration has not been achieved on GBM, and more effective adjuvant therapy should be expected. This time, I would like to show how it is possible to promote to more effective surgery, including our new intraoperative MRI system (Surgical Suite) and review the problems with glioma surgery. Future glioma treatment, including pathophysiology of glioma and future advancement of technology against glioma, will be addressed in my talk.
In glioma surgery, maximal cytoreduction followed by chemoradiotherapy is regarded standard of care for glioma treatment at present. Insular gliomas, however, are by many still considered to be inoperable. Anatomical localization, vascular supply, and the potential devastating complications make a complete resection highly impossible. We present our experience with the operative treatment of patients with insular gliomas in an “awake craniotomy” setting.
Sixteen consecutive patients with an insular tumor were operated awake during the period 2003 and 2008. Preoperatively, an extensive anesthesiologic and neuropsychologic/linguistic workup was performed. All patients underwent MRI with neuronavigation and when possible fMRI. After cortical stimulation, de Sylvian fissure and periinsular sulci were opened. Tumor resection was performed under speech (dominant hemisphere) and motor movement surveillance.
The patients’ average age was 41.6 (SD 10.2) years. Pure insular lesions were seen in two patients, a medial temporal base-insular glioma in one, insular frontoopercular and orbito-frontal-insular-temporal polar in five and eight patients, respectively. Presenting symptoms included epilepsy (94%), dysphasia (19%), and cognitive problems (19%). In 11 patients, the resection was near total (95–98%), and less than 95% in the remaining five patients. At histological examination we encountered 13 low-grade and 3 high-grade gliomas. The average follow-up was 1.8 (SD 1.4) years since surgery. Perioperatively seven patients clinically deteriorated. However, all patients with a low-grade glioma recovered to preoperative status. Two patients with a high-grade glioma have died during follow-up.
Insular glioma surgery, facilitated by (sub)cortical stimulation in an awake setting is feasible to acquire maximal cytoreduction in a safe manner. A dedicated surgical team is required, next to neurosurgeon, anesthesiologist, and patient interaction.
In the treatment of malignant brain tumors, maximum resection is recommended. However, aggressive resection tends to cause neurological deterioration such as paresis, aphasia, and many deficits of higher brain functions. For the maximum tumor resection without any neurological deficit, we performed intraoperative neurological monitoring of not only language and motor function, but also various higher brain functions in awake craniotomy.
From April 2003 to December 2008, 80 tumor resections under awake craniotomies were performed in Komagome Metropolitan Hospital. The tumors included 28 gliomas, 39 metastatic brain tumors, 4 radiation necroses, and 9 other brain tumors. Intraoperatively, we evaluated motor function in 47 surgeries, language in 26 surgeries, and higher brain functions such as imitation, writing, motor programming, visual memory, emotion, and so forth, in 19 surgeries according to the location of tumors. Functional MRI (f-MRI) and fiber-tracking integrated neuronavigation was also used during tumor resection. Neurological examination was performed before and after the surgery.
The mean extent of tumor resection was not inferior to that of our prior series of surgeries under general anesthesia. Of 47 patients harboring tumor within or adjacent to motor area, only four patients experienced permanent deterioration of paresis after surgeries. Among the 19 patients with tumors located nearby language area, only one patient suffered worsening of aphasia after surgeries. In the evaluation of higher brain function during the tumor resection, we found various symptoms such as imitation disorder, writing disturbance, memory loss and so on. But the symptoms were improved after discontinuation of the resection, and no permanent disorder was recognized.
Awake craniotomy contributes to the preservation of neurological functions with acceptable extent of tumor resection and keeps high performance status of the malignant brain tumor patient after surgery.
MRI during surgery has been proven to improve surgical resection rate and to reduce postoperative complication during brain tumor surgery. However, current commercially available intraoperative MRI system is relatively large and requires extra room for installment. To distribute such technique in world wide, we need to build a small intraoperative MRI system with good imaging quality.
With collaboration with Yoshida Dental Mfg. Co., Ltd., we created a new MRI system. Our concept in creating the machine is (1) the 5 gauss line is within the circle of 3 m in diameter and (2) the system weighs less than 3 tons, which will be the limit of routine hospital elevator carriage. (3) Also, we desired imaging time to obtain essential intraoperative imaging series should be less than 30 min.
Our system called “Vesalius,” fulfilled our requirements. The size of the system is half of the currently available systems and the 5 gauss line was confined within 2.6 m in diameter around the center of the machine. However, to reduce outer magnetic influence, we need to limit the interpolar gap to 32 cm. With this limitation, we improved operative bed and head clamping system. High-quality images (including T1, T2, FLAIR) with less image distortion was obtained within 30 min. We are now starting clinical application.
We have built a new intraoperative small low-magnetic-field MRI system, with high imaging quality. We believe this system will significantly improve future brain tumor surgery.
Disclosures: The first and second authors have no financial relationship with the manufacturing company. The last author is an employee of the company involved in the development of this system.
Information of glioma mass and involved pyramidal tracts are invaluable clinically to surgical trajectory plan and therapeutic outcome. We used diffusion tensor tractography in a stereoscopic virtual reality (VR) environment to perform presurgical planning for cerebral glioma with pyramidal tract (PT) involvement.
45 glioma cases of Patient-specific Digital Imaging and Communications in Medicine (Dicom) data from diffusion tensor imaging (DTI) and three-dimensional MRI scans were transferred to the workstation (Dextroscope; Volume Interactions Pte. Ltd., Singapore). The tumor and adjacent PT were segmented and reconstructed for presurgical planning and intraoperative guidance. The number of effective pyramidal tracts (EPT) and the motor strength after operation are compared with those before operation.
On this VR workstation, diffusion tensor tractography and three-dimensional MRI scans were integrated. The brain tissue, glioma mass and PT were obviously displayed. An individual presurgical project was meticulously planned in every case. There was no significant difference between presurgical and postsurgical motor strength, as well as the presurgical EPT and postsurgical EPT in our study. A positive relationship was found between presurgical EPT and follow-up motor strength.
Diffusion tensor tractography in this VR environment exhibits the information of glioma mass and the involved PT, which is invaluable in designing a suitable presurgical trajectory. According to the presurgical tractography we can predict as well as optimize the therapeutic outcome.
To assess the ability of preoperative DTI-Fiber Tracking (DTI-FT) to predict the extent of resection achievable during surgery of gliomas performed with intraoperative brain mapping.
Seventy patients with lesions located in the frontal, temporal and insular lobes of the dominant hemisphere were included. All of the patients, after preoperative neuropsychological assessment, underwent DTI-FT of the corticospinal tract (CST) and of the inferior fronto-occipital (IFO) fasciculus. Each of the tracts (CST and IFO) was scored by two independent observers as being inside or outside of the tumor. Awake intraoperative language and motor brain mapping included EEG, ECoG, EMG, and MEP monitoring. Surgery was carried out according to the functional boundaries of the lesion. Then, for each patient preoperative and postoperative MR images and DTI-FT were loaded into the neuronavigation software and image fusion was used in order to evaluate the extent of resection as well as the correspondence between the resection boundaries and the preoperative DTI-FT. A correlation between the preoperative score (inside/outside of the lesion) of each tract (CST, IFO) and the extent of resection was also investigated.
The CST was inside the lesion in 38 patients (54%) and the IFO in 23 patients (33%). In all the cases in which the CST (38 patients) or IFO (23 patients) was inside the lesion, we found only incomplete resections (subtotal 19, partial 31). When CST and IFO were both outside the lesion, complete resections was carried out (20 patients).
DTI-FT of the CST and IFO can help to predict the possibility of total resection. In particular, when CST and IFO are inside the tumor a radical removal is rarely possible, while when outside of the lesion, an extensive resection in the dominant hemisphere with the aid of brain mapping may be possible.
Optical coherence tomography (OCT) is a noninvasive imaging technique with a micrometer resolution. It operates in no tissue contact and generates images based on the reflection of infrared light. We have demonstrated that OCT discriminates tumor and normal brain in experimental preclinical settings. We have developed an OCT-integrated operating microscope and we have performed a clinical study in 20 patients.
The optical path of an operating microscope was optimized for the near-infrared and conventional light spectrum. The OCT-integrated microscope was used to analyze brain tumor tissue and areas of the resection cavity during resection of 20 malignant gliomas. We have used post-image acquisition processing to compensate for movements of the brain and to calculate light attenuation factors for brain and brain tumor tissue.
OCT imaging of cortex and white matter showed a typical light attenuation profile. Tumor tissue was identified by loss of the normal light attenuation and appearance of a prominent microstructure. OCT analysis allowed discrimination of normal brain tissue, invaded brain, and solid tumor tissue. In a clinical study, OCT analysis of the resection cavity demonstrated residual tumor following macroscopically complete resections, which was confirmed by histological analysis. The blinded scoring of 41 tissue samples for OCT characteristics (microstructure and light attenuation) showed a high correlation with conventional histology (r=0.99). The OCT integrated microscope remained fully functional for microsurgery and allowed OCT analysis of a 3–50 mm scan line at a zoom factor of 2–14.3.
Integration of OCT into an operating microscope allows continuous tissue analysis of the resection edge during resection of brain tumors and identifies residual tumor prior to secondary changes such as those induced by tissue contusion. OCT allows the detection of residual tumor and extends the information provided by conventional operating microscopes today.
A variety of cancers, including malignant gliomas, overexpress transforming growth factor β (TGF-β), which helps tumors evade effective immune surveillance through inhibiting cytolytic activity of natural killer cells and CD8+ cytotoxic T lymphocytes (CTL), and enhancing the generation of regulatory T cells (Tregs). We hypothesized that inhibition of TGF-β would improve the efficacy of vaccines targeting glioma-associated antigen (GAA)-derived CTL epitopes by reversal of immunosuppression. In mice bearing intracranial GL261 gliomas, intraperitoneal (i.p.) administration of the TGF-β-neutralizing monoclonal antibody (mAb), 1D11, was combined with subcutaneous vaccinations of synthetic peptides for GAA-derived CTL epitopes, GARC-1 (77–85), and EphA2 (671–679) emulsified in incomplete Freund’s adjuvant (IFA). Mice receiving the combination regimen exhibited significantly prolonged survival compared with mice receiving either 1D11 alone, GAA-vaccines alone or mock-treatments alone. 1D11 administration enhanced the systemic induction of antigen-specific CTLs in GL261-bearing mice. Flow cytometric analyses of brain infiltrating lymphocytes (BILs) revealed that i.p. 1D11 administration increased GAA-reactive/IFN-γ-producing CD8+ T cells and reduced CD4+/FoxP3+ Tregs in the glioma microenvironment. Systemic 1D11 administration also upregulated plasma levels of interleukin (IL)-12, macrophage inflammatory protein (MIP)-α and interferon (IFN)-inducible protein (IP)-10, suggesting a systemic promotion of type 1 cytokine/chemokine production. Furthermore, 1D11 treatment upregulated plasma IL-15 levels and promoted the persistence of GAA-reactive CD8+ T cells in the GL261-bearing mice. These data suggest that systemic inhibition of TGF-β by 1D11 can reverse the suppressive immunological environment of intracranial tumor bearing mice both systemically and locally, thereby promoting the therapeutic efficacy of GAA vaccines.
Recently, a novel spontaneous mouse glioma model has been developed, where Sleeping Beauty (SB) transposase-mediated stable transfection of oncogenes in the mouse brain induces CNS tumors recapitulating human malignant glioma. This system allows us to address the mystery of early events in glioma genesis. We have previously demonstrated critical roles of type 1 adaptive immunity in control of established intracranial gliomas in mice. Although decreases in type 1 cytokine levels have been reported in glioma patients, it has not been determined whether such changes merely reflect consequences of established glioma and/or whether preexisting alterations play any roles in occurrence of gliomas. We hypothesized that the interferon (IFN)-α signaling pathway might play a pivotal role in immune surveillance protecting against glioma development. We induced gliomas in IFN-α receptor (R)1/− or wild-type neonatal mice via intracranial injections of plasmids encoding NRas, short hairpin RNA for p53, and EGFRvIII, along with the fourth vector encoding SB transposase. Spontaneous tumors in IFN-αR1/− mice grew and ceased hosts more rapidly than those in immune-competent mice. When tumor-infiltrating lymphocytes were analyzed, IFN-αR1/− mice showed increased accumulation of CD4+FoxP3+ regulatory T cells and CD11b+Gr-1+ myeloid-derived suppressor cells in tumors. Furthermore, following intravenous infusions of Pmel-I mouse-derived CTLs, IFN-αR1/− mice exhibited reduced accumulation of transferred CTLs at the tumor site compared with wild-type mice. Of great interest is that when primary tumor cells were cultured and subsequently injected into the brain of immune-competent hosts, IFN-αR1/− glioma cells demonstrated a remarkably lesser degree of tumorigenicity compared with those from wild-type mice, suggesting a possibility that IFN-αR1/− hosts allowed growth of tumors that could be immunologically rejected or inhibited in immune-competent mice. These results suggest a potentially pivotal role of IFN-α reactivity in CNS immunological surveillance of gliomas, and warrant well-designed immunoepidemiological studies in humans.
To perform a prospective phase II clinical trial using adjuvant autologous dendritic cell-based immunotherapy for malignant gliomas.
Patients aged 14–70 years with malignant gliomas and initial treatment KPS score>50 were enrolled. Eighteen patients with newly diagnosed malignant gliomas underwent conventional treatments (surgery, radiotherapy, and chemotherapy) and adjuvant autologous dendritic cell immunotherapy. The control group (n=18) underwent conventional treatments only.
The follow-up time ranged from 18 to 50 months (median 30 months). Fourteen of 18 patients (78%) who received immunotherapy survived, but only 17% survived in the control group (p<0.01). The median survival time for the immunotherapy group was 25.5 months and for the control group 16.7 months (p<0.05). Four of nine patients with glioblastoma multiforme (GBM) (44%) who received immunotherapy survived more than 24 months compared with none in the control group (p=0.03). The survival fraction was significantly higher in the immunotherapy group compared with the control group based on Kaplan-Meier analysis (p<0.01). The remarkable response rate in the immunotherapy group was 56% but only 6% in the control group (p<0.01). The recurrence rate was lower in the immunotherapy group (8/18) than in the control group (17/18) (p<0.01). The KPS maintenance rate was significantly better in the immunotherapy group (80%) than in the control group (50%) (p=0.02).
Adjuvant autologous dendritic cell immunotherapy may improve survival time and rate, reduce recurrence rate, and maintain better quality of life. It seems to be a safe and effective adjuvant treatment for malignant gliomas. Multimodality treatment for malignant gliomas is essential.
To investigate the safety and clinical responses of immunotherapy targeting WT1 gene product, 31 patients with WT1/HLA-A*2402-positive recurrent glioblastoma (GB) were evaluated in a phase II clinical study.
All patients were intradermally injected with an HLA-A*2402–restricted, modified 9mer WT1 peptide every week for 12 weeks. The responses on MRI were analyzed by RECIST criteria 12 weeks after the initial vaccination. Progression-free (PFS) and overall survival (OS) after initial WT1 treatment were estimated. The percentage of numbers of WT1 specific cytotoxic T lymphocytes (CTL) among those of all lymphocytes in the patients’ peripheral blood were drawn using flow-cytometry, to clarify the effect of concomitantly used steroids and antiepileptic drugs (AEDs). The same analyses were done with peripheral blood from another group of 14 patients with WT1/HLA-A*2402–positive malignant gliomas, who have been receiving oral administration of temozolomide, to give a rationale of the combined chemoimmunotherapy.
The protocol of phase II study was well tolerated. The clinical responses included: none with complete response (CR), 2 patients with partial response (PR), 13 with stable disease (SD), and 16 with progressive disease (PD). The overall response rate (CR+PR) was 6.5%, and the disease control rate (CR+PR+SD) was 48.4%. The median PFS was 16.0 weeks, and PFS at 6 months was 29.0%. The median OS after initial vaccination reached to 267 days. The analyses of WT1-specific CTL revealed that neither concomitantly used agents (steroids, AEDs) nor temozolomide affects the percentage of WT1-specific CTL.
This study showed that WT1 vaccine therapy for patients with GB was safe and had a clinical response compared with any other therapeutic modalities. Analyses of WT1-specific CTL raised the possibility of concomitant use of considerable dose of steroids and the rationale of combined chemoimmunotherapy planning in the future.
Autologous formalin-fixed tumor vaccine (AFTV) could induce killer lymphocytes in vivo and exhibited prophylactic effect against recurrence of hepatocellular carcinoma after the surgery in a randomized clinical trial. AFTV was combined with radiotherapy (RT) for primary glioblastoma (GBM) in a phase I/IIa trial.
Adults after resection of GBM were treated with standard RT and AFTV. AFTV was prepared with fixed GBM fragments, PPD, and BCG extracts suspended in saline. When the cumulative radiation dose reached 32–36 Gy, AFTV treatment was started and injected three times weekly. Second delayed type hypersensitivity tests (DTH2) were performed 2 weeks after third and last vaccination. The primary end point was overall survival (OS) and secondary end point was progression-free survival (PFS).
Twenty-two eligible patients from two hospitals were accrued with a median follow-up time of 18.9 months (M). Median OS and PFS were 18 M and 7.6 M, respectively. OS and PFS of RT with AFTV were not significantly different from those of RT with temozolomide in the same hospital. The patients whose DTH2 reaction was more than 12 mm showed significantly longer OS and PFS than those with DTH2 reaction less than 12 mm (OS, 22.5 months vs. 18 months; PFS, 14 months vs. 7.5 months). There was no significant adverse event whose grade was more than grade 2.
Concomitant use of AFTV in the late stage of radiotherapy resulted in comparable median times of PFS and OS compared to our historical controls of RT with temozolomide without severe adverse events.
Our previous preclinical studies have demonstrated that intramuscular (i.m.) administration of a Toll-like receptor 3 ligand poly-ICLC remarkably enhances induction of type 1 cytotoxic T-lymphocytes (CTLs) and improves therapeutic efficacy of vaccinations against glioma-associated antigen-derived CD8+ T cell epitopes. Based on these studies, we have developed a phase I/II trial. Human leukocyte Antigen (HLA)-A2+ participants with recurrent malignant glioma received intra-lymph-node injections of type 1 dendritic cells (DCs) loaded with HLA-A2 binding peptides EphA2 (883–891), IL-13Rα2 (345–353: 1A9V), YKL-40 (202–211), and GP100 (209–217: 2M) at 2-week intervals. Participants also received twice weekly i.m. injections of 20 μg/kg poly-ICLC. Participants who demonstrated positive radiological response or stable disease without major adverse events were allowed to receive booster vaccines. Primary end points were assessments of safety and immunological responses. Clinical and radiological responses were also evaluated. To date, 13 participants (six with glioblastoma multiforme [GBM], five with anaplastic astrocytoma [AA], and two anaplastic oligodendroglioma [AO]) have received vaccinations with no major adverse events. Increased CD8+ cells reactive to HLA-A2.1EphA2 (883–891) or HLA-A2.1IL-13Rα2 (345–353) tetramers were detected in postvaccine peripheral blood mononuclear cells (PBMC) in seven of nine participants evaluated. These patients also demonstrated upregulation of a chemokine receptor CXCR3 on CD8+ PBMC following vaccines, indicating that the vaccine regimen induced type 1 CTL responses. One of these participants with recurrent GBM exhibited partial radiological response, which persisted for 7 months with booster vaccines. Biopsy of tumor site after vaccination in this participant revealed intensive infiltration of CD8+ T cells and macrophages. Six patients achieved progression-free at 6 months (two GBM, two AA, and two AO). Fisher’s exact test indicated an association between positive tetramer response and 6-month progression-free survival, suggesting a possible correlation between tetramer-detected immune responses and clinical response. These interim data demonstrate preliminary safety and immunological activity of poly-ICLC-assisted type 1 DC-based vaccines.
Grants 1R21CA117152, 1P01NS40923.
Replication-conditional viral mutants promise an alternative treatment modality for malignant gliomas because of their potential to selectively eradicate tumor cells through oncolysis and spread their progeny throughout the tumor tissues. Initial data obtained from clinical trials have demonstrated their clinical safety, but not their effectiveness, probably because of the still inefficient delivery, propagation, and spread of such viruses in tumor tissues. Efficient delivery of oncolytic viruses to tumor regions is desirable due to the diffusely invasive property of malignant gliomas as a major obstacle in their treatments. Cell-based delivery of oncolytic viruses has become a promising strategy to target the diffused tumors. Some cell types like neural and mesenchymal stem cells have a homing ability to pathological regions including tumors. However, oncolytic viral carrier cells may not survive long enough to migrate to scattered tumor regions, because of viral lytic cycle. To overcome this limitation, we have developed a novel strategy to improve carrier cell-based virotherapy. This new method consists of two parts: (1) transformation from replication-deficient prooncolytic virus to replication-competent oncolytic virus, (2) inducible viral transformation by tumor microenvironment. Our results showed viral gene expression was suppressed in prooncolytic virus, but activated by recombinase-based inducer in the infected cells. Once prooncolytic viruses changed to replication-competent viruses, these induced viruses efficiently propagated in glioma cells compared to noninduced viruses. This strategy enables viral carrier cells to survive for a longer period of time, compared to cells carrying general oncolytic viruses. We believe that this promising method will enhance the development of cell-based virotherapy.
Oncolytic HSV-1-derived viruses (OVs) are being investigated as treatment modalities for many cancers. However, a clear understanding of the action of this biological therapy in human patients is currently not feasible. We have been investigating OV therapy induced changes in the tumor extracellular matrix.
Transcript pro-filing of secreted proteins involved in angiogenesis revealed a significant induction of cystein-rich 61 (CYR61) gene expression in tumors treated with OV in vivo. CYR61 is a secreted ECM protein that can bind to and activate integrins αvβ5 and αvβ5. Increased CYR61 in the ECM of breast cancer cells has also been shown to activate an autocrine loop resulting in upregulation of its own receptor αvβ5 resulting and increased sensitization to integrin anatgonists (Menendez JA, et al. Oncogene. 2005). Consistent with this we found that glioma cells stably overexpressing CYR61 were much more sensitive to the integrin antagonist, cilengitide, than control cells. Hence we hypothesized that OV treatment, which induces secreted CYR61, would increase integrins αvβ5 on both infected and uninfected glioma cells, and sensitize them to cilengitide. We tested the effect of OV treatment on glioma cells for the expression of integrin receptors αvβ5. Moreover, we detected synergistic killing of glioma cells treated with OV and cilengitide compared to cells treated with either agent alone. We tested this hypothesis in immunocompromised mice with intracranial tumors (U87dEGFR). Five days after OV treatment mice were injected systemically with a single dose of cRGD (5 mg/kg) or PBS. Mice treated with OV and cilengitide survived significantly longer than mice treated with OV alone (p<0.01 between OV and cRGD + OV).
These results indicate that OV therapy can sensitize glioma cells to cilengitide and lays a rationale for designing future clinical trials combining the two agents.
Medulloblastoma is one of the most malignant brain tumors. From 1995 to 2009 we treated 80 cases of medulloblastoma. There were 48 female and 32 male in our series, ranging from 2 months old to 55 years. In all cases, total removal was attempted and achieved in 68 cases (85%). Total removal was impossible with brainstem attachment. Radiotherapy following surgery was routine procedure (except in very young infants), and starting from 2005 we used chemotherapy. In our series, recurrent rate was 10% with total removal, compared with 80% in the other group. Complications included lower cranial nerve palsy, cerebellar dysfunction and mutism, which all recovered.
To optimize physical and biological parameters of radio-therapy in average-risk medulloblastoma.
Twelve children with average-risk medulloblastoma were treated with hyperfractionated radio-therapy (HFRT). Radiotherapy consisted of two fractions daily of 1 Gy each, 6 h apart, 36 Gy/36 fractions of craniospinal irradiation (CSI), followed by conformal tumor bed boost (32 Gy/32 fractions) for a total dose of 68 Gy/68 fractions over 6.5 weeks. Chemotherapy was reserved for relapse.
The median age of the cohort was 8 years (range, 5–13 years). Acute hematologic toxicity was self-limiting, not requiring interruption of treatment, growth factor support or platelet transfusions in any patient. Grade II neutropenia and thrombocytopenia was seen in 33% and 16% patients, respectively. No one developed grade III–IV thrombocytopenia. Five of 12 children had suboptimal mean Intelligence Quotient (IQ) scores (mean IQ less than 90) even before initiation of radiotherapy. Two patients relapsed, at 18 and 21 months from diagnosis, both supratentorially and in the leptomeninges, of which one died at 25 months despite salvage chemotherapy. With a median follow-up of 19 months (range, 6–30 months), the 2-year event-free survival and overall survival were 65% and 100%, respectively. Mean IQ scores for all tested domains were preserved in all the 10 children evaluable at 1 year. Two children developed hypothyroidism necessitating thyroxine supplementation, while no patient developed significant sensorineural deterioration.
Preliminary analysis suggests that HFRT without chemotherapy for average-risk medulloblastoma is a promising option with acceptable self-limiting acute toxicity, without an unduly increased risk of relapse. Long-term results on larger number of patients are required to confirm its role in reducing the known late sequelae of radiotherapy.
Incidence of invasive cancer rises from adolescence (SEER data). Brain tumor types change with age; embryonal brain tumors become less common in adolescence (CBTRUS data). Our hypothesis is that medulloblastomas (MB), an embryonal tumor, is biologically different in adolescence due to differences in host biology. Adolescence is defined as ≥9, ≤15 years in this study.
mRNA expression pro-files of 71 MB and 6 normal brain specimens were characterized using Affymetrix HG-U133 plus 2.0 Gene-Chips. Unsupervised hierarchical clustering identified four clusters. Cluster C comprised six normal brain tissues and one tumor (nonadolescent). Class comparison was performed between adolescents and nonadolescents, between clusters to establish significant differences in gene expression (using FDR=0.001, CI=95%). Differentially expressed genes were categorized in Gene Ontology groups using the DAVID tool.
Adolescents accounted for 44.1% (15/34) of cluster D, cluster A (6/27), cluster B (0/9), p=0.019 (Fisher’s Exact Test). There were 21 adolescents; 71.4% of adolescents have cluster D signature. Cluster D has three subgroups; majority of adolescents occurred in subgroup D1 (11/20), D2 (1/11), D3 (3/3). No significant difference in gene expression of adolescent MB compared to nonadolescents across clusters. Tumor clusters A, B, D compared among each other; cluster A significantly overexpressed genes involved in brain development, organ, cellular, neurite morphogenesis, neurogenesis and axon guidance, compared to cluster D (p=0.0000007–0.04). Class comparison between subgroup D1 and D2 revealed 153 significantly differentially expressed genes. D1 significantly underexpressed genes of light perception (p=0.0008), pathways (p=0.0004–0.001) involved in T-cell receptor (TCR) locus.
Adolescent MBs have different molecular signatures, a majority showing the D1 signature, underexpressing genes of light perception and TCR compared to D2. The presence of T-cells and its potential significance in the prognosis of MB is being explored.
Although the outcome of brain tumor in children has improved significantly over decades, the high-risk (HR) brain tumors such as recurrent, residual or infant disease still have a poor prognosis. Autologous hematopoietic stem cell transplantation (AHSCT) has been applied to improve the outcome of HR brain tumors. We report on the outcome for HR brain tumors receiving AHSCT at a single institution.
Between October 1999 and August 2008, 20 patients received AHSCT using peripheral blood stem cells for HR brain tumor at Asan Medical Center. Their medical records including transplant data were reviewed and the outcome was analyzed as of December 2008.
Of 21 patients, eight had medulloblastoma (two infant, two metastatic, three recurrent), six had supratentorial PNET (one infant, one residual, four recurrent), two had pineoblastoma (one infant, one recurrent), two had ATRT, and two had recurrent GCT. At transplantation, 13 patients were in CR and 7 in PR. Nine patients (CR 5 and PR 4) received conditioning with cyclophosphamide/melphalan (CM) and 11 (CR 8 and PR 3) with carboplatin/thiotepa/etoposide (CTE). After AHSCT, five patients (four in CM, one in CTE) relapsed and four patients (three in CM, one in CTE) died of causes other than disease progression. With a median follow-up of 33.7 months, 3-year EFS and OS were 50.8% and 56.1%, respectively. CR at the time of AHSCT tended to have better EFS compared to PR (64.5% vs. 24.6%, p=0.103). The 3-year EFS for 11 patients conditioned with CTE was better than that of nine with CM (79.5% vs. 22.2%, p=0.014).
AHSCT with CTE is highly successful treatment modality for HR brain tumors in children. Given the limitation of this study including small number of patients as well as heterogeneous population, further study is warranted to clarify the role of AHSCT in pediatric brain tumors.
The clinicopathological features of atypical teratoid/rhabdoid tumors (AT/RTs) in Japan, and the importance of the differential diagnosis from PNET/medulloblastoma and germ cell tumor are described.
Forty-one patients with AT/RT in Japan were analyzed about clinicopathological features.
Forty-one AT/RTs (23 male and 18 female) were aged from 1 month to 37 years (mean 2.9 years) and tended to show increased intracranial pressure by obstructive hydrocephalus and/or tumor volume. The tumor locations were in the posterior fossa (63%), brain hemispheres (22%), spinal (12%), and pineal region (3%). Interestingly, younger patients (under 2 years) with AT/RT tended to be located in posterior fossa or spinal region. Leptomeningeal dissemination showed more than 50% of AT/RTs. Histologically, AT/RT is defined as a polymorphous neoplasm often featuring rhabdoid, PNET, epithelial, and mesenchymal components. AT/RTs usually include PNET components and occur mainly in the posterior fossa, so mimic medulloblastoma. AT/RT is characterized by the cytogenetic finding of monosomy 22 rather than i(17q). The tumor is similarly mistaken for PNET at supratentorial immunophenotypic diversity, particularly features indicative of epithelial and mesenchymal differentiation. Nonetheless, the remarkable spectrum of tissues that typify teratoma is absent in AT/RT. The prognosis of this tumor is far less favorable than that of PNET/medulloblastoma or germ cell tumor. For the diagnosis of AT/RT, negativity of INI1 protein is important and IGF-1 receptor tended to express in this tumor cells.
This study describes the clinicopathological features of 41 AT/RT in Japan, and we emphasize the necessity for distinguishing this unique tumor from other pediatric central nervous system neoplasms. And INI1 is important for this tumor diagnosis. Moreover, IGF-1 may play a role as growth factor of AT/RT cells.
The purpose of this study was to determine whether combination treatment involving radiation therapy (RT) and valproic acid (VA), a histone deacetylase (HDAC) inhibitor, provides superior antitumor effect than either therapy alone, when evaluated in the context of an intracranial xenograft model of atypical teratoid rhabdoid tumor (ATRT).
300,000 ATRT cells, modified with luciferase for bioluminescence imaging (BLI), were injected into the supratentorial compartment in each of 35 mice. Tumor cells were monitored by BLI until exponential growth was indicated, at which time mice were randomized into four treatment groups: (1) vehicle only treatment; (2) RT only (0.5 Gy/day × 5 days); (3) VA only (150 mg/kg/BID × 7 days); or (4) RT and VA administered concurrently as indicated for monotherapy administrations. BLI was used for continuous monitoring of tumor response to therapy, and all mice were followed until euthanasia was required due to presentation of symptoms associated with neurological compromise from tumor burden.
Mice receiving concurrent RT + VA treatment experienced significant survival benefit relative to vehicle treated control group animals (p=0.0009), as well as with respect to mice receiving either treatment as a monotherapy (p=0.0017 compared with VA alone, and p=0.0223 compared with RT alone). Bioluminescence imaging revealed that combination therapy reduced the incidence of neuraxis dissemination of intracranial tumor.
Our results indicate that combination therapy using an HDAC inhibitor + RT has enhanced antitumor activity relative to either therapy used alone, and therefore warrants consideration for use in the young pediatric patients that are affected by this highly malignant CNS tumor.
Atypical choroid plexus papilloma (APP, WHO grade II) were newly defined by the WHO in 2007 in addition to the previously defined choroid plexus tumors (CPT), which were choroid plexus papilloma (CPP, grade I), and choroid plexus carcinoma (CPC grade III).
The CPT-SIOP-2000 study is an international observation and randomized treatment trial for all patients with CPT. After maximal surgical resection, all patients with APP and residual tumor, with metastases or with CPC regardless of residual tumor, were treated with chemotherapy (VP16, vincristine, and either carboplatinum or cyclophosphamide). Irradiation was given to patients older than 3 years of age between the second and third cycle. The other CPT patients were followed without further treatment.
Until December 2008, 23 nations registered 152 patients with CPT. In the latest interim analysis, patients with APP (n=32) had a median age of 0.7 years and were thus younger than patients with CPP (2.2 years; n=42) or CPC (2.2 years; n=34). Primary metastases were frequent in CPC (20%) and APP (16%), but also occurred in CPP (5%). Complete resection was achieved in 81% of CPP, 65% of APP, and 50% of CPC. Patients with APP had an intermediate 5-year EFS (84±10% SD) compared to CPP (91±8% SD) and CPC (28±20% SD; p<0.0002). Objective early response (complete remission, partial response) after two cycles of chemotherapy among APP patients receiving further treatment was seen in 6 of 10 patients.
The intermediate position of APP defined by histology was confirmed by clinical data of tumor resection and frequency of metastases but not in age at diagnosis. Further molecular studies are necessary to determine if there is a further subgroup hidden with different biology among those very young children. APP seem to respond to chemotherapy.
Choroid plexus tumors are rare, and the pattern of dissemination has not been described in details yet.
A retrospective analysis of the CPT-SIOP-2000 database (Choroid Plexus Tumor-International Society of Pediatric Oncology 2000) included patients randomized and treated as well as those registered only.
By June 2008, 100 patients were formally registered. In 21 of those, metastases were reported (10 of 21 male; median age, 2.7 years; range, 0.1–11.8). This includes five patients documented as metastatic but with a confirmation pending, 12 tumors confirmed metastatic at presentation, 3 tumors presenting as localized but later recurring as metastatic (secondary), and 1 presenting metastatic and again recurring metastatic. Primary location was the lateral ventricle in 20 and the 4th ventricle in 1 patent. The histology was choroid plexus papilloma 3 (CPP), atypical choroid plexus papilloma 5 (APP), and choroid plexus carcinoma 11 (CPC). Primary surgery was gross total resection 12, partial resection 6, or biopsy 2. All tumors were spread through cerebrospinal fluid. Eight patients received radiation and chemotherapy, 10 chemotherapy only, and 3 neither. Those three had complete resections of either CPP (2) or APP (1), all are alive, and one recurred. Among the other 18, the median event-free survival (EFS) was 1.33 years, and the 4 year EFS 51% (±14 SD), 4 year overall survival 72% (±19 SD, median not reached). In all recurrences leptomeningeal spread was documented.
Leptomeningeal spread is common in choroid plexus tumors, in particular when the primary location is supratentorial. The prognosis seems to be better than other metastatic brain tumors.
Gliosarcoma (GS) represents a glioblastoma with a sarcomatous component, which is presumed to be a metaplastic differentiation of the glioma cells. We studied the clinical relevance of this histological glioblastoma subentity within the pediatric population.
Patient data were obtained from the German HIT-GBM database where clinical data of more than 600 pediatric patients with centrally reviewed high-grade gliomas are accumulated. By applying defined inclusion criteria (diagnosis of GS proven by central neuropathological review; patient age 0–21 years) five patients were identified. In addition, after review of English medical scientific literature 19 additional cases were found.
The relative frequency of GS in the German HIT-GBM database was only 2.4%. In the whole series of 24 pediatric GS including previously reported cases the male-to-female e ratio was 1.3:1. GS was found in all pediatric age groups with a median age of 11 years but there was an unexpectedly high accumulation in infants (7 of 24 less than 3 years of age, 29%). GS showed a strong predilection of the cerebral hemispheres (23 out of 24 cases). Symptoms of increased intracranial pressure were the leading symptoms of a short clinical history with a median duration of 1 month. Interestingly, six patients (25%) were reported with a history of cranial radiotherapy prior to GS diagnosis. In 62% of GS in our series, gross total resection was achieved. Median overall (OS) and event-free survival (EFS) of the total cohort were 13.3 and 9.8 months, respectively, but infants showed superior survival.
GS is a very rare tumor entity in children. Literature review suggests a relatively higher incidence in infants and in patients with a previous history of radiotherapy.
To assess the efficacy of temozolomide (TMZ) on diffuse brainstem glioma as concomitance with radiotherapy and as adjuvant treatment postradiotherapy in children.
Eighteen children (median age at diagnosis, 8.3 years) with newly diagnosed brainstem tumors that were treated with TMZ at Taipei Veterans General Hospital from January 2004 to December 2008 were eligible for the study. They were divided into two groups by the different treatment modalities. RT+TMZ (radiotherapy alone followed by adjuvant TMZ) group received conventional radiation after initial diagnosis. CCRT+TMZ (concomitant chemoradiotherapy followed by adjuvant TMZ) group received concurrent chemotherapy during radiation with TMZ (75 mg/m2/day). After the completion of radiotherapy, TMZ (150 mg/M2) was administered orally once daily for 5 consecutive days for all enrolled patients. Treatment cycles were repeated every 28 days. We evaluate the progression-free survival in both groups of patients.
There were 10 patients in RT+TMZ group and 8 in CCRT+TMZ group. All patients experienced progression of disease. Twelve patients (75%) died in the study period and all deaths were attributed to disease progression. The median progression-free survival (PFS) was 7.4 months for the RT+TMZ group and 6.4 months for CCRT+TMZ group. The 6 months and 1 year PFS in RT+TMZ group were 70% (SD, 14%) and 30% (SD, 14%) respectively and in CCRT+TMZ were 50% (SD, 17%) and 0% respectively. The log-rank test in PFS between the two groups was p=0.09.
In this study, RT followed by TMZ or CCRT followed by TMZ did not alter the poor prognosis of the newly diagnosed diffuse brainstem glioma in children. In addition, CCRT followed by TMZ did not achieve a better outcome than RT followed by TMZ.
Optic pathway/hypothalamic pilocytic astrocytomas in children are usually treated with chemotherapy following a surgical biopsy. In this report, we retrospectively considered the role of surgical intervention. In a series of 32 patients without neurofibromatosis type 1, the median age at initial treatment was 2.7 years (range, 0–12 years). Twenty-five cases were verified by histology, and seven cases were diagnosed by MRI findings. Twenty-eight patients received chemotherapy. All patients were alive at median follow-up of 65 months. Aims of surgery at the initiation of treatment were biopsy in 15 cases (1 stereotactic and 14 craniotomies) and debulking in nine cases. The 14 open biopsies revealed pilocytic astrocytoma; however, noticeable complications occurred in six children after the biopsies. Review of preoperative MRIs showed that all had typical findings indicating pilocytic astrocytoma. The open biopsy offered no noteworthy benefit for the patients despite surgical risk and delay of chemotherapy. The extent of the nine resection surgeries was 80% or less removal, and postoperative adjuvant therapy was needed for seven of the nine patients. The remaining seven children who did not undergo surgery obtained remission with chemotherapy alone. One child underwent decompression of the optic canal that yielded amelioration of decreasing vision. After relapse in 11 patients, 19 bulk-reduction surgeries were performed. Surgical resection was not curative in any patient. In six patients, mostly older children, cystic expansion of tumor was partially resected, resulting in additional remission. In six children beyond the age of adolescent, spontaneous tumor involution was observed. In conclusion, considering the risk of open surgery and the effectiveness of chemotherapy, the role of surgical intervention is restricted to bulk-reduction surgery only when it is inevitable, especially at relapse after chemotherapy.
To discuss the clinical feature, treatment, and prognosis of intracranial nongerminomatous malignant germ cell tumors (NGMGCTs).
The records of 39 patients who were treated in Shanghai Huashan Hospital between 1995 and 2007 were reviewed retrospectively. According to the classification of Matsutani in 1997, they were grouped into intermediate prognosis group and poor prognosis group based on the histology of the tumor. Clinical manifestation, diagnosis, treatment and outcome were analyzed in each group.
In these 39 cases, there were 15 mix germ cell tumors, 15 immature teratomas, 7 embryonal carcinomas, and 2 yolk sac tumors. All patients were treated surgically. The tumor was totally removed in 29 cases, subtotally in 5, and partially in 3. Biopsy was performed in the other two cases. Thirty-four patients (87.2%) were followed up. The common 5-year survival rate was 36.8%. The 5-year actuarial overall survival rate for patients in the intermediate prognosis and poor prognosis groups were 42.6% and 0%. Chemotherapy combined radio-therapy has significant relationship with the prognosis of intermediated prognosis group (p=0.039). The 5-year survival rate of immature teratoma patients who received Gamma Knife surgery was 100%. It had significant difference (p=0.0049) compared to the 5-year survival rate of patients who did not receive Gamma Knife surgery.
NGMGCTs can be divided into intermediate group and poor group based on the prognosis (p=0.0003). Embryonal carcinoma can be classified to the intermediate prognosis group because of its similar prognosis with immature teratoma and mixed tumors mainly composed of germinoma or teratoma. Surgery is the first choice for NGMGCTs because treatment should be based on tumor histology. For patients in the intermediate prognosis group, combined treatment, including surgical resection, radiotherapy, chemotherapy and Gamma Knife surgery, was effective.
To review the optimum treatment for newly diagnosed intracranial germ cell tumors.
A retrospective review of medical records at our department from April 1990 to March 2007 identified 108 patients, 86 males and 22 females aged from 2 months to 45 years (median 14 years), with newly diagnosed intracranial germ cell tumors. Their diagnoses were germinoma in 83 patients, and nongerminomatous germ cell tumor in 25.
In the patients with germinoma, the 10-year overall and progression-free survival rates at median follow-up period of 99 months were 86% and 74%, respectively. Tumor recurrence was at 6–153 months (median 26 months) after starting the initial therapy. Patients treated with chemotherapy alone had shorter progression-free survival rate, and patients treated with chemotherapy followed by reduced-dose radiation therapy to the whole ventricle, whole brain, or craniospinal axis had significantly better progression-free survival rate compared to radiation alone or reduced-dose radiation therapy to focal field. Nongerminomatous germ cell tumors were divided into good, intermediate and poor prognosis group as proposed by Japanese Pediatric Brain Tumor Study Group. In the good and intermediate prognosis group, the 10-year overall and progression-free survival rates were 93% and 100%, respectively. In the poor prognosis group, the 3-year overall and progression-free survival rates were 56% and 29%, respectively. In both groups of nongerminomatous germ cell tumors, all of the patients in whom radiographic lesions disappeared after combination therapy of surgical resection, radiation therapy, and chemotherapy were alive without recurrence.
Chemotherapy followed by reduced-dose radiation therapy to the field covering the whole ventricle at least was effective in the treatment of germinoma. In the initial treatment of nongerminomatous germ cell tumor, combined treatment, including salvage surgery, is essential to achieve the complete disappearance of radiographic lesions.
To review the effectiveness of reduced-dose and restricted volume radiation-only therapy in treatment of intracranial germinoma (IG), and to assess the feasibility of reducing or eliminating the use of chemotherapy.
Between January 1996 and March 2007, 51 patients with IGs underwent variable treatments at Taipei Veterans General Hospital. A retrospective analysis was done of 38 patients with either reduced radiation alone (30 Gy for 26 patients) or reduced radiation with chemotherapy (systemic vinblastine, bleomycin, etoposide, and cisplatin for 12 patients). All 38 patients received extended focal (including whole-ventricle) irradiation, and were followed to February 2008 with characterizing overall (OS) and relapse-free survival (RFS) rates. Variables associated with survival were evaluated by univariate Cox proportional-hazards regression.
Median follow-up was 62.4 months (range, 10.1–142.5 months). The total 5-year overall survival rate was 93.7%. The 5-year OS and RFS rate for patients receiving radiation only were 100% and 96.2%, respectively. For those receiving radiation plus chemotherapy were 83.3% and 91.7% (no statistical significance). There was no predictive factor that was significantly associated with OS or RFS rates, and chemotherapy had no significant effect on survivals but higher incidence of treatment-related toxicity.
Further decrease of radiation dose to 30 Gy with whole-ventricle irradiation is sufficient to treat selected intracranial germinoma. Use of wide-field irradiation or chemotherapy should be avoided as unnecessary. We recommend exploring further radiation volume and dose reductions to eliminate the risk of radiation-induced treatment complications. Systemic chemotherapy is only recommended for patients at high risk of recurrence or metastasis. Reduction of radiation doses to 30 Gy may be feasible, even without chemotherapy.
Germ cell tumor (GCT) in the basal ganglia is a curious neoplasm. Atypical clinical presentation and peculiar radiological findings are common. In the present study, we analyzed the clinical outcomes of basal ganglia GCTs. Furthermore, the diagnostic utility of positron emission tomography (PET) was analyzed.
The clinical data of 21 consecutive patients from one institution were retrospectively analyzed. Nineteen patients were male and two were female. The median age at diagnosis was 13 years (range, 7–19 years). Progressive hemiparesis was the most common symptom, followed by bulbar symptoms, psychiatric problems, symptoms related to the increased intracranial pressure (IICP), and hyperkinetic movements. [18F]-fluorodeoxyglucose (FDG) PET was performed in 12 patients. [11C]-methionine PET was performed in five patients (two for primary tumors and three for recurred metastatic tumors).
Nineteen tumors were germinomas and two were mixed GCTs. The median period of diagnostic delay was 7 months (1–31 months). [18F]-FDG PET showed a hypometabolic lesion in 11 patients and a hypermetabolic lesion in only one patient with a mixed GCT. In contrast, [11C]-methionine PET showed a hypermetabolic lesion in all patients. The two patients with a mixed GCT expired within 1 year from the diagnosis. The mean follow-up period of the 19 patients with a germinoma was 46 months (1–106 months): four patients (21%) had tumor progression and three patients (16%) expired at the last follow-up. The patients with poor outcomes had a significantly prolonged period of diagnostic delay (p=0.003). The hyperkinetic movement disappeared after treatment in all patients. The hemiparesis slightly improved in five patients through rehabilitation, but was unchanged or even deteriorated in 14 patients.
Timely diagnosis of basal ganglia GCTs could affect the clinical outcomes. [11C]-methionine PET may be helpful in the early diagnosis.
Fortunately, survival rates for children diagnosed with malignant brain tumors have improved with advances in surgery, radiation therapy, and chemotherapy. As a result, understanding the long-term impact of therapy including the neurocognitive and psychological effects of cancer treatment are important as most children who often present with various deficits. Significant sequelae, including intellectual decline, as well as variable attention-concentration, slow processing speed, and motor deficits have been observed as a result of treatment, with specific risk factors including young age at the time of treatment, higher dose and larger brain volume involved in radiotherapy, longer follow-up time since therapy, and treatment with methotrexate. In an effort to preserve intellectual functioning and quality of life, many current treatment studies have been utilizing reduced doses and volumes of radiotherapy, while others are employing postoperative chemotherapy to delay or avoid radiation therapy in these children. Two prominent multicenter intervention studies in the United States involve the use of methylphenidate (Ritalin) medication and the other utilizes a Cognitive Remediation Training Program in order to address symptoms such as attention, memory and academic achievement difficulties following treatment for pediatric cancer and have demonstrated modest success. The emotional effects of cancer upon a pediatric patient vary depending on the diagnosis, the location of the tumor and the treatment type, and on the patient’s family dynamics and preexisting psychological conditions. The vast majority of studies to date indicate that survivors of pediatric cancer are generally at no greater risk for long-term emotional sequelae than one would expect to see in a healthy population. It is important to note, however, that these studies often do not include brain tumor patients, mainly because of the physical effects that brain tumors and their treatments have on the CNS and the apparent difficulty in separating organic and psychological causes of discrepancies in patient QoL. Of those studies that do include brain tumor patients, approximately half concluded that these patients are at no greater risk than are the healthy controls, whereas the other half concluded that they are at greater risk. However, studies have found that pediatric brain tumor patients consistently score lower in the domains of psychosocial health and of physical health than do other, non-CNS tumor patients. Thus, while information about brain tumor patients’ QoL continues to emerge, there does appear to be an increased risk for deficit in certain domains of physical and psychosocial functioning. It will be important to provide counseling and support before and during the transition to adulthood with experiences such as employment, marriage, and parenthood.
I will discuss meeting the needs of the family member as a way of better caring at home for patients with brain tumors. We will go over interventions to improve caregiving at home as well.
Endocrine complications are the most frequent and clinically relevant among the late effects in childhood brain tumors. Hypothalamopituitary hormonal axis can be impaired by the neighboring brain tumor itself or by therapeutic modalities, especially radiation and surgical treatments. In many cases endocrinopathy can be predicted by the risk factors such as site and nature of the tumor and dosage of irradiation in the hypothalamic area. Clinicians should also be aware that it may develop several years or even a decade after the end of oncologic treatments. Endocrinopathy in childhood and adolescence often presents growth disturbance and/or abnormal pubertal development (i.e., precocious, delayed, or absence of puberty). Therefore, statural and pubertal monitoring as well as regular measurements of thyrotropin, thyroxin, and morning cortisol is essential during the long-term follow-up of the survivors at risk for endocrine complications. Treatments for endocrine disorders are usually effective and often helpful in improving quality of life (QOL). Children and adolescents, however, need induction of therapy at their pin-point timings corresponding to their specific developmental stages in order to achieve better final results. In this context early detection of endocrine disorders through the adequate monitoring and then involvement of pediatric endocrinologists at a suitable stage are mandatory.
Survival rates of children with brain tumor have significantly improved over the years due to developments in diagnostic techniques, neurosurgery, chemotherapy, radiotherapy, and supportive care. However, neoplasms occurring in the CNS remain an important cause of cancer-related deaths in children. This lecture presents the current problem in the medical care system for children with brain tumor in Japan and introduces some modern multidisciplinary programs for treatment of difficult tumors. In our country, due to a lack of registration system, there are no exact data concerning the incidence of pediatric brain tumor. At the time of initial diagnosis, the child is seen at a local community hospital where the knowledge of brain tumors is very small. Unfortunately, often the child does not get sent to the major treatment center that specializes in childhood brain tumors and the outcome for them is not sufficient. It has been emphasized that we do need to establish specific center hospitals and educational programs to encourage young pediatric neurooncologists in our country. Prognosis is still highly dependent on clinical characteristics, such as the age of the patient, histological type, stage, and localization of tumor. This uniqueness makes the treatment strategy complex in individual patients. However, increased knowledge about the long-term functional outcome of children with brain tumor and late effects of treatment is being obtained and has been useful to further improve outcome for patients. The process to establish the current treatment protocol in Japan will be introduced, focused on intracranial germ cell tumors.
Primary brain tumors include various histological types. For most types of tumors in most locations except deep-seated tumors such as brainstem gliomas, surgical resection is recommended. And their removal should be as complete as possible with preservation of neurological function. Total resection of intraaxial infiltrating tumors by surgery alone is generally impossible. Additional options of radiation therapy and/or chemotherapy vary according to histology and degrees of resection. Patients with incurable or unresectable tumors are considered as candidates for clinical trials.
The Pediatric Brain Tumor Foundation (PBTF) is honored to be here today. For the past 26 years, we have supported the search for the cause of and cure for childhood brain tumors. We want to find ways to improve those treatments, and we hope to do that with your help. A Mother’s Words: “We cling to hope, and hope lies in research.” That is why the PBTF is going to enormous lengths to fund research to save the lives of these precious children. The Numbers: I won’t burden you with too many numbers, but here are some important ones to remember about brain tumors: this year in the United States alone, the Central Brain Tumor Registry tells us, 3,750 children will be diagnosed with a primary brain tumor. That means that today, while we’re at this meeting, 10 families are getting the devastating news that their baby has a brain tumor. Even worse, 1,360 of the children diagnosed this year will die an early death. Forty percent of kids with a brain tumor will be dead within 5 years of their diagnosis. A Difficult Journey: The journey these kids face is daunting. To begin with, brain tumors are complicated to diagnose. The symptoms often mimic that of other childhood diseases, with nausea, headaches, and hearing and vision problems. These tumors are sometimes inoperable, and pathological diagnosis is sometimes difficult, especially with 126 different types of tumors in existence. And of course, once the tumor is diagnosed, the treatment options are not great. As you know, even if a brain tumor is benign, its treatment is not. Surgery is invasive and can damage vital areas of the brain that control neurological functions. Few drugs will cross the blood-brain barrier. Those that do can weaken the immune system and leave children susceptible to a host of other infections and possible hearing loss. Radiation to a developing brain can lead to long-term cognitive challenges and physical disabilities. The long-term effects on children and their families are devastating. We have only yet begun to see the consequences of cancer treatment as these children age. There is a fine line between a healthy life and one filled with daunting medical challenges. This is the line that children diagnosed with a brain tumor walk each day of their lives. Young brain tumor patients face countless challenges that are nearly overwhelming for them compared to physically healthy children. Furthermore, if they survive the rigors of this all-consuming disease, and the devastation of their treatment, they then embark upon a long and often immensely frustrating journey to adulthood. These children can experience an ongoing decline in mental and physical capabilities, which impacts their self-esteem, their ability to retain learned information, and their ability to engage in physical activities. Often they face other cancers brought on by their treatment. The things that come as second nature to other children can be monumental obstacles in their daily lives. The emotional and financial tolls on the families of brain tumor patients are also devastating. Jobs and homes are lost. Bankruptcies are not uncommon. Children require special long-term care by parents, but what happens when parents die? Who then cares for the patient? Parents’ or caregivers’ marriages sometimes break up under the strain. Siblings who feel ignored may have behavioral problems. Grieving takes place during and after the illness. Even if a child’s brain tumor is not fatal, there is loss with living. Research Challenges: All of these children, living and dead, are my heroes. They are the heart of the Pediatric Brain Tumor Foundation’s efforts. Until a cure is found, we will continue our search for new noninvasive and more effective treatments. We are working to overcome considerable research challenges: few centers of excellence; limited understanding of the disease; lack of tissue samples; widely varying response to treatment; insufficient funding for research; difficulties in attracting qualified clinicians and researchers; more oncologists need to be trained in this specialty. Funding Challenges: There are also many funding challenges. To begin with, there is a lack of funding for childhood diseases. A long-term commitment to funding is required to have an impact. And few organizations are wholly dedicated to childhood brain tumor research and family patient issues. In addition, low population numbers make it unattractive for pharmaceutical companies to invest in research funding and new drug development. Only one new drug has been developed for the treatment of pediatric brain tumors in the past 30 years. It is also difficult to achieve research collaborations, something that we are working very hard to do. The PBTF is one of the largest funding sources for childhood brain tumor research, but we can’t do it alone. The federal government’s help is needed to get pharmaceutical companies involved in research funding and drug development, not just in the United States but all over the world. Reasons for Hope: The work we are doing in the PBTF Institute labs is turning research into hope. When we began our work in 1984, long-term effects were not an issue because most children diagnosed with a brain tumor died. Today many brain tumor patients are living longer. We have vastly improved radiation delivery methods, meaning that fewer kids suffer the awful effects of whole-brain radiation. The U.S. government has made a stronger commitment to advocates in most major peer reviews over the past few years, thereby ensuring greater care and outcomes for the children. Our advocacy work in Washington, DC, and around the world, has led to increased visibility of the issue of childhood brain tumors. Thanks to the efforts of the PBTF and other interested parties, there are new research programs under way at universities where pediatric brain tumor research was never done before—we know because we’re receiving grant applications from more and more institutions each year. We have driven an increase in collaboration among researchers at the international level through our institutes, and we’re encouraged by the possibility of new biological treatments for brain tumors. The PBTF has raised more than $50 million in this fight so far. But we have a long way to go before we are finished. PBTF Research Funding: We fund cutting-edge medical research at dozens of the world’s leading institutions from coast to coast in the United States to Canada, Australia, Germany, and the United Kingdom. PBTF Institute Program: We believe that research collaborations are the answer to the fast-forwarding of brain tumor research. Researchers at the Pediatric Brain Tumor Foundation Institutes at Duke University, the University of California, San Francisco, and the Hospital for Sick Children in Canada are working together to share their findings as well as tissues and data. In May 2008 we announced an additional $6 million grant to our research institute at Duke University. Funding for Duke and our two other institutes now exceeds $14 million. We are greatly encouraged by the way the doctors and researchers at these institutions are working together. Fostering Global Collaboration: We are already seeing the fruits of these collaborations. We held our second annual PBTFI research conference last spring, and the excitement over the shared discoveries was uplifting. The researchers’ level of enthusiasm is the highest we’ve seen in 26 years of grant funding. These researchers, many of whom are here for this meeting in Yokohama, will get together again this October to share their successes. And that same month, the PBTF will sponsor the first-ever International Pediatric Basic and Translational Research Invitational Conference. We see increased interest in pediatric brain tumor research that we have never seen before, fostered by groups such as the International Brain Tumor Alliance, which is here today to help spread the understanding of brain tumors, including pediatrics, to all corners of the globe. Many, many basic and clinical researchers are working on new discoveries, including Mitch Berger and his team at UCSF, Darell Bigner’s group at Duke, and many others in the United States. Research is also occurring in other countries, led by Jim Rutka at Sick Kids Hospital in Toronto; Masayuki Sawamura, Masao Matsutani, S Nishikawa, and their colleagues here in Japan; Stefaan van Gool in Brussels; David Ashley in Australia; David Walker in Nottingham, England; Manfred Westphal in Germany; and many, many more. This tells us that we have much to be encouraged about regarding the future outcomes of childhood brain tumors. This type of scientific collaboration is a bright new source of hope for every child with a brain tumor, from Chitose to Chicago to Chelsea and on around the world. Partners in Hope: We also support researchers by funding data collection and a variety of professional meetings, including the Central Brain Tumor Registry of the United States, for which we provided the founding grant; the World Federation of Neuro-Oncology; the Society of Neuro-Oncology, including the founding grant for the prestigious journal Neuro-Oncology, which is the official neuro-oncology journal of the United States, Europe, and Japan. We also sponsor the Society of Neuro-Oncology Foundation, fund its Basic Research Award for Excellence, and provide grants to support its annual meetings; Brain Tumor Epidemiology Consortium meetings; International Pediatric Neuro-Oncology Symposium; International Symposium on Brain Tumor Research and Therapy; Peter Steck Memorial Research Award and Lecture; and Childhood Neurology Society. PBTF/Japan Partnerships: Our partnership with Japan is strong. We have sponsored medical conferences here over the years, and we also gave the Children’s Cancer Association of Japan a grant to bring a social worker to the United States for cross-cultural training. PBTF Family Support: In addition to research, the Pediatric Brain Tumor Foundation believes that it is critical to offer emotional and educational support to families in their time of need. Our programs include educational resources, patient services, college scholarships, advocacy efforts, and the Ride for Kids program. We strongly encourage you to share information about our services. We primarily serve the United States, but our materials are also printed in Spanish and we offer resources to people worldwide. Online Services: Our Web site, www.pbtfus.org, gives families a great number of resources and easy access to information. The site also includes new survivor outreach information. Our family support program includes a social worker who is available to assist patient families by phone or e-mail. Here in Japan, Dr. Sawamura at Hokkaido University runs a Japanese-language Web site, Directory Database of Research and Development Activities (ReaD), with information for patient families (e-mail the ReaD office at pj.og.tsj.oykot@daer to gain access to the ReaD Web site). The Children’s Cancer Association of Japan is also an excellent resource for patient families. Advocacy in Action: The Pediatric Brain Tumor Foundation also believes that raising awareness of this disease creates hope. As an active member of the brain tumor advocacy community, the PBTF is a founding member of the Alliance for Childhood Cancer, patient advocate on the National Cancer Institute’s (NCI) Specialized Program of Research Excellence (SPORE) for brain tumors, representative on the Patient Advocates Research Team (PART) Council for the NCI’s SPORE in cancer, and patient advocate, National Institute of Neurological Disorders and Stroke (NINDS). We also make congressional visits to educate U.S. lawmakers on childhood cancer issues. We invite you to join us in advocating for quality care in each of your countries for pediatric and adolescent brain tumor patients, development of new pediatric oncology drugs, long-term follow-up care for survivors, and representation of patients’ needs and viewpoints when research efforts are translated into clinical trials. Fundraising Efforts: Research is incredibly costly and requires million of dollars to fund; we hope you will join in the effort to raise the funds needed in your respective countries for pediatric brain tumor discoveries. One of our fundraising efforts is our Ride for Kids motorcyclists’ charity program. It is the backbone of our medical research and family support program funding. In 1984, the first year of our work, we raised $4,000. Last year PBTF Ride for Kids programs in 37 cities resulted in more than $5.2 million in donations to help cure the kids, and we’re on track to raise even more with 39 events in 2009. This program is also an important part of our family support efforts, because it brings together the community of brain tumor survivors and their families. Our Celebration of Life program following each ride is just that: a celebration of the lives and accomplishments of these children. In addition to Ride for Kids, we raise awareness and funds through: direct mail, which helps us appeal to individuals who support similar charitable causes; a radiothon program, which lets us partner with radio stations to broadcast patient stories and solicit donations from the public. We also have thousands of people around the world who raise funds for us in many original ways. Some hold walkathons, others run marathons. We’ve even had schoolchildren send us their pennies. Every cent makes a difference in our fight against childhood brain tumors. Faces of Hope: One of the most visible signs of hope is that brain tumor survivor-ship is increasing. For instance, when we began raising funds in 1984, we did not need scholarships for brain tumor survivors. Now more children are living longer, and making plans to attend college. However, a brain tumor places an enormous financial burden on a family. The PBTF helps make higher education possible by offering scholarships. Here are a few of the “faces of hope” that we are helping send to college. To date we have awarded 397 scholarships to brain tumor survivors, and the program is growing dramatically with the addition of the Tim and Tom Gullikson Family Support Fund. How’s that for hope? Here is a perfect example. Michelle Higa was our first scholarship recipient to graduate from college. She is now working on her master’s degree in psychology. Like so many young adults who had a pediatric brain tumor, Michelle wants to help children going through medical crises. Celebrating life is what the Pediatric Brain Tumor Foundation is all about. We celebrate the lives of those children who are no longer with us, as well as the lives of the young people who continue to suffer, by working harder each day to find a cure for this dread disease. Thank you for your time today, and for everything you are doing to help us in our fight. Please visit our Web site, www.pbtfus.org, to learn more about how you can help us turn research into hope.
There are many challenges facing the brain tumor patient. Brain tumors attack the very essence of who we are. Brain tumors affect the physical, emotional and cognitive capabilities of a person, thus impacting quality and length of life, employment, and personal relationships. They strike people of any age, from tiny babies to the elderly, male or female, in any geographic locality. Based on current knowledge, brain tumors cannot be prevented or realistically screened for. Their causes are, as yet, largely unknown. Beyond the patient’s own personal situation, and across Europe and beyond, there is a desperate lack of funding for research into more effective therapies for brain tumors. Although some progress in treatment has been made, there can be inequity in accessing promising new therapies. There are sometimes misdiagnoses and late diagnoses. There are not enough specialists and specialist centers. There is a lack of consistency in registries from country to country. For the caregiver of someone with a brain tumor, the challenges are different but also extremely daunting. A woman caring for a relative with a brain tumor said: “It can be hard to put someone else first all the time—to let someone go but be there for them still. It’s like standing on the shore watching (him) out at sea. Sometimes the waves bring him back closer to me and sometimes the tide takes him away again.” In this difficult setting, brain tumor support, awareness-raising, advocacy and fund-raising are crucial. The European brain tumor patient and caregiver group community is rising to these challenges and achieving progress through various activities. In some European countries—though by no means all—brain tumor foundations and groups, where they exist, are responding to a desperate need by energetically raising money for research. They offer support and information and, very importantly, advocate with determination for a better, more equitable deal for patients and caregivers. Through its two primary activities—the promotion of the International Brain Tumor Awareness Week (November 1–7, 2009) and the Walk Around the World for Brain Tumors (January 1–November 7, 2009)—the International Brain Tumor Alliance (IBTA) unites patient groups in Europe (and other parts of the world) in a global and very public effort to raise funds and awareness of this illness. More than 30 European patient organizations relevant to brain tumors supported these two projects in 2008. Around the world, more than 150 patient and professional groups and organizations were supporters. Through the achievements of such initiatives, across Europe and beyond, the situation for brain tumor patients and their caregivers is improving, giving hope to those whose lives are touched by this devastating disease.
“Why my child?” is the first reaction any parent will have when told of their child’s illness. In 1968, our foundation, Children’s Cancer Association of Japan, first established in Japan to support families with children afflicted with childhood cancer. Since then, the social workers at CCAJ have received countless phone consultations. Fears, anxieties, and uncertainties expressed by parents are constant and have not changed since its inception. “Is the hospital we are going to trustworthy?” “Aren’t there alternative treatments?” “We are worried about our financial situation.” “What is the cure rate of this illness?” “How could we explain the disease to our child?” “Who can look after the siblings?” “What will happen to his/her studies while in treatment?” “We have to make a decision on the child’s treatment options and we don’t know which is best.” “We want to meet others who have the same illness.” Those are but some of the information they want. In recent years, the number of consultations regarding childhood brain tumors have been alarmingly increasing and the contents of questions varied and urgent. Childhood leukemia, which has been the main concern in the medical field, has seen not only the remarkable improvement in the cure rate, but also due to the acceptance in the concept of total care involving professionals, it has become possible for the leukemia patient to receive a variety of social support from the time the child is admitted to a hospital. On the other hand, the dissemination of information and support system has not been well developed for childhood brain tumors despite the fact that childhood brain and spinal cancers have the second highest morbidity rate among childhood cancers. The reasons are their numerous varieties, existence of different treatments and prognosis depending on the location and the age of patients, poor prognosis, and complexity of communication due to treatments involving multiple fields of medicine. Even though the methods of controlling the disease have much improved due to progress of research and treatment methods, often patients are exposed to late-effects, have to face drastic change in their lifestyle, and overcome huge obstacles trying to be socially independent. In addition to the above, the proliferation of conflicting information given on Internet and lack of appropriate information, and shock of diagnosis and huge problems that the families face in the treatment process are the main reasons we see the increase in the consulting cases. Children’s Cancer Association of Japan was founded in 1968 by parents whose children died of cancer. The members of CCAJ are striving to help families and children afflicted with cancer alleviate difficulties and worries. We also help improve their quality of life by coordinating support from medical fields and other supporting institutions and individuals. CCAJ now has 19 associate organizations nationwide and membership of more than 3,000 and will celebrate its 41st anniversary this year. The main activities are as follows: (1) funding for medical research projects for doctors, nurses, and other institutions on the subjects of treatment and/or care of childhood cancers; (2) consultations by social workers, assisting different groups such as different cancer subgroups, parents who lost their children to cancer, cancer survivors, siblings of cancer patients, coordinating and educating volunteers and planning workshops; (3) supporting families financially for treatments, and management of lodging facilities; (4) dissemination of information by publishing periodicals, newsletters, informational booklets. We also hold publicity activities to advocate the disease. Even if the medical environment is well developed and support system established, it is important that interaction and sharing of information among families continue. Also vitally important is the cooperation of medical professionals and the patients’ families. Groups such as Japan Brain Tumor Alliance and Japan Pediatric Brain Tumor Network involving all brain tumor patients have been established and expanding their activities. Our mission is to spread our activities internationally together with above mentioned groups, and not only strive for the patients and the families to be able to receive quality treatment but work hard to improve our supporting role including enrichment of their daily lives after the treatment.
We performed a comprehensive genetic profiling in a large series of astrocytic and oligodendroglial tumors to identify specific molecular signatures that could help define biologically relevant subtypes of human gliomas. We studied 402 tumors (38 pilocytic astrocytomas [PA], 22 diffuse astrocytomas [A], 62 anaplastic astrocytomas [AA], 183 glioblastomas [173 primary GB [pGB], 10 secondary GB [sGB]), 34 oligodendrogliomas [O], 20 anaplastic oligodendrogliomas [AO], 20 oligoastrocytomas [OA], 23 anaplastic oligoastrocytomas [AOA]). All tumors were examined by 1-Mb array-CGH. TP53, CDKN2A/p14ARF, RB1, PTEN, and IDH1 were investigated for mutation. Total 1p/19q loss were predominantly found among O and AOs but rare in astrocytic tumors. TP53 mutations were common among A, AA, and sGB but less common in OA, AOA, and pGB and rare among O and AO. TP53 and total 1p/19q deletions were mutually exclusive. IDH1 mutations were frequently found among all oligodendroglial tumors, A, AA, and sGBs but rare in pGBs and not found in PAs. IDH1 mutations were strongly correlated with either TP53 mutation or total 1p/19q loss. RB1 pathway alterations were exclusively found among grade III or IV tumors, and almost always associated with p53 pathway alterations. PTEN mutations or EGFR amplifications were exclusively found among AAs and GBs. The data suggest that IDH1 mutation is the earliest change in both oligodendroglial tumors and adult astrocytic tumors with the exception of pGBs. IDH1 mutations followed by total 1p/19q loss may characterize oligodendrogliomas while IDH1 mutations followed by TP53 mutations may characterize astrocytomas. RB1 pathway alterations are associated with malignant progression in all types. pGB develop through distinct pathways by acquiring concurrent alterations of the RB1 and p53 pathway, with or without PTEN mutations or EGFR amplification but not IDH1. Thus, we demonstrate that sequential acquisition of particular genetic abnormalities may define distinct subtypes of gliomas.
Glioblastomas (GBM) are highly malignant and heterogeneous gliomas with very poor prognosis. The biological and molecular characterization of these tumors is still challenging and impacts their therapeutic management. Previous genomic surveys have revealed the highly rearranged nature of GBM genome and transcriptome. However, the impacts of tumor DNA aberrations on gene expression remain unclear.
We investigated copy number alterations (CNA) and gene expression to identify causal genetic events in GBM. High-resolution maps of somatic chromosomal alterations were obtained for 20 GBM. Gene expression profiling was carried out on the same tumor samples, and compared to those obtained on nonneoplastic brain samples. Concordance between CNA and gene expression was identified by two complementary approaches (correlated or targeted probes). The resulting GBM signature was validated with an independent microarray data set of 81 GBM and 23 normal brains.
Loci targeted for high-priority minimal common regions (MCR) of recurrent CNA were defined and combined with gene expression profiles performed on the same tumor samples. Genes with concordant changes in CNA and expression levels were defined as over/under-expressed genes located in amplified/deleted regions, or as MCR genes with expression correlated to the corresponding genomic state. After validation, we found that the expression of 318 genes was significantly affected by CNA. Associated enriched GO process annotations were related to cell cycle disorder, cellular adhesion and angiogenesis. The gene signature included well-known GBM genes such as EGFR, PDGFA, and p16INK4 but also novel candidate genes. Two tumor suppressor genes PCDH9 and STARD13, involved in tumor invasiveness and resistance to etoposide, were validated by qPCR in an independent set of 57 glioblastoma.
This study shows the power of combining genomic alterations and gene expression to identify robust transcriptome signature and putative tumor biomarkers in GBM.
The influence of inherited genetics on outcome of therapy in malignant brain tumors remains poorly understood. In this study, using a linkage disequilibrium (LD)-based approach, we examined the relationship between genetic diversity in 45 DNA repair genes and survival in 301 adults with anaplastic astrocytoma (AA) and glioblastoma multiforme (GBM) treated on temozolomide and carmustine-based protocols. We genotyped a subset of 1,267 tagging SNPs within or near 45 genes in high LD with a region of 20 kb flanking each gene, on a customized Affymetrix platform. Associations were performed after testing for Hardy-Weinberg conformity and gender mismatches. The results identified tSNPs in and around EXO1, DDB2, LIG1, MGMT, MSH2, MSH4, PAN3, RAD52, TDG, XRCC1, and XRCC5 to be highly associated with survival. Although, several gene variants were associative in both AA and GBM, a subset of SNPs were associative in only one category. The identification of highly polymorphic regions of DNA repair genes in association with outcome of temozolomide/ carmustine therapy in malignant gliomas is consistent with the role of DNA repair in mechanisms of tumor resistance to these agents. Resequencing is ongoing to establish the specific causative SNPs and to examine the biology underlying their association with treatment outcome in malignant gliomas.
Supported by grants RO1 CA127872, RO1 CA 112519, P50CA108786, and P30CA14236 from the NIH and the Pediatric Brain Tumor Foundation of the United States.
Identification of GBM subgroups might improve an aimed individualized therapeutical approach. Consequently, in our study we looked at patients with early versus late tumor recurrences in order to identify molecular markers, which might be associated with those different growth patterns. These markers, in turn, could be new therapeutic aims.
Study population comprised 10 pairs each of primary and recurrent tumors occurring within 180 days after initial diagnosis (short-term recurrences, STR) and later than 300 days after initial diagnosis (long-term recurrences, LTR), respectively. Exons 5–8 of the p53 gene and the entire coding region of PTEN were investigated by sequence analysis. EGFR amplification was determined. Loss of heterozygosity (LOH) analysis of markers known to be involved in glioma pathogenesis was performed. A chip-based expression analysis was performed in six samples (three STR and three LTR) using Affymetrix, GeneChip Human genome U133 Plus 2.0. Two candidate genes were further evaluated by real-time PCR-based mRNA expression analysis.
p53 mutations are less common in STR than in LTR, while genetic variants of the PTEN gene are more frequent in STR. In STR, p53, and PTEN variants occur in combination, while LTR harbor either p53 or PTEN variants. Clustering after chip analysis revealed clear differentiation between STR and LTR. Construction of a six-gene-comprising predictor is possible. A set of 53 differentially expressed genes was identified. GAS6 and SNTG turned out to be promising new candidates for determination of early recurrence.
STR and LTR can be distinguished by p53 and PTEN mutation status and by gene expression profile. Out of a set of 53 differentially expressed genes, GAS6 and SNTG appeared to be suitable markers for early recurrence and potential therapeutic approaches. Of course, our study sample is rather small, and data need further confirmation.
Unlike select examples in oncology such as imatinib for CML, single-agent signal transduction modulators have been ineffective in most cancers including glioblastoma. Combinations of agents based on established signaling pathways may be more effective. However, performing individual phase I/II trials for each combination is inefficient. We developed a three-arm phase I/II trial using a sequential accrual design combining sorafenib with erlotinib, temsirolimus, or tipifarnib.
The North American Brain Tumor Consortium (NABTC) conducted a phase I studies of sorafenib (VEGFR/PDGFR/Raf inhibitor) in combination with erlotinib (EGFR inhibitor), temsirolimus (mTOR inhibitor), or tipifarnib (farnesyltransferase inhibitor) in recurrent GBM. Eligibility criteria included histologically proven GBM, radiologic progression, ≥18 years old, KPS≥60, adequate bone marrow reserve and organ function. There was no limit on the number of prior therapies. No enzyme-inducing antiepileptic drugs were allowed. Dose-finding used a standard 3+3 design with the MTD was defined as the dose with DLTs in 1/6 or fewer patients. Serum pharmacokinetic (PK) studies were performed. Accrual was sequential, decreasing study pauses for maturation of toxicity data.
Overall 49 patients were enrolled. Based on 17 patients, the MTD was sorafenib 400 mg BID and erlotinib 100 mg qD. Based on 13 patients, the MTD was temsirolimus 25 mg qweek and sorafenib 400 mg BID. The MTD was sorafenib 200 mg BID and tipifarnib 100 mg qD×21 days based on 19 patients, but the dose of tipifarnib is below therapeutic levels. PK studies showed no drug-drug interactions for sorafenib with tipifarnib or temsirolimus. However, there was no accumulation of erlotinib, suggesting a drug-drug interaction with sorafenib altering erlotinib metabolism or clearance.
This study demonstrated that the sequential accrual design accelerates study completion and decreases logistical problems of phase I studies in multicenter trials. Phase II doses were successfully determined for sorafenib with temsirolimus and with erlotinib, permitting enrollment onto the phase II component. The new phase I study using an alternating week schedule of tipifarnib will be performed to permit a daily dose more likely to reach therapeutic concentration.
Many end points have been used in the context of clinical trials in neuro-oncology including objective response, overall survival, median time to tumor progression, progression-free survival and other landmark end points, such as PFS-6 months or PFS-12 months. End point determination is critical to assess the efficacy of drugs or treatment strategies in both phase 2 and phase 3 trials. There are pros and cons with each of these end points, and the challenge remains as to whether one or another should be used in a given situation, and whether there is acceptance across the neuro-oncology community, regulatory agencies or industry sponsors. Given the costs of conducting trials, and the impact of end points on patient accrual and expectations, protocol design has become more and more relevant, particularly with the advent of novel agents and strategies that may influence our interpretation of MR-based imaging that is critical for end points that use response or progression as major determinants of futility or success of treatment. We will present examples of the use of progression-free survival as one potential end point in both newly diagnosed and recurrent disease, and the implication of landmark PFS as a potential valid end point for phase 2 trials.
Recent data suggest that the glutamatergic system is important in the proliferation and migration of glioblastoma cells. Talampanel is a well-tolerated oral AMPA receptor blocker that could be beneficial when added to conventional glioblastoma therapy.
The primary purpose of this safety and activity (phase II) trial was to estimate overall survival in adults with newly diagnosed glioblastoma treated with talampanel in addition to standard radiation and temozolomide (RT+TMZ). A secondary purpose was to evaluate talampanel toxicity in this setting. Talampanel was initiated with RT+TMZ and discontinued for toxicity or disease progression.
Survival results were compared to historical controls. Seventy-two patients were enrolled from December 2005 to July 2006. Their median age was 60 years (range, 37–85), median KPS was 90 (70–100), and 77% had a debulking procedure. With a minimum of 18 months follow-up, 42 patients (58%) have died yielding a median survival of 18.3 months (95% CI=14.6–22.5 months). When the 60 patients who were 18–70 years old on this trial were compared with the EORTC (RT+TMZ) data, the median survival (20.3 vs. 14.6 months) and percent surviving at 24 months (41.7% vs. 26.5%, p=0.02) appeared superior. Data on MGMT methylation and postprogression treatment with VEGF-targeted therapies for this population will be available for presentation.
Talampanel was well tolerated and did not appear to increase the known hematologic or nonhematologic toxicities of TMZ. Talampanel can be added to RT+TMZ without significant added toxicity. These encouraging survival results in this study suggest that blocking AMPA receptors may be a useful strategy in glioblastoma.
Because of the proposed sensitivity to chemotherapy of oligodendroglial tumors both the RTOG and EORTC had investigated if these tumors benefit from adjuvant PCV chemotherapy. These studies, EORTC study 26951 and RTOG 9402, both showed that the addition of PCV chemotherapy (consisting of procarbazine, CCNU, and vincristine) to 59.4 Gy radiotherapy does increase progression-free survival without improving overall survival in anaplastic oligodendroglioma and anaplastic oligoastrocytoma. A major finding of both studies is the large difference in prognosis of patients with and without combined 1p/19q loss. Based on these differences in survival and the clear different outcome in anaplastic oligodendroglioma with 1p/19q loss, EORTC and the collaborative groups felt that it was no longer rational to treat these patients according to histology without taking the genotype of these tumors into account. For studies in anaplastic gliomas it was therefore proposed to classify into anaplastic glioma without 1p/19q loss and anaplastic oligodendroglial tumors with 1p and 19q codeletion. Another challenge is the definition of a proper end point for these trials. Overall survival seems to be the most relevant outcome parameter, even at progression. The outline and initiatives in grade III gliomas (EORTC 26053/22054, CATNON plus the codeleted trial) are presented. Standard therapy for glioblastoma is surgical resection aimed to be as complete as possible, respecting neurological function followed by chemoirradiation with temozolomide. TMZ given as concomitant and adjuvant therapy to RT has shown to increase progression-free survival (PFS) (rate at 6 months, 53.9% vs. 36.4%) and median survival (14.6 vs. 12.1 months) compared to adjuvant treatment with RT therapy only (EORTC 26981/22981 NCIC CE.3 trial). Still, many patients do not respond to therapy. The resistance of cells against DNA damage caused by nitrosoureas and temozolomide is at least in part mediated by the DNA-repair enzyme O6-methylguanine-DNA methyltransferase (MGMT). Epigenetic silencing of the MGMT gene by promoter methylation compromises this DNA repair and has been associated with longer survival in (glioblastoma) patients who are treated with alkylating or methylating agents. An analysis of the EORTC 26981/22981 NCIC CE.3 trial showed, that indeed patients with glioblastoma containing a hypermethylated MGMT promoter benefited from TMZ (overall survival [OS] rate at 24 months, 46% vs. 23%), whereas those who did not have a methylated MGMT promoter did have a significantly worse survival rate and less benefit from the addition of temozolomide to RT (OS rate at 24 months, 14% vs.<2%). This raises the question if the small benefit from chemoirradiation observed in this group outweighs the toxicity and costs of the temozolomide treatment, and calls for the development of more effective drug regimens for this specific group of patients. Although there may be small numbers of patients with an unmethylated MGMT promoter that do benefit from combined chemoirradiation, for the entire subgroup of these molecularly defined GBM patients the overall benefit is questionable. Most interestingly, the phase II trial with the integrin inhibitor cilengitide also demonstrated a marked benefit mainly in the patients with glioblastoma containing a methylated MGMT promoter. Consequently, the current Merck/EORTC phase III trial is designated to delineate the role for cilengitide in glioblastoma with methylated MGMT. Even earlier, Eli Lilly took the approach to examine the protein kinase C-β inhibitor, enzastaurin, together with radiotherapy but without TMZ in patients with glioblastoma containing an unmethylated MGMT promoter. This raises the general question whether treatment in glioblastoma trials should not only be stratified according to MGMT but entry into those trials limited by MGMT status. This would call for different approaches of GBM patients, depending on the MGMT promoter gene status. The primary question to address in GBM with unmethylated MGMT promoter gene is the identification of drugs that provide more survival benefit compared to TMZ. The current EORTC trial initiatives are presented.
Japan Clinical Oncology Group (JCOG) Brain Tumor Study Group conducted a multiinstitutional randomized controlled trial on malignant gliomas entitled, a randomized controlled phase II study of chemoradiotherapy using ACNU versus procarbazine and ACNU for astrocytoma grade 3 and 4, with the support of the Health and Labour Sciences Research Grants of the Ministry of Health, Labour, and Welfare in order to establish a standard therapy for malignant gliomas in Japan.
The patients with newly diagnosed supratentorial astrocytoma grade 3 or 4 were enrolled and randomized into two groups. The patients in group A were treated with ACNU (80 mg/m2 iv) during the postoperative radiotherapy (RT, 60 Gy local), while those in group B received procarbazine (80 mg/m2 for 10 days per os) preceding administration of ACNU. Each regimen was continued every 8 weeks for 2 years if it was tolerable for the patients and their disease did not progress. The primary end point was the overall survival rate and the secondary end points were the response rate on the MRI and the frequency of the adverse events. Procarbazine is expected to reduce O6-methylguanine-DNA methyltransferase (MGMT) and enhance the anticancer activity of nitrosoureas. The protocol was activated in April 2004 and 111 patients were registered by the end of August 2006 from 19 collaborating neurosurgical institutes of JCOG-BTSG.
The overall survival of the patients treated with ACNU+RT was 16.2 months and that of procarbazine+ACNU+RT was 18.7 months, while PFS of both groups were 6 months. CTCAE grade 3/4 was observed in 40–60% of the patients.
ACNU-based chemoradiotherapy was an effective but toxic treatment.
There has been no qualified sponsor-investigator clinical trial program, and the standard therapies have been all we could do for the treatment of malignant glioma patients in the Korean Brain Tumor Society. We have just started to join two international clinical trials since 2008. In this article the past and current status of the neuro-oncology field in Korea as well as eastern and northern Asian countries will be introduced, and clinical outcomes of concurrent radiotherapy and temozolomide chemotherapy for 100 patients of four university hospitals of Korea (Advisory Board of S-P Korea) will be presented.
A recently developed histogram analysis of relative cerebral blood volume (rCBV) from the entire tumor has been reported to offer excellent interobserver agreement for quantitative analysis and demonstrate the heterogeneous morphologic features of glioma vascularity. We aimed to determine whether histogram analysis can be adopted in the assessment of tumor response during glioma therapy.
We retrospectively studied 51 dynamic susceptibility contrast 3-T MR imaging data of 29 patients (mean age 50.5 years, range, 18–76) with histologically confirmed gliomas (9 low grade, 20 high grade). rCBV maps were created and normalized to unaffected white matter. Histogram width (HW), peak height position (PHP), and maximum value (MV) of the entire tumor were measured from normalized histogram distribution.
The values (mean±SD) of HW, PHP, and MV were 4.64±2.03, 4.58±2.63, and 6.29±2.79 for the preoperative imaging of high-grade gliomas (n=8), and 3.83±1.96, 2.66±1.66, and 4.73±1.96 for the final imaging, which showed definite radiological tumor progression or confirmed tumor recurrence by biopsy (n=8). Thirty-two imaging data obtained during the median imaging follow-up of 3.7 months were divided into two groups (progression vs. stable/radiation necrosis) according to the follow-up result, and three parameters were compared. All three parameters were positively correlated with tumor progression (HW, 3.05±2.18 vs. 1.02±0.50; PHP, 2.39±1.71 vs. 0.94±0.28; MV, 4.13±2.83 vs. 1.56±0.52) and MV was the most predictive with multivariate analysis.
Our results suggest that histogram analysis of rCBV can be a more objective and useful diagnostic tool to determine tumor response during the treatment of gliomas, compared with region of interest method for maximum rCBV, which can be highly operator dependent.
Tumor progression of glioblastomas entirely depends upon the radiologically contrast-enhancement on MR images. Disease progression can be difficult to interpret because it looks like pseudoprogression. The aim of this study was to study the real incidence of pseudo-progression during/after concomitant chemoradiotherapy (CCRT) for newly diagnosed glioblastomas and the predictability of dynamic susceptibility-weighted contrast-enhanced perfusion MR scans for the pseudoprogression/true progression.
This present study was a single arm, prospective cohort study designed to evaluate the incidence and outcome of pseudoprogression during CCRT followed by adjuvant temozolomide (TMZ) treatment for newly diagnosed glioblastoma.
Between July 2004 and February 2008, 76 consecutive patients (median age 50 years) who underwent CCRT followed by adjuvant TMZ treatment were enrolled in this study. MGMT promoter was methylated in 26 patients (34%) and unmethylated in 50 patients (66%). Enlargement of lesions, evidenced at the MRI scan in 59 of 76 patients (77.6%), was subsequently classified as pseudoprogression in 31 patients (52.5%) and true progression in 28 patients (47.5%). Of these 31 patients, 18 patients developed pseudoprogression within the period of CCRT (hyperacute phase) and 13 patients during the adjuvant TMZ treatment (early phase). Lower value less than 1.75 on the perfusion images could be predictive of pseudoprogression with the sensitivity of 91% and the specificity of 82%.
The present study supports high incidence of pseudoprogression during the treatment for newly diagnosed glioblastomas. Perfusion MR imaging can be used to predict the possibility of pseudoprogression in the treatment of glioblastomas.
Retinoic acid derivatives exert antitumor efficacy by inhibiting cell proliferation, inducing cell differentiation, apoptosis, and growth arrest. All-trans-retinoic acid (ATRA), one of retinoic acid derivates, has been extensively used for the treatment of acute promyelocytic leukemia (APL). Am80, a novel synthetic retinoic acid known to be more potent than ATRA, is allowed only for ATRA resistant APL patients.
We evaluated the antitumor efficacy of CED of micelle-Am80 against the rodent intracranial U87MG xenograft model in conjunction with systemic TMZ administration.
Cytotoxicity of Am80 for U87MG cell line was examined in vitro by MTS assay. Cells were incubated with 50, 100, 200, and 500 μM of Am80 for 2, 4, and 6 days. The expression level of phospho-Akt and phospho-MAPK in the cells treated by Am-80 were evaluated by Western blotting. Toxicity of micelle-Am80 after CED was evaluated after infusion into the rat brain striatum. The antitumor efficacy was also evaluated using rodent intracranial U87MG xenograft model.
Am80 at 50 μM revealed no cytotoxicity against U87MG cell. However, cell viability decreased as concentration increases from 100 μM to 500 μM. Enhanced cytotoxicity was observed when Am80 was combined with TMZ. Western blotting revealed decreased expression of phospho-Akt and phospho-MAPK in cells treated with Am80. When cells were treated with a combination of Am80 and TMZ, further decrease of these phosphorylated proteins were observed. In vivo toxicity study revealed no toxicity with CED of Am80 in rodent striatum. Survival of U87MG intracranial xenografted rodents treated with CED of micelle-Am80 combined with systemic administration of TMZ was significantly longer than those treated with CED of Am80 or systemic TMZ as well as control rats (P<65308;0.01).
CED of micelle-Am80 synergistically enhanced the antitumor efficacy of systemic TMZ treatment.
Convection-enhanced delivery has been used with limited success in the treatment of malignant gliomas. Although direct delivery of biological agents to malignant glioma patients has been performed safely with good volume of distribution, increased survival has not been demonstrated. One reason for this lack of success is in the limitation of current convection enhanced delivery techniques, as multiple externalized catheters are inserted into the tumor for limited time periods (i.e., 5–7) days and removed; repeat infusions require another surgery for catheter placement. For the past 2 years, we have been designing and performing proof of principle experiments on an implantable magnetic breather pump system for malignant gliomas. The system has five components: (1) an intratumoral delivery catheter (for unresectable tumors) or an intracranial pouch (for tumor cavity in resected tumors), (2) a reservoir for tapping the catheter or intracranial pouch, (3) a connecting catheter to link the reservoir, (4) an internalized pump with canisters containing biological agents of choice, capable of delivering drugs at microliters/hour, (5) a lab-on-a-chip to sample tumor fluid to measure specific cytokines (i.e., vascular endothelial growth factor) or drug levels to monitor treatment progress. The magnetic breather pump will be implanted much like a shunt, with the delivery pump secured at the chest wall. The canisters in the delivery pump may be accessed via subcutaneous ports to “refill” the drugs, and can be wire-lessly reprogrammed to deliver at specific time intervals and different cycles using Medical Implant Communication Service (MICS) interface. We are currently developing a “pathfinder” with all five components connected to demonstrate proof of principle in vitro. If the “pathfinder” is successful in vitro, it will be reduced to an implantable system, and tested in vivo in animal models. Successful completion of animal studies will result in human phase I studies.
Immunostimulating oligodeoxynucleotides with CpG motifs (CpG-ODN) have shown a promising efficacy in cancer models. Intratumoral infusions of CpG-ODN in glioblastoma patients were well tolerated at doses up to 20 mg in a phase I clinical trial. A phase II trial was designed to study the efficacy of a local treatment by CpG-ODN in patients with recurrent glioblastomas (GBM).
Patients with recurrent GBM occurring at least 3 months after radiotherapy and previously treated with chemotherapy received 20 mg CpG-ODN (CpG-28) by convection-enhanced delivery. The primary end point was the percentage of the patients without tumor progression at 6 months after inclusion. Secondary end points were tolerance, survival, and radiological response.
Thirty-four patients were enrolled in two centers between November 2004 and March 2006. Thirty-one patients received the treatment. The 6-month progression-free survival (PFS) was 19%. One partial response and three minor responses were seen. The median overall survival was 28 weeks. Eight patients (24%) were alive after 1 year and five patients (15%) were alive 2 years after inclusion. Treatment was usually well tolerated. As previously reported, the most common toxicities were lymphopenia, mild fever, seizures, and transient neurological worsening.
Despite a few cases of radiological responses, CpG-28 showed a modest activity on the 6-month PFS in this population. However, the presence of long-term survivors suggests that this treatment might have been beneficial in some patients, whose molecular or clinical characteristics remain to be defined.
High-grade gliomas (HGGs) are malignant brain tumors for which no cure is available. To identify new therapeutic targets we performed an extensive mutation analysis. We sequenced 39 genes in 113 HGG (including 109 glioblastoma) tumor samples and 16 HGG cell lines. We analyzed 200 exons belonging to 35 kinases, isocitrate dehydrogenase 1 (IDH1), NRAS, PTEN, and TP53 genes. At least one somatic mutation was found in 90 out of 129 HGG samples (70%). Most mutations were observed in genes belonging to the PI3K-AKT pathway; in more than 45% of HGGs we detected a mutational activation of the pathway, indicating that the PI3K-AKT axis represents a relevant therapeutic target for HGG. In addition, five different types of somatic mutations affecting the IDH1-R132 residue were detected in 20% (23 of 113) of HGG tumor samples. IDH1 mutations were predominantly observed in secondary glioblastoma (11 of 94 vs. 11 of 15, p-value=0.0000016). It has been established that IDH1 mutations occur early in gliomagenesis and that the overall survival in both primary and secondary glioblastoma is better in patients with IDH1 mutations. However, it remains to be determined whether the mutations functionally activate or inactivate the IDH1 enzymatic activity. To address this issue we performed metabolic mapping (enzyme histochemistry) on wild-type and mutated HGG samples. Our initial results suggest that IDH1 mutations are inactivating. These data have relevant implications to exploit IDH1 mutations for diagnostic, prognostic, and therapeutic purposes.
Microregional hypoxia is a well-known characteristic of glioblastoma that appears as the tumor mass outgrows the existing vascular supply. Hypoxic areas of solid tumors are resistant to traditional chemo- and radiotherapies, highlighting the need to develop new therapies for their targeting. A common property of these cells is the expression of HIF-1, a key transcription factor that coordinates adaptive responses to hypoxia, providing cancer cells the means to survive and proliferate under hypoxic conditions. To identify specific small molecule inhibitors of HIF-1, we screened a combinatorial library of 10,000 natural product-like chemical compounds. We show that KCN1, a lead inhibitor identified in this screen, potently inhibits HIF-1 activity in various cancer cell lines (IC50 ≈ 4 mM), while exerting minimal effects on the levels of HIF-1b, other short-lived proteins, or control proteins. To identify the molecular mechanism of action of this novel inhibitor, we compared the changes in global gene expression caused by KCN1 to those caused by a variety of previously characterized drugs using the Connectivity Map database. The gene expression signature of KCN1 was found to be similar to those produced by known inhibitors of Hsp90. Biochemical analyses confirmed that KCN1 modulates Hsp90 function by inhibiting the association of the chaperone with a model client protein. However, in contrast to known Hsp90 inhibitors such as geldanamycin, KCN1 does not inhibit the enzymatic activity of Hsp90 via direct binding to the ATPase pocket. Rather, the small molecule alters the availability of Hsp90 to its clients, including HIF-1a, by a novel mechanism of action. We found that KCN1 augments the interaction of Hsp90 with a cytoskeletal protein called plectin-1. This changes the bioavailability and subcellular distribution of the chaperone. Detailed experimental evidence supporting this novel mechanism will be presented. In an effort to translate KCN1 to the clinic we evaluated its toxicity and antitumor efficacy in an animal model. Systemic administration of KCN1 in nu/nu mice harboring sc human LN229 glioblastoma xenografts evidenced strong antitumor effects in the absence of noticeable toxic effects. These data suggest that KCN1 is a promising new agent for the treatment of malignant gliomas in a clinical trial.
In malignant gliomas, the c-jun N-terminal kinase (JNK) pathway regulates tumor growth and resistance to apoptosis, and correlates with histologic grade and epidermal growth factor receptor (EGFR) tyrosine kinase expression. We recently reported that glutathione S-transferase P1 (GSTP1), a major metabolizing and stress response signaling protein that is frequently overexpressed in brain tumors, is a downstream target of EGFR and undergoes EGFR-dependent tyrosine phosphorylation, in vitro and in vivo, resulting in an enhancement of its enzymatic function and increased drug resistance. Intracellularly, the dimeric GSTP1 exists in a reversible equilibrium with its monomeric form, the predominant form under most conditions. Monomeric GSTP1 binds to JNK and acts as an endogeneous JNK inhibitor. In this study, we investigated the impact of the EGFR-dependent tyrosine phosphorylation of GSTP1 ion the activation of JNK in glioma cells. We showed that the EGFR-phosphorylation of GSTP1 shifts the GSTP1 dimer-monomer equilibrium toward the monomeric state. Kinase analyses showed that the c-Jun phosphorylation by JNK was enhanced by phosphorylated-GSTP1 and suppressed by unphosphorylated-GSTP1. EGFR activation resulted in dissociation of GSTP1 from the GSTP1–JNK1 complex in a cell-free system and in glioma cells. Mutation of the phospho-acceptor Tyr198 to phenylalanine in the GSTP1 protein and overexpression of the Tyr198Phe mutant GSTP1 protein decreased JNK activity and c-Jun phosphorylation in EGFR- overexpressing glioma cells. Together these results indicate that in malignant glioma cells the tyrosine phosphorylation of GSTP1 by EGFR shifts GSTP1 from the dimeric to the monomeric form leading to dissociation of GSTP1 from the GSTP1JNK complex and reactivation of JNK, which is independent of JNK upstream signaling. This regulation of JNK signaling function by the EGFR-GSTP1 cross talk, defines a novel signaling network with a potential to regulate the biology and therapeutic outcome in malignant gliomas.
This work was supported by NIH grants RO1 CA127872, RO1 CA 112519, P50CA108786, and P30CA14236.
EGFRvIII is the exon 2–7 deletion mutant and commonly identified in glioblastoma patients. This leads to the ligand-independent tyrosine kinase activity that activates persistent downstream phosphatidylinositol 3-kinase pathway and promotes tumor growth, which makes EGFRvIII a therapeutic target. However, the functional role of EGFRvIII in the stem cell biology, another therapeutic target, has not been evaluated. In this study, we aim to quantitate the expression level of EGFRvIII and stem cell marker genes, and identify the stem cell marker that might have a functional association with EGFRvIII. First, using recent development real-time RT-PCR assay for detecting EGFRvIII, we evaluated EGFRvIII expression in 105 GBM samples. We could detect and quantitate EGFRvIII expression in 22 samples (20%) with variable expression level. Next, to evaluate the expression of stem cell marker genes, we selected 21 EGFRvIII-positive samples and 23 negative samples, and performed SYBR-based real-time RT-PCR assay. The stem cell marker gene included in this study are as follows: CD133, Nestin, BMI-1, MELK, Notch1, Notch2, Notch3, and Notch4. We also examined the expression level of other growth factors such as EGFR, EGFR2, EGFR3, EGFR4, PDGFRA, and PDGFRB. We evaluated the statistical association of each stem cell marker gene expression with EGFRvIII. Of the 14 genes, the results showed that EGFR (p<0.01), Notch2 (p=0.03), Notch3 (p=0.02), and Notch4 (p=0.01) were significantly highly expressed in the EGFRvIII-positive cases. This suggests that downstream pathway of Notch 2–4 may be functionally associated with EGFRvIII activation.
Aberrant EGFR signaling is a major contributing force to glioma progression and treatment resistance. The most prevalent mutation, EGFRvIII, promotes growth and survival of cancer cells. More recently, new point mutations in the extracellular domain of EGFR have been identified by the Cancer Genome Atlas. We are investigating the signaling of these abnormal receptors to identify novel druggable targets. We have analyzed the EGFRvIII specific signal using shotgun phosphoproteomics based on recovery of tyrosine-phosphopeptides and mass spectrometry. Glioma cell lines expressing EGFRvIII, a kinase inactive form of EGFRvIII and wild-type EGFR (LN428 PTEN+; LNZ308 PTEN−) were compared by this approach. Spectral intensity and count were used in a label-free quantification, and revealed 150 and 180 phosphoproteins, respectively, in the two cell lines used. Several phosphorylations showed a significantly greater level in EGFRvIII cells compared to wild-type EGFR. These include c-Met, Stat5, and Gab1. The signaling of EGFRvIII is ligand-independent and low intensity. Whether it dimerizes has remained controversial. Furthermore, the low-intensity signal makes it challenging to study. We have created a chimeric EGFRvIII that can be dimerized experimentally, using a variant FKBP12 domain and cognate small molecule, a process termed chemically induced dimerization (CID). CID increases EGFRvIII signaling severalfold in intensity, without leading to increased internalization or changes in the signaling nodes activated downstream, and is allowing us to investigate the nature of the EGFRvIII signal in greater depth than before. In itself, the observation that forced dimerization increases the signal of EGFRvIII is strong evidence that it ordinarily does not dimerize to a significant degree. Another possible mechanism behind EGFRvIII's impact on glioma biology is differential cellular localization and we have found that it localizes to the nucleus in an activity-dependent manner and requires Y1173 and Y845.
Both genetic and epigenetic mechanisms contribute to meningioma development by altering gene expression and protein function. The epidermal growth factor receptor (EGFR) is dysregulated or over-expressed in brain tumors and its expression is directly correlated with tumor malignancy and unfavorable prognosis. The leucine-rich repeats and immunoglobulin-like domains (LRIG) family proteins are membrane-bound proteins that have been proven to inhibit EGFR signaling through interaction with ubiquitination. The human LRIG2 is prominently expressed by myelinating cells and cytoplasmic expression of LRIG2 is associated with poor survival of oligodendroglioma patients. The aim of this study was to test the hypothesis that polymorphic variations in the EGFR and LRIG2 genes influence risk of meningioma.
376 meningioma cases, and 1103 controls recruited in the Nordic Interphone study were genotyped. Eighty-nine SNPs in the EGFR gene and six SNPs in the LRIG2 gene were selected to capture most of the polymorphic variants in these genes.
The EGFR gene SNP (rs759171) showed a higher minor allele frequency in meningioma (3%) compared with controls (1%). The highest risk of meningioma was associated with homozygosity for the EGFR variant (OR=3.10 CI, 95% 1.33, 7.25). When analyzing the LRIG2 gene SNP (rs1216801), the minor allele frequency was higher in meningioma (16%) compared with controls (12%). A significant association between risk of meningioma and homozygosity for the LRIG2 variant (OR=1.50 CI, 1.05, 2.15) was also evident.
As far as we know this is the first study where meningioma risk has been analyzed in association with polymorphic variants of the EGFR and LRIG2 genes. Functional studies in this area are important to assess whether polymorphic variants are also associated with meningioma tumor phenotype.
P53 tumor suppressor expression is lost in a variety of brain tumors. P53 is a transcription factor that regulates the expression of a large number of proteins, some of which are critical effectors of its tumor suppressive activities. In the present study we demonstrate the existence of a novel mechanism whereby the tumor suppressor p53 can control the growth of malignant gliomas and other solid tumors. We provide evidence that activation of p53 by gamma or ultraviolet irradiation or chemotherapy in wtp53 cells can induce cell death specifically in bystander cancer cells (but not normal cells) through p53-controlled release of a death-promoting factor. To identify this factor we performed proteomic analyses (2GE-MS and ICAT) on supernatant of glioma cells with inducible p53 (tet-on system) and identified a number of candidate secreted proteins. Using antibody neutralization and siRNA approaches we demonstrated that galectin-3 (Gal-3), a 31 kDa β-galactoside-binding proapoptotic factor was the mediator of this bystander effect. To identify the mechanism, we first showed that Gal-3 was not a direct target of wt-p53; rather its secretion is facilitated by p53 transcriptional activation of TSAP6, a key mediator of the nontraditional secretory pathway. We further demonstrated the biological importance of p53 controlled Gal-3 secretion by showing that it inhibits anchorage-independent tumor cell growth in vitro and can strongly reduce malignant glioma formation in vivo. We next investigated the mechanism of death and confirmed its specificity for tumor cells. We first showed that Gal-3 induces a caspase 9–dependent apoptotic pathway in more than 20 different tumor cell lines of different origins investigated so far, while normal cells were unaffected. Second, we demonstrated that Gal-3 induced apoptosis in tumor cells could be inhibited by depletion of β1 integrin using siRNA. Further investigations are ongoing to unravel the mechanism underlying the tumor cell specificity. In conclusion, our data demonstrate that p53 can exert cell extrinsic control over tumor growth through a novel mechanism leading to increased secretion of galectin-3. Gal-3 in turn induces apoptosis in adjacent tumor cells, suggesting that it can be directly exploited for tumor therapy.
This work is supported by the Goldhirsh Foundation.
Activated epidermal growth factor receptor (EGFR) signaling is one of the most prominent molecular alterations in malignant gliomas and sustains many of their key oncogenic capabilities. However, activation of EGFR signaling also increases energy demands by enhancing protein synthesis and nutrient consumption. Our recent finding that EGFR inhibition confers protection from hypoxia-induced cell death is congruent with this. Other recent work has demonstrated that p53 has an important role in the processing of starvation signals and that the p53-dependent molecular mediators of the Warburg effect, synthesis of cytochrome c oxidase 2 (SCO2) and TP53-induced glycolysis and apoptosis regulator (TIGAR), regulate glucose consumption and mitochondrial function. Notably, EGFR amplification and p53 mutations are almost mutually exclusive events in glioblastoma. We therefore hypothesized that the presence of wild-type p53 in glioma cells with activated EGFR is necessary in order to limit metabolic demands induced by deregulated signal transduction processes in the presence of hypoxia and nutrient depletion. We here report that shRNA- mediated gene suppression of p53 as well as the mutant temperature-sensitive dominant-negative p53 VAL135ALA increase glucose consumption and lactate production in p53 wild-type human malignant glioma cells and enhance hypoxia-induced cell death. These findings suggest that glioma cells may benefit from retaining p53 wild-type status by reducing their vulnerability toward tumor hypoxia and offer an explanation for the dichotomy of EGFR amplification and p53 mutations in glioblastoma.
To determine whether p53 deficient subventricular zone (SVZ) cells have the potential to undergo spontaneous transformation in vivo, and whether transformed cells can be used in therapy response experiments.
Athymic nu/nu mice were injected with either luciferase modified p53 null SVZ cells, or the same cells modified with human telomerase reverse transcriptase (hTERT) and human papilloma virus E6/E7. Cells were monitored for proliferation by bioluminescence imaging (BLI), and resultant tumors were harvested for use in orthotopic allograft therapy response experiments.
BLI revealed steady cell growth in all mice receiving intracranial (IC) injection with hTERT + E6/ E7 modified cells in contrast to six mice showing stable (n=3) or lack of (n=3) luminescence following injection of p53 null cells. In addition to the IC injections, one mouse each received subcutaneous (SC) injection with each type of cell. p53 null + hTERT + E6/E7 cells showed progressive SC growth, producing a tumor that has been resected and whose cells have been used for tumor propagation, both in SC and IC locations. SC-injected p53 null cells showed no indication of growth for 100 days, but thereafter showed rapid growth, producing a large tumor that has been used as a source of cells for intracranial injection in a series of 20 mice, 10 of which were treated with temozolomide. Temozolomide treatment significantly extended the survival of mice with IC tumor (p<0.0001).
Our study makes the following points regarding the use of p53 null CNS cells from genetically modified mice: (1) Cells can be used to assess tumorigenicity effects of transferred genes, such as hTERT and E6/ E7; (2) Cells have the ability to spontaneously transform in vivo; and (3) Spontaneously transformed cells can be used in an orthotopic engraftment model approach for studying therapeutic response.
A common pathobiological feature of malignant gliomas is the insidious infiltration of single tumor cells into the brain parenchyma, rendering these deadly tumors virtually incurable with available therapies. The ADP-ribosylation factor 6 (ARF6), which belongs to ARF family of small GTP-binding proteins with multiple roles in fundamental biological processes, has recently been shown to play an important role in tumor cell invasion. In gliomas, exogenous expression of EFA6A, a guanidine exchange factor (GEF) for ARF6 in glioma cells enhanced cell motility and invasiveness in vitro. However, whether ARF6 exerts a direct impact on glioma cell invasion is largely unknown. In this study, we report that ARF6, a Ras superfamily small GTPase, is abundantly expressed in invasive human glioma cells. In vitro, cellular depletion of ARF6 by siRNA decreased Rac1 activation, impaired HGF- and serum-stimulated glioma cell migration. In vivo and ex vivo, stable knockdown of ARF6 by siRNA markedly decreased the invasive capacity of invasive glioma in the brain. Ectopic expression of ARF6 by glioma cells promoted cell migration through activation of Rac1. Upon stimulation, IQGAP1, a key regulator of cell adhesion and migration was recruited to the membrane of the leading edges of migrating cells together with ARF6, allowing forward protrusion. Cellular depletion of ARF6 by siRNA abrogated the recruitment of IQGAP1 into cell membrane and attenuated the formation of the protrusions at the invasion fronts. Finally, using coimmunoprecipitation assays, we found that ARF6 was associated with Rac1 and IQGAP1 in glioma cells upon HGF stimulation. Knockdown of IQGAP1 by siRNA significantly inhibited the ARF6-promoted Rac1 activation and cell migration. In conclusion, these data suggest that ARF6 signaling is pivotal for glioma cell invasion in the brain and IQGAP1 is required for ARF6-mediated Rac1 activation and glioma cell invasion.
Sphingosine-1 phosphate (S1P) is a bioactive lipid that signals through a family of five G protein-coupled receptors, termed S1P1–5, and regulates cellular proliferation, migration, and survival. We investigated expression and role of S1P receptors in glioma.
We analyzed the expression and production of S1P receptors in 58 human glioma specimens by quantitative RT-PCR, Western blotting, and immunohistochemistry. The relevance of S1P receptors expression levels to survival in glioblastoma was examined by Kaplan-Meier method. We examined the mechanism of S1P receptors pertaining to glioma cell proliferation and invasion by experiments manipulating gene expression in glioma cell lines.
Expression of S1P1 receptor was significantly lower in glioblastoma than those in normal brain (p<0.01) and diffuse astrocytoma (p<0.05). Immunoblot demonstrated that normal brain expressed more S1P1 receptor protein than glioblastoma. S1P1 receptor was immunolocalized predominantly to the astrocytes in normal brain, but faint staining was observed in glioblastoma. Downregulation of S1P1 receptor expression correlated with poor survival of patients with glioblastoma (p<0.05). S1P1 receptor small interfering RNA promoted cell proliferation in high-expressor glioma cell lines (T98G, G112). Cell proliferation was promoted by pertussis toxin, an inactivator of Gi/o type of G proteins, to which S1P1 receptor is coupled exclusively. Forced expression of S1P1 receptor in low-expressor cell lines (U87, U251) resulted in decreased cell growth both in vitro and in vivo. In migration and invasion assays, no significant change was observed by manipulating S1P1 receptor gene. Furthermore, we found a significant association between the expression of S1P1 receptor and early growth response-1, a transcriptional factor that has a tumor suppressor function through PTEN in glioblastoma (p<0.05).
These data show that downregulation of S1P1 receptor expression enhances the malignancy of glioblastoma by increasing cell proliferation and correlates with shorter survival of patients with glioblastoma.
Eph receptors constitute the largest subfamily of receptor tyrosine kinases and interact with membrane-bound ligands termed ephrins. Eph and ephrins have many vital functions including cell adhesion, migration and axon guidance. Eph and ephrins have been found to be aberrantly expressed in many malignancies including brain tumor. The purpose of this study was to investigate Eph receptor function in glioma brain tumor stem cell neurosphere cultures.
Gene expression was investigated by Q-PCR and immunoblot analysis in high-grade glioma surgical specimens and cell lines. Targeted reduction of Eph expression was performed using an inducible shRNA system. Eph receptor function was also inhibited using soluble ephrin-Fc protein and the EphA3 mAb IIIA4.
In this study we investigated Eph and ephrin family expression in a screen of 37 high-grade glioma tumor specimens. Results highlighted elevated expression of EphA3 and the high-affinity ligand ephrin A5 in a significant proportion (70%) of samples. To further investigate EphA3 function the receptor was downregulated using shRNA in an EphA3+ glioma neurosphere cell line. Inducible downregulation of the EphA3 receptor resulted in initiation of neuronal and glial cell differentiation following activation of the ERK/MAPK pathway. A reduction in total stem/ progenitor cell proliferation was also observed when treated with EphA3 shRNA (46%) or soluble ephrin A5Fc (33%) to inhibit EphA3 function. Furthermore, treatment with the EphA3 mAb, IIIA4, inhibited cell proliferation (>50%) and induced rapid internalization of the receptor, cell spreading and increased adhesion. The reduction in cell numbers was not associated with apoptosis or cell cycle arrest. CFSE division tracking identified slower cell divisions in populations in which EphA3 signaling was attenuated.
We propose EphA3, in part, regulates cancer stem cell self renewal and cell division rate in glioma and could prove a potential therapeutic target.
During proliferation of normal human somatic cells, telomeres progressively shorten, leading eventually to senescence. However, the majority of cancers evade this by activating either telomerase or the alternative lengthening of telomeres mechanism (ALT) to prevent telomere shortening. Why ALT, rather than telomerase, is switched on in certain tumors is unknown and its presence has been associated with good and bad prognosis, depending on tumor type. Our goal in this study was to examine whether the presence of ALT in GBM tumors has prognostic significance.
We performed a retrospective review and analysis of 432 cases of GBM diagnosed from 1998 to 2007 at Royal North Shore and North Shore Private hospitals, Australia, and Auckland Hospital, New Zealand. To assess ALT, ALT-associated promyelocytic leukemia (PML) bodies (which are a known hallmark of ALT) were detected by combined PML immunofluorescence and telomere fluorescence in situ hybridization.
Of the 432 cases, 61 patients received surgery only, 110 patients were postoperatively treated with radiotherapy (RT), while 261 patients were postoperatively treated with RT plus chemotherapy administered either concurrently with RT or post-RT. ALT was detected in tumors from 61 patients (14.1%). In all treatment arms, ALT-positive patients displayed longer median survival compared to patients without ALT: No treatment, 3.8 months compared to 2.2 months, p=0.696; RT only, 10.1 months compared to 7.3 months, p=0.026; RT + chemotherapy, 17.8 months compared to 13.9 months, p=0.051. Approximately 33% of ALT-positive patients survived more than 2 years postdiagnosis as opposed to 12% of all ALT-negative patients treated with RT and chemotherapy.
The presence of ALT in GBM tumors correlates with patient survival. ALT positivity in GBMs not only predicts better outcome for these patients but also identifies a subgroup that will not benefit from telomerase inhibitors.
MicroRNAs (miRNA) are short noncoding RNAs associated with diverse biological processes. They are also involved in tumorigenesis and function as both tumor suppressors and oncogenes. Large-scale miRNA expression in glioblastomas, however, has not been quantitatively estimated to date. Here we determined global miRNA expression profiles of 12 malignant gliomas (8 glioblastomas [GBM] and 4 anaplastic astrocytomas [AA]) using TaqMan Human MicroRNA Array v1.0 (Applied Biosystems) by real-time PCR method. In total, the expression of 365 mature human miRNAs was examined. Expression of the 16 miRNAs was significantly altered in the two histological subtypes. Of them, the most significantly regulated miRNA was miR-196a (p=0.0038). In addition, miR-196b (family member of 196a) and miR-21, which was previously reported to be upregulated in GBM, were also included. This result was validated in another panel of 105 gliomas (74 GBMs, 18 AAs and 13 diffused astrocytomas) with individual Taqman microRNA assay (Applied Biosystems) by real-time PCR method. Furthermore, we examined the relationship between expression levels of these miRNAs and overall survival in 39 primary GBM patients. GBM patients with high miR-196 expression level showed significantly poor survival by Kaplan-Meier method (p=0.0073). Multivariate analysis demonstrated that high level of miR-196 was an independent and significant predictor of overall survival in GBM patients (p=0.036, HR=2.81). In conclusion, our results suggest that miR-196 overexpression could be relevant for the malignant progression of glioma. Consequently, miR-196 is of significant prognostic value for predicating survival in GBM patients.
Animal models of cancer have been very instructional in understanding the genesis of the tumors and very useful to test the new strategy of treatment. Among the brain tumors, glioblastoma multiforme (GBM) is the most malignant, and there is still a need to generate the animal model for this tumor that can recapitulate the abnormal molecular events causing the formation of human GBMs. We have recently developed a novel method to induce GBMs in adult immunocompetent mice by injecting Cre-loxP-controlled lentiviral vectors expressing oncogenes in a small number of cells. By injecting Cre-loxP-controlled lentiviral vectors into the specific region of the adult GFAP-Cre mice that express Cre recombinase specifically in GFAP-expressing cells, we established cell type and region specific expression of oncogenes. Activated forms of H-Ras and AKT were induced in fewer than 60 GFAP+ cells in hippocampus, subventricular zone or cortex of GFAP-Cre/p53+/− mice. More than 70% of the mice developed tumors when transduced in subventricular zone, and 100% of the mice developed tumors when transduced in hippocampus. However, tumors were rarely found when transduced in cortex. These tumors from hippocampus and subventricular zone showed the characteristics of human GBMs including pleomorphism, hypervascularity, necrosis, and pseudopalisading. Transplantation of fewer than 100 brain tumor cells into the naive recipient mouse brain led to the formation of GBM-like tumors. When these tumor cells were cultured in the stem cell medium, they formed neurospheres and contained CD133+ cells. Moreover these tumor cells can differentiate into neurons and astrocytes in response to the addition of fetal bovine serum to the stem cell medium. These results indicate that our novel mouse model can be used for a wide variety of brain tumor research including the brain tumor stem cell investigation, and might be useful for the deeper understanding of GBMs.
Glioblastoma animal models still remain a big challenge in basic and translational research as the most commonly used cell line-based models and do not recapitulate the invasive growth of human glioblastomas in patients. Recently, we have established a human glioblastoma xenograft model in nude rats that is highly invasive. Here, we characterize this model with an emphasis on neuropathological and radiological features. Immunodeficient nude rats were xenografted with biopsy spheroids derived from 29 patients diagnosed with primary brain tumors. Amongst these, 25 were primary glioblastoma multiforme (GBM), three were grade II gliomas and one was a grade III oligoastrocytoma. We monitored the animals with MRI and 11 of the resulting tumors were subsequently passaged in vivo for up to 19 generations while tumor take rates and survival data were recorded. The tumor take rate for rats xenografted with primary GBM biopsies was 77% and remained close to 100% at subsequent in vivo passages, whereas only one out of four lower grade tumors engrafted. Average time from transplantation to the onset of symptoms was 125 days. Culture time of the spheroids correlated inversely with tumor take rates, but not with survival. After four to five in vivo passages the tumors changed to a complete glioblastoma phenotype becoming more vascular, with increased cell proliferation, less apoptosis but more necrotic areas. By extended passages, the tumors gradually became less invasive and appeared more circumscribed (after eight to nine passages). In conclusion, the model provides reproducible high take rates and survival times if parameters related to tumor burden and the operative procedure are standardized. Furthermore, passaging in vivo modulates these tumors phenotypically, resulting in different combinations of angiogenic and invasive growth patterns. We therefore propose that the presented model provides a powerful tool for dissecting aspects related to brain tumor growth and progression.
We have been developing specific delivery vectors to Glioblastoma Multiforme (GBM). Interleukin 13 receptor alpha 2 (IL-13-Rα2), a cell membrane receptor, has been found to be overexpressed in more than 75% of GBM patients and absent in normal brain. Recently, we have begun a search for small peptides binding specifically IL-13-Rα2.
To isolate the peptides, we used the phage-display library with peptides presented on the surface of filamentous M13 phages. We screened Ph.D-C7C (New England Biolabs) library, with a diversity of 1.2 billions. The phage library was panned using GBM cell lines that overexpress IL-13-Rα2 and corresponding cell lines, which do not express the receptor. Phages were eluted from the IL-13-Rα2-positive cells and subjected to additional binding/amplification cycles.
We have isolated three different peptide phage clones. ELISA experiments confirmed that these three peptide phage clones bound to the recombinant IL-13-Rα-Fc receptor chimera protein and not to an irrelevant IgG-Fc control or bovine serum albumin protein. Moreover, ELISA experiments including the other IL-13 receptor protein, IL-13-Rα1, demonstrated that one of the peptides is truly specific for IL-13-Rα2. Furthermore, a 1000 fold excess of the IL-13 ligand did not block the peptide’s binding to the IL-13-Rα2. Also, cell-binding phage titer experiments were carried out on selected cell lines. We found that 200 times more peptide phages bound to the IL-13-Rα2-positive cell lines compared to the cell lines that do not overexpress the receptor.
We have identified specific heptapeptides that bind to IL-13-Rα2. The peptides are binding to the receptor at the site different from that used by the natural ligand, IL-13. These peptides will be further developed for diagnostic, imaging and molecular therapeutic interventions in GBM.
Glioma is the most common primary malignant CNS tumor in adults and remains a deadly disease despite of the development of the surgical procedure, radiotherapy and chemotherapy in the last decade. We used the serological identification of antigens by recombinant cDNA expression cloning (SEREX) to determine the novel serum markers in glioma. The phage expression library, constructed using mRNA derived from the U-87 MG glioblastoma cell line, was screened using sera from 48 patients with glioma and 31 independent genes were isolated. The enzyme-linked immunosorbent assay (ELISA) using recombinant proteins indicated that the levels of serum antibodies to SH3-domain GRB2-like 1 (SH3GL1) were significantly higher in patients with low-grade glioma than in healthy volunteers (p=0.0045) or in patients with high-grade glioma (p=0.0243). The independent validation test, using other sera from the patients, presented similar results. ELISA in the purified recombinant proteins of deletion mutants of SH3GL1 showed that approximately 10–20 amino acids in the C-terminal were indispensable as the epitope site. In immunohistochemical staining, overexpression of SH3GL1 proteins in cytoplasmic region of glioma cells according to its histological grade, but not in glial cells, was observed. Consequently, the levels of anti-SH3GL1 autoantibodies can be a novel low-grade glioma-specific serum marker.
Glioma patients are at high risk for venous thromboembolism (VTE). However, predictive laboratory parameters have not been identified. High platelet count (PLC) and increased soluble P-selectin (sPsel) were reported as risk factors in cancer patients. We investigated sPsel and PLC as risk markers in glioma patients.
The Cancer and Thrombosis Study (CATS) is a prospective observational study, whose end point is the occurrence of objectively confirmed VTE. sPsel was measured in the third week after neurosurgical intervention using a human sPsel Immunoassay (R&D Systems, Minneapolis, MN, USA). Multivariable Cox regression analysis was applied to calculate hazard ratios (HR) for VTE including PLC, sPsel, age, sex and extent of surgery.
One hundred forty patients with newly diagnosed high-grade glioma were analyzed (52 women; median age 54.5 years; interquartile range [IQR], 42.8–5.1) during a median observation time of 309 (range, 3–1664) days. Twenty patients developed VTE (6 women, 14 men; thereof two events were fatal pulmonary embolisms. The cumulative probability of VTE was 10% at 6 months and 15% at 12 months. sPsel levels (ng/ml) were higher in patients with VTE compared to those without (median=51.8, IQR, 36.9–66.0 vs. median=38.8; IQR, 30.7–52.1; p=0.011). PLC (G/l) was significantly lower in patients with (median=214; IQR, 166–248) than in those without VTE (median=255, IQR, 200–327; p=0.011). In multivariable regression analysis high sPsel (75th percentile, 55.1 ng/ml) and low PLC (25th percentile, 198G/L) were significant risk markers of VTE (HR=3.4, 95% CI 1.3–9.0, and HR=3.3, 95% CI 1.2–8.8, respectively).
Our study revealed two strong predictive risk markers for VTE in glioma patients. Elevated sPsel is associated with a 3-fold increased risk of thrombosis. In contrast to patients with other solid tumors, low PLC is associated with increased thrombosis risk.
The aim of this study was to investigate clinicoradiographic aspects together with pathologic features of an institutional series of gliomatosis cerebri (GC) and to analyze potential prognostic factors. Between 1997 and 2004, 50 patients underwent radiotherapy with diagnosis of GC based on radiographic feature and confirmed with biopsy. A retrospective analysis was conducted on these patients including the entire medical records, radiographic data and pathologic features. Survival outcome and potential prognostic factors such as the patient age, sex, performance status, the location of major involvement, grade of contrast enhancement, performance of decompressive surgery, and pathologic subtype and grade were analyzed. The median progression-free survival and overall survival lengths were 15.7 and 25.9 months, respectively. Poor prognostic indicators included higher pathologic grade, disease with contrast enhancement, lesions involving deep gray matter beyond the hemisphere or involving cerebellum and brainstem, and poor patient performance status. Given a wide variety on clinicoradiographic feature and histopathologic characteristics in GC, the prognosis should be fairly predicted in every individual patient to direct an optimal management.
Survival of patients with glioblastoma (GBM) remains poor. However, a small percentage of primary GBM patients live ≥3 years. Relatively little is known about the patterns of care and outcomes of these patients.
Nested case-control study of long-term survivors (cases) compared to standard survivors (controls) of GBM. Pathology was confirmed by a single neuropathologist who reviewed all cases and controls included in the study. Each long-term survivor (LTS) was matched to standard survivors (SS) by age and postoperative KPS. Clinical characteristics, treatment regimens and outcomes were reviewed. Logistic regression models were used to assess potential associations between baseline factors and the probably of long-term survival.
Matching yielded 27 cases (LTS) with 81 controls (SS). Baseline factors were as follows: female LTS 40.1%, female SS 34.6%, male LTS 59.3%, male SS 65.4%, LTS with hemispheric tumors 96.3%, SS with hemispheric tumors 93.8%, LTS with callosal tumors 3.7%, SS with callosal tumors 2.5%, LTS with multifocal GBM 0, SS with multifocal GBM 3.7%, LTS with subtotal resection (STR) 51.9%, SS with STR 59.3%, LTS with gross total resection (GTR) 29.6%, SS with GTR 23.4%, LTS with biopsy 18.5%, SS with biopsy 17.3%, percent of LTS receiving a vascular endothelial growth factor (VEGF) inhibitor 26.1%, percent of SS receiving a VEGF inhibitor 23.4%. There were no significant associations between the baseline factors considered and long-term survival. Median overall survival for LTS was 4.76 years (range, 3.15–10.54 years) and for SS was 1.30 years (range, 0.54–2.99 years).
Analysis yielded no significant associations between gender, tumor location, extent of resection or treatment with a VEGF inhibitor and the probability of long-term survival. Other factors, such as molecular or genetic differences, may help account for differences in survival and are being assessed.
WHO grade II gliomas (GIIG) are progressive tumors characterized by a continuous growth before malignant transformation. During the initial period, their radiological mean diameter increases at an average rate of 4 mm/year and the Velocity of Diametric Expansion (VDE) has been shown to be predictive of prognosis. The recent optimization of the therapeutic management of GIIG has allowed an improvement of their prognosis and an increase of the number of women considering a pregnancy. The study goal was to determine whether pregnancy could modify the natural history of GIIG, using a quantitative approach of the VDE.
We asked the French Glioma Study Group (REG) and the ANOCEF group to collect all cases of pregnancy in the adult female population of GIIG for whom MRI longitudinal follow-up was available. Repeated measurements of the mean tumor diameter were performed before, during and after pregnancy. Then, VDE was fitted by linear regression.
Seven pregnancies in seven females harboring a GIIG were reviewed (1992–2007). Six oligodendrogliomas and one mixed glioma were discovered before and during pregnancy in four and three cases, respectively. The median VDE was at 4.7 mm/year before or after pregnancy. During pregnancy, the VDE increased markedly in five cases (median VDE 8.5 mm/year) and remained stable in two cases. In one of the five patients in the negative group, radiotherapy was performed shortly after delivery and can be responsible for the VDE decrease.
The present results, using a quantitative approach of VDE measurements, suggest a possible negative interaction of pregnancy on the natural history of GIIG. We propose to inform the patient about the possible risks. For each envisaged pregnancy, the benefit-to-risk ratio should be carefully weighted. If a pregnancy is decided, we advise to perform a close neurological and MRI follow-up. Besides, cases collection continues.
The EORTC designed a phase III trial to define the role of adjuvant whole brain irradiation (WBRT) after local treatment (surgery or radiosurgery) of 1–3 brain metastases from solid tumors.
Patients eligible for radiosurgery had metastases ≤2.5–3.0 cm in diameter and in case of surgery a complete resection was mandatory. Only patients with absent or stable systemic disease or with asymptomatic synchronous primary tumors, and with WHO PS 0–2 were allowed. Patients were randomized to receive either adjuvant WBRT (30 Gy in 10 fractions) or observation (OBS). Primary end point was time to WHO PS deterioration to ≥3. Secondary end points were time to intracranial progression, frequency of neurologic death and overall survival. Analysis is by intent-to-treat (logrank, two-sided α=0.05).
Since 1996 –2007, 359 patients were recruited. Median time to WHO PS ≥3 was 9.8 months (95% CI 8.0–11.7) in the OBS arm and 9.8 months (95% CI 7.8–12.6) in the WBRT arm (p>0.5). Overall survival was 10.9 months (95% CI 9.2–14.6) in both arms (p>0.5). Cumulative incidence of intracranial progression at 6 and 24 months was 39.7% (95% CI 32.5–46.8) and 54.0% (95% CI 46.7–61.3) of the OBS patients, but only 15.2% (95% CI 9.9–20.5) and 31.4% (95% CI 24.5–38.2) of the WBRT patients (p<0.0001). Intracranial progression was a cause of the death in 77/179 patients (43%) of the OBS group and in only 45/180 patients (25%) of the WBRT group.
After radiosurgery or surgery of a limited number of brain metastases, adjuvant whole brain radiotherapy reduces the frequency of intracranial relapses and neurologic deaths but does not prolong the time period of functional independence and overall survival time.
The purpose of this study is to examine the impact of cyst formation and tumor hemorrhage on the local recurrence after surgery for single brain metastasis.
Between July 2001 and February 2007, 90 patients went through surgery for single brain metastasis at our institution. Sixty-three patients (70%) were male, and 27 (30%) were female. Mean age was 55 (+12) years. The primary sites were lung in 58 (64%), colon in 11 (12%), breast in 6 (7%), kidney in 5 (6%), and liver in 5 (6%). Hemorrhagic brain metastasis was noted in 11 patients. Among nonhemorrhagic 79 patients, 12 patients had large cystic brain metastasis that was determined by diameter of the tumor cyst (greater than 3 cm). Postoperative RT was given in 4 patients (32%) in cystic group and 21 (30%) in solid group (p=0.45), and was given in 24 (30%) in nonhemorrhagic group and in 4 (36%) in hemorrhagic group (p=0.73)
Among 11 hemorrhagic brain metastases, local recurrence was developed only in one patient. Among 79 nonhemorrhagic brain metastasis, 32 (40%) showed local recurrence. Twelve-month progression-free survival was 90% in hemorrhagic group, and 46% in nonhemorrhagic group (p=0.04). Seven patients (58%) showed local recurrence among 12 large cystic brain metastases, and 25 (37%) in the other group. Six-month progression-free survival in large cystic group was 57% and 75% in the other group. Twelvemonth progression-free survival was 0% in large cystic group and 52% in the other group (p=0.02).
It was found from the results that hemorrhagic metastasis rarely recurred; however, large cystic metastasis frequently recurred after surgery regardless of adjuvant therapy. Therefore, it seems reasonable to conclude that postoperative RT and close follow-up should be necessary in large cystic metastasis.
Although CSF cytology and MRI are standard methods for the diagnosis of neoplastic meningitis (NM), this complication of neoplastic disease still remains to be difficult to detect in some cases. We therefore reevaluated the sensitivity of gadolinium (GD)-enhanced MRI and cerebrospinal-fluid (CSF)-cytology for the diagnosis of LM differentially for solid and hematological malignancies and for normal or elevated cell counts.
We identified retrospectively 101 cases of NM diagnosed in our CSF laboratory since 1990 with complete data of both MRI and CSF cytology. Thirty-four had hematological; 67, solid neoplasms. CSF-cell counts were increased in 63 and normal in 35 patients.
For hematological neoplasms, MRI was positive in 53%. CSF cytology was positive in 97%. In solid tumors, we found MRI-sensitivity of 0.81 and cytological sensitivity of 0.76. With normal CSF-cell-counts, MRI was positive in 63% (0.57 hematological, 0.75 solid malignancies), CSF-cytology 78% (0.9 in hematological, 0.64 in solid neoplasms). In cases of increased cell-counts, MRI-sensitivity was 0.75 (0.52 for hematological, 0.87 for solid malignancies), and sensitivity of CSF cytology was 0.89 (1.0 for hematological and 0.82 for solid neoplasms). Twenty-three patients were treated with intrathecal MTX or Ara-C, 16 patients with liposomal Ara-C. Sixty-two patients were not treated intrathecally.
We confirmed here the high overall sensitivity of MRI for the diagnosis of neoplastic meningitis. The best sensitivity, however, was seen in solid tumors and elevated cell counts. In hematological malignancies, a markedly lower sensitivity of MRI was seen. Of note, we consider the very high sensitivity of cytology in hematological malignancies to be artificial due to methodological reasons of this retrospective study. We conclude that MRI is a very sensitive method to detect NM especially in solid tumors and elevated cell counts. With normal cell counts and hematological neoplasms, CSF-cytology remains to be superior to radiological methods.
Phosphorothioate oligodeoxynucleotides containing CpG motifs (CpG- ODNs) are immunostimulating agents with antitumor effects in animal models. Neoplastic meningitis is a devastating disease, with no efficient therapeutic available. A phase 1 trial was conducted in patients with neoplastic meningitis to define the safety profile of subcutaneous injections, combined or not with intrathecal administration, of a CpG ODN.
Cohorts of three patients were treated for 5 weeks with escalating doses of CpG-28 (level 1, 0.1 mg/kg/week subcutaneously; level 2, 0.3 mg/kg/week subcutaneously; level 3, 0.3 mg/kg subcutaneously associated with 3 mg intrathecally every other week). The diagnostic of neoplastic meningitis was based on CSF cytology or clinical symptoms with meningeal enhancement on MRI. The primary end point was tolerance. Secondary end points were time until neurological progression and survival.
Nine patients (3 per level) were treated between March 2007 and May 2008. Primary cancer was malignant glioma, breast cancer, melanocytoma, ependymoma, and small cell carcinoma. Median age was 51 years and median KPS was 70%. In patients who were treated with subcutaneous injections only, no significant improvements in clinical or radiological symptoms were seen. In the three patients who were treated with both subcutaneous and intrathecal administrations, clinical improvement was observed. However, these patients received concomitant treatment with spinal radiotherapy or systemic chemotherapy, which might have impacted the outcome. Adverse effects possibly or probably related to the studied drug were moderate and consisted in grade 2 lymphopenia, anemia, and neutropenia, local erythema at injections sites, fever, seizure, and back pain. The median time until neurological progression was 9 weeks. The median survival was 17 weeks. One patient is still alive.
CpG-28 was well tolerated at doses up to 0.3 mg/kg subcutaneously and 3 mg intrathecally. Main side effects were limited to local erythema, lymphopenia and fever. Escalating doses of CpG-28 for intrathecal administration is ongoing.
Many children diagnosed with brain tumors in lower-income countries may not have access to appropriate neurosurgical and/or radiotherapy standards. The potential of high-dose chemotherapy with autologous stem cell rescue in the context of childhood brain tumors to improve treatment efficacy while reducing long-term toxicities and costs has not been fully explored and might be usefully applied in middle-income countries to compensate for inadequate neurosurgery and/or radiotherapy. Since August 2008, the Cure2Children Foundation has supported both financially and professionally the development of two centers in Pakistan for stem cell transplantation applied to the cure of thalassemia major, a very prevalent (50,000 cases in Pakistan) and curable deadly genetic disease. The methodology employed consists of matched-related allogeneic bone marrow transplantation after administration of myeloablative chemotherapy (thiotepa 10 mg/kg, busulfan 14 mg/kg, and cyclophosphamide 200 mg/kg). Management standards for central venous access, severe pancytopenia, immune suppression, hospital infection control, and other relevant issues have been addressed by local training as well as with Web-based data management and videoconferencing. A total of seven transplants have been performed so far, three in an established center (National Institute of Blood Diseases, Karachi) and four in the two newly developed services mentioned above. No transplant-related deaths or untoward morbidities have occurred at present, detailed clinical and cost analysis will be presented. We believe that this experience might be relevant to the possible implementation of autologous transplantation for the cure of some brain tumors of childhood in lower-income regions.
The Philippines is a very diverse country because of the large socioeconomic gaps. Thus, we have centers with very advanced technology and medical expertise at par with international standards for treating brain tumors but are inaccessible to a large majority of people because of the high cost. For the low-income groups, free medical expertise as well as minimal hospital fees are available in tertiary care government hospitals, such that benign brain tumors can be treated. However, there are only a handful of these, hence the long waiting time for imaging studies, elective surgeries, and whole-brain radiotherapy. For tumors requiring adjuvant treatment with chemotherapy and other types of radiotherapy, the high cost of medicines and lack of radiotherapy equipment in government hospitals coupled with minimal insurance coverage limits care for these patients. To standardize care, several brain tumor centers offering multidisciplinary treatment have been set up. Regular brain tumor board meetings are also held in various centers. A year ago, the Philippine Society for Neuro-oncology was started to enhance brain tumor education and management. Each doctor seeing brain tumor patients tries to help by doing surgeries or chemotherapy in government hospitals where they can be done for free or at one-third of the cost, by referring patients for assistance to receive medicines at a reduced cost from either government agencies or other funding agencies. Despite this, the majority of low-income patients will not get the care they need, not because of a lack of medicines nor a lack of medical expertise but mainly because of the lack of money to pay for diagnostic tools, chemotherapy or radiotherapy. Long-term solutions are needed such as increasing imaging and radiotherapy facilities in government hospitals, lowering the cost of chemotherapeutic drugs or improving accessibility through various foundations. We can continue to improve the quality of care by continuing education programs and by involvement in clinical trials to increase patient options.
Establishing a pediatric neuro-oncology service can be challenging given the complexity of brain tumors in children. Creating such a service requires the involvement of different disciplines such as neurosurgery, neuropathology, and radiation oncology. These challenges are tougher and more multifaceted in a developing country than a developed one.
The basic factors required to build and sustain a successful pediatric neuro-oncology service are common to developed and developing countries. These include establishing a dedicated multidisciplinary team, empowering nursing, providing palliative care, and adhering to approved disease-specific protocols and guidelines. For developing countries, twinning initiatives with developed countries have been shown to enhance and speed the progress of newly established services in leukemia and lymphoma, but their contribution remains to be tapped for brain tumors. Integrating telemedicine into a twinning program can improve the quality of care of patients, especially if implemented in a prospective fashion.
It is a major challenge to coordinate different disciplines to work in conjunction and approve standard therapeutic approaches in developing countries. Another major obstacle is the inherent attitude toward nursing in many developing countries. Although twinning and telemedicine can positively influence a newly developed neuro-oncology service, it still can provoke sensitive issues to local teams in developing countries. To avoid unconstructive impact of telemedicine and twinning the mentoring team should be perceptive of such issues. This can be achieved through frank and continuous communication.
There is an urgent need to improve the care for children with brain tumors in developing countries. Immediate steps need to be taken at the local and international levels and individual plans developed for each program and country.
There were some important obstacles regarding pediatric brain tumors registrations and care in the Kyrgyz republic.
study childhood brain cancer incidence, age, sex, and ethnic differences.
Childhood cancer registration in Kyrgyzstan is carried out by a network of population-based regional registries, National Center of Statistics, since 1983. Collected date from forms submitted along morphological findings and death certificates. The population figures and cancer incidence rates were provided in for age (0–4, 5–9, 10–14), ethnic groups, for each sex and calendar years. Counted crude, age-standardized rates (ASR) per 1,000,000. Estimated population relative risk in the urban and rural areas.
There were 270 registered with new diagnoses of brain cancer in children (1983–2007): 150 (ASR 7.5) boys, and 120 (ASR 6.9) girls. Male/female proportion is 1.25. Total ASR annual childhood cancer incidence was 74.8. The most frequent diagnostic groups were leukemia (30.3%, ASR 20.8), non-Hodgkin lymphomas (9.9%, ASR 7.3). Nonregistered patients—17%. Analyses of geographical variations showed highest incidence of brain tumors in urban (Bishkek, Chui) than rural mountains regions (Osh, Naryn, Issykkul), (RR=2.2, 95% CI=1.57–8.99). Incidence rate was significantly higher in Russians (ASR 10.8), compared with native Kyrgyz’s (5.7) and Uzbeks (8.5). Only one pediatric oncology department (30 hospital beds) is in Bishkek (capital) in whole republic. Surgical treatment performed at the National (adult) hospital without radiation or chemotherapy. About 60% of childhood population (South part—Osh, Djalalabat, and Batken areas) have no possibility to receive any treatment, even chemotherapy or adequate initial brain surgery.
Childhood brain cancer incidence in Kyrgyzstan is low and similar to those reported from some Asian developing countries. The setting (radiotherapy) at the National cancer center is unsatisfactory both with regard to trained physicians, medical supplies and supportive care. Too many patients are seen in an inadequately staffed and equipped department. High-risk patients with brain tumors are at a definite loss, especially since supportive care is basically absent. There is an urgent need to improve the care for children with brain tumors.
Cancer pain is often undertreated even in developed countries with abundant resources and easy access to oral, parenteral, and transdermal opioids. In developing countries, where opioids are generally not available, the challenges to providing adequate analgesia to patients with cancer are often insurmountable. In spite of concerted efforts by the World Health Organization and other organizations to make oral opioids available and to educate physicians and government officials, little progress has been made in relieving pain in cancer patients in this important area. Major obstacles include (1) a limited budget for pharmaceutical agents, (2) preference for agents that prevent (i.e., vaccines) or cure (i.e., antibiotics) disease, (3) a focus on children and young adults who have many years to live, and (4) concerns about compliance (as immediate-release opioids must be taken every 3–4 h), storage (especially in poor households), and drug diversion. The documented association between drug diversion, addiction, violence, and AIDS in the West is worrisome to public health officials in developing nations who face many other important health issues. Nevertheless, the management of cancer pain is of high priority given the rapidly increasing incidence of cancer in the developing world and the fact that approximately 70% of patients with cancer experience severe pain during the course of their illness. As a result, novel approaches that address fundamental concerns regarding opioid availability in these countries are needed. One such approach employs an inexpensive, nonbiodegradable, polymeric implant designed to provide continuous hydromorphone to the subcutaneous tissue for 1–3 months. This subcutaneous implant is constructed of materials that are all approved by the FDA, can be implanted by a physician extender, and releases at a continuous rate without an initial “burst.” The materials and manufacturing are designed to provide an inexpensive product suitable for use in developing nations. It has the potential to reduce concerns about patient compliance, drug storage, and opioid diversion while making opioids available to patients in rural areas and reducing the number of follow-up visits necessary for medication refills.
Disclosure: Dr. Grossman holds several patents on the implantable hydromorphone polymer and is a principal in Axxia Pharmaceuticals.
We have evaluated 715 glioblastoma patients who were diagnosed in the Canton of Zurich, Switzerland, between 1980 and 1994, to provide information on how patients were treated at the population level. Despite a general policy during the study period of treatment by surgical intervention aimed at maximum tumor removal followed by radiotherapy, there was a marked tendency toward limited intervention and lack of treatment with advancing patient age. Of those younger than 65 years, 82% were treated either with surgery followed by radiotherapy, surgery alone or radiotherapy alone, versus 47% of patients 65 years or older. Only 25% of patients older than 75 years underwent surgery and/or radiotherapy, while the remaining patients were given best supportive care (BSC). The mean age of patients was 54.5 years for those treated with surgery and radiotherapy, 58.3 years for surgery alone, 62.2 years for radiotherapy alone, and 69.2 years for BSC. Among patients who were treated with surgery plus radiotherapy and those treated with radiotherapy alone, younger patients (<60 years) had a significantly higher survival rate than older patients (≥60 years). In contrast, no significant difference in survival was observed between younger and older patients treated with surgery alone or receiving BSC, suggesting that lower survival rates in elderly patients with glioblastoma may be at least in part due to a lesser response to radiotherapy.
The pseudocapsule can be found in the transition zone between the adenoma and surrounding normal pituitary tissue. We investigated the precise histology of the pseudocapsule and evaluated the remission rate, pituitary function and recurrence rate after intensive resection of the pseudocapsule.
In 616 patients with pituitary adenomas (Hardy type I–III) over a period of 14 years, we have introduced intensive resection of the pseudocapsule to achieve complete tumor removal. A combined pituitary function test and radiological study were performed before surgery, 1 year after surgery and at subsequent 1.5-year intervals.
Pseudocapsules were identified in 343 (55.7%) patients and the distinct pseudocapsules were observed in 180 (52.5%) patients. In the remaining 163 patients, the pseudocapsules were incompletely developed. Tumor infiltration was present in the pseudocapsule in 71 (43.6%) patients. The presence of a pseudocapsule was more frequent in prolactin (PRL)-secreting tumors (70.9%) than in growth hormone (GH)-secreting (55.0%) and adrenocorticotropin (ACTH)-secreting (40.0%) tumors. In the 243 patients of the total resection group, surgical remission rate was 99.1% in clinically nonfunctional tumors (CNPTs), 88% in GH-secreting, 70.6% in PRL-secreting, and 100% in ACTH-secreting tumors. The surgical remission rate was 86.2% in the presence of a pseudocapsule and 94.3% in the absence of a pseudocapsule. Preoperative hypopituitarism improved in 140 patients (57.6%), persisted in 47 patients (19.3%), and was aggravated in 33 patients (13.6%). The tumor recurrence rate was 0.8% in the total resection group and was 42.1% in the subtotal resection group.
We have shown that tumor tissue is frequently present within the pseudocapsule, suggesting that any tumor remnant in the pseudocapsule could be a source of recurrence and an obstacle to achieving complete remission. These results indicate that intensive resection of the pseudocapsule could result in higher remission rate without deteriorating pituitary function.
To evaluate local control and toxicity for pituitary adenomas treated with stereotactic radiotherapy (SRT).
Patients with pituitary adenomas referred and treated 1997–2007 were retrospectively reviewed (n=83). Median age was 47 years (range, 14–73), with 46 males and 37 females. Twenty patients had functional and 63 had nonfunctional tumors. Median follow-up was 42 months (range, 1–137). Two patients received SRT as their primary treatment, 38 received it postoperatively, and 9 for raised hormones. Thirty-four patients received SRT for radiological progression despite prior surgery with median time to progression following surgery being 12 months (range, 1–275). Before SRT, hormone replacement therapy was observed in 37% (thyroid), 35% (cortisol), and 30% (testosterone, males only). SRT dose was 50 Gy in 25 daily fractions using the GTC frame, and CT-MR fusion for planning (Radionic). The GTV and sella contents were treated, with no expansion from CTV for PTV margin. The prescription guideline was >95% coverage of the CTV by a minimum dose of 47.5 Gy, and maximum dose <52.5 Gy.
The 3-year progression-free survival for functional and nonfunctional adenomas was 94% and 92% respectively (p=0.90). Four patients had progression (three nonfunctional and one functional); among these, two had metastatic spread. One patient had salvage excision, one had radiosurgery, one patient required temozolomide for leptomeningial disease and one required palliative radiation to treat lumbar bony metastases. Post-SRT 43 patients (52%) had hypothyroidism, 35 (42%) required cortisol and 20 (24%) required testosterone. One patient had severe optic neuropathy. To date there were no second cancers.
Though with a relatively short follow-up, this study suggests fractionated stereotactic radiotherapy with a narrow margin is safe and effective for the treatment of pituitary adenomas.
The authors retrospectively reviewed the long-term outcome of linear accelerator radiosurgery for patients with acromegaly in our hospital in the view of biochemical remission and further analyzed the associated factors.
We reviewed patients who received LINAC radiosurgery between 1994 and 2004 due to residual or recurrent growth hormone (GH)-secreting functional pituitary tumor in our hospital. All patients were followed >3 years. The residual or recurrence of the tumor was defined as a persistent high level of growth hormone or insulin- like growth factor-1 (IGF-1). The biochemical remission was defined as basal growth hormone <5 ng/ml with a normal sex-and-age adjusted IGF-1.
A total of 22 patients were included in our review. The mean follow-up after radiosurgery was 94.68 months. The overall mean biochemical remission time was 49.2 months (median 42 months). Nineteen patients (86.4%) achieved the biochemical remission throughout the follow-up period. The 3, 5, and 8 year biochemical remission rates were 36.4%, 54.5%, and 72.7% respectively. The initial GH at diagnosis and the pre-SRS GH correlated with biochemical remission (p=0.005, p<0.0001). Overall post-SRS hormone deficit was in five patients (22.7%).
Compared to other radiosurgery facilities, LINAC radiosurgery also provided a competent outcome. SRS has maximum effect in the first 2 years. Moreover, SRS showed long-term biochemical effects and needs longer follow-up for better biochemical remission.
We have reported the effectiveness of intracystic chemotherapy with bleomycin for cystic craniopharyngiomas. This result has been confirmed by the following reports. However, relevant complications, including cerebral ischemia, visual and hypothalamic damages, associated with local bleomycin therapy have been also reported. We have, so far, experienced no such significant complications with our methods. Here we report the way to avoid such complications.
Intra-cystic bleomycin administration was performed on 14 children since 1988 in our institutes. When radiological examinations indicated cystic craniopharyngiomas, we removed the cyst wall partially to confirm the pathological diagnosis and then placed the Ommaya reservoir tube into the tumor cavity. It is important to place all the side holes of tube within the cyst cavity and to tighten its entrance so as to avoid leakage of infused bleomycin. Bleomycin was administrated 2 weeks postoperatively via the Ommaya reservoir connected to the tube. Cystography should be performed prior to bleomycin administration. A smaller dose (5 mg or less) per injection or infrequent (every other day) injections would lessen the complications. In addition, the concentration of bleomycin in the cyst is most important. If the dilution of bleomycin is not enough, the drug may leak through the wall of the cyst. We suggest an appropriate concentration of bleomycin solution is 1 mg/ml or less and the timing of injection is immediately after the aspiration of cystic fluid. In nine children the cysts have almost disappeared and the children have achieved a good school life. Four children are also achieving a good life condition after additional stereotactic radiosurgery. No severe complications were observed.
We recommend several tips for appropriate usage of bleomycin to avoid complications and to achieve a good result.
Craniopharyngioma is a histologically benign tumor that originates from epithelial nests or from areas of squamous metaplasia. The incidence of craniopharyngioma is 2–10% in pediatric primary brain tumors. Surgical removal and radiation therapy are associated with significant risk for morbidity in the pediatric population with craniopharyngioma. Intracystic bleomycin therapy (IB) has been proposed as a treatment for predominantly cystic craniopharyngioma. The aim of the study was to review the effectiveness of intracystic bleomycin therapy in our craniopharyngioma children.
Pediatric craniopharyngioma patients treated with IB at Taipei Veterans General Hospital had been enrolled to participate in the retrospective review from 1999 to 2008. Brain MRI or CT was performed every 3 months to evaluate the objective response of intracystic bleomycin injection.
Seven boys and two girls received the intracystic bleomycin therapy. The median age is 7.8 years (range from 3.3 to 11.8 years). The median follow-up duration was 44 months (range from 9 to 79 months). For the total nine patients, eight achieved good response to IB therapy, and one achieved progression. The patient with progression showed no response to IBT and tumor relapsed quickly after 2 months of treatment. The median progression-free survival was 29 months (range from 2 to 79 months.).
Our preliminary results showed intracystic administration of bleomycin is a valid and effective therapy for certain children with craniopharyngioma. IBT may delay the need for aggressive surgery or radiation therapy for several years. However, further prospective randomized studies are needed to evaluate the feasibility and effectiveness of this treatment.
While total resection is a desirable goal for the treatment of craniopharyngioma, it is not always accomplished, and is associated with permanent endocrinological and visual deficits. This knowledge has led us to perform intratumoral bleomycin injection as a primary therapy for cystic craniopharyngioma. Five patients (four adults and one child) in whom cystic craniopharyngioma was recently diagnosed were treated primarily with a schedule of 3 mg of intracystic bleomycin three times a week for 5 weeks. One patient was excluded due to bleomycin leakage in reservoir permeability test. The remainder did not require subsequent resection and radiotherapy. Median follow-up period was 36 months. There was a reduction in cyst size greater than 90% with the preservation of visual and endocrinologic function. Intracystic bleomycin administration could be a feasible option for the treatment of predominantly cystic craniopharyngioma in terms of tumor control and tolerability.
Despite meningiomas being the most common benign intracranial tumor little is known about their association with epilepsy and outcome. The aim of this study was to define their epidemiology, identify their relation with epilepsy, and the factors predisposing to recurrence as well as prediction of outcome.
The case notes of 1,101 patients treated in the National Hospital for Neurology and Neurosurgery, London, UK, 1975–2000, were analyzed. Data collected included age, sex, clinical symptoms, GCS, tumor location, type of pre- and postoperative epilepsy, Simpson grade, treatment, recurrence and outcome using KPS.
Mean follow-up: 36 months. Females: 65%; males, 35%. Age at presentation ranged from 7 to 97 (mean 56) years. Meningiomas were more likely to present with visual symptoms (25.28%), headaches (21.19%), seizures (17.20%), and paresthesia (10.97%). Presenting symptoms and tumor location were statistically linked (p<0.0001). Symptoms and recurrence were not strongly related (p<0.475) as opposite to symptoms and postoperative KPS (p<0.0001). Simpson resection grade (I and II are less likely to recur), and pre- and postoperative epilepsy are related to tumor recurrence. Preoperative epilepsy is linked to tumor location (p<0.013) and postoperative epilepsy, age and histopathology results were linked to recurrence (p<0.0001) as well as tumor location (p<0.006) and laterality (left-sided tumors recur more than right-sided) (p<0.006). The presence of psammoma bodies was related to a benign histological diagnosis (p<0.004). Postoperative radiotherapy reduced recurrence rate (p<0.0001). Recurrence peaks on the 24th month and 10th year from the first treatment date.
Assessment of the identified factors may aid in outcome prediction and preoperative counseling and decisions related to patient treatment.
Nonbenign (atypical and malignant) meningiomas are rare tumors and only few studies on larger series have been performed so far. Because of their tendency to recur and thus poorer prognosis, they constitute a separate biological and clinical entity compared to benign meningiomas.
92 patients with nonbenign meningiomas operated on in years 1990–2002 at the University Medical Centre Ljubljana were analyzed in our retrospective study. Adopting the newest (2007) WHO classification criteria, 86 tumors were microscopically reclassified as primary atypical or malignant meningiomas. A histological, immunohistochemical and statistical analysis was undertaken to determine the correlation of several clinical, radiological, histological, and immunohistochemical factors with survival.
Mean and median overall survival times were 6.9 and 6.3 years, respectively. Overall 5- and 10-year survival rates were 74.7% and 54.3%, respectively. Univariate analysis of 86 nonbenign meningioma patients confirmed female sex (p=0.035) and age under 56 years (p=0.004) were significantly related to prolonged survival. Among histological factors, microinvasion into brain parenchyma (p=0.011), dense cellularity (p=0.030), prominent nucleoli (p=0.020), and sarcomatous appearance (p=0.019) all correlated with decreased survival. There was no difference in survival between patients with low (equal or less than 5%) or high (greater than 5%) proliferation index (p=0.329).
In patients with totally excised atypical and malignant meningiomas, presence or absence of microinvasion into brain parenchyma is a better predictor of survival than Ki67 proliferation index.
To review the skull base chordomas and chondrosarcomas and analyze comparative outcomes between these tumors. Between 1991 and 2005, 30 consecutive patients with pathological diagnosis of chordoma (n=19) or chondrosarcoma (n=11) of the skull base were managed by multimodal treatment combining one or more surgical approaches with conventional radiotherapy and/or GKRS. A retrospective analysis was conducted on these patients. Follow-up data were complemented by a most recent telephone interview with the patients or their family members. Age of the patients ranged from 3 to 69 years (median, 36 years). Seventeen patients were female and 13 male. Average length of follow-up was 56 months (range, 2–172 months). A total of 43 surgical approaches (31 for chordoma vs. 12 for chondrosarcoma) were performed at the initial or recurrent setting. Adjuvant radiotherapy and GKRS were performed on 21 and 6 patients, respectively. Four deaths occurred during the follow-up, two of which resulted from progression or recurrence of chordoma, one from pulmonary embolism, and another one from the sepsis unrelated primarily to the tumor. Recurrence of chordoma was observed in 5 patients out of 13 (38%) and 6 out of 10 (60%), each within 3 and 5 years after the initial treatment, whereas only one recurrence of the chondrosarcoma was observed in 2 years after the surgery with no radiation. Among 15 survivors with the chordoma, four patients were suffering from severe disability with progressive disease. In contrast, most patients with the chondrosarcoma harbored stable disease and less disabling symptoms except one instance of recurrence. Outcomes in the functional status as well as survival were much worse in chordomas than those in chondrosarcomas. The authors recommend a maximum safe resection, possibly to the extent of total resection at the initial setting, combined with complementary treatment, especially GKRS for the management of the skull base chordomas.
The purpose of this study was to evaluate the effect of stereotactic radiosurgery (SRS) for chordomas and chondrosarcomas of the skull base and to determine the optimal marginal dose to control these tumors.
Fourteen patients with histologically confirmed chordomas or chondrosarcomas underwent 16 sessions of SRS using Gamma Knife. The mean tumor volume was 12 cm3 and the marginal doses ranged from 10 to 20 Gy (mean, 15 Gy). Lower marginal doses of 12 Gy on average (range, 10–12.5 Gy) was applied for four patients since they underwent prior fractionated radiotherapy, and partial treatment for which parts of tumors were excluded from planned target volume (PTV) because of their proximity to critical normal structures such as the optic pathway and the brainstem was applied for five patients. The whole tumor volume was covered with higher marginal doses of 18 Gy on average (range, 16–20 Gy) for six patients. The mean follow-up period after SRS was 55 months.
Progression-free survival (PFS) rates at 1, 2, and 5 years after SRS were 93%, 51%, and 34%, respectively. Five-year PFS rate for patients whose tumors were completely covered by PTV was 53 months and tended to be higher than that of 17 months for partially treated patients (p=0.11). Five-year PFS rates for patients who underwent SRS with higher and lower marginal doses were 60% and 14%, respectively, which were significantly different (p=0.023). Tumor progression after partial irradiation mainly occurred from sites where delivered doses were reduced.
Sufficient marginal dose at least 18 Gy appears crucial to obtain tumor control in case of chordomas. Proper combination with surgical resection to detach tumors from critical structures and to reduce tumor volume is necessary to completely deliver sufficient marginal doses to lesions at the time of SRS.
To assess the results of proton beam therapy (PBT) in the treatment of skull base chordomas and chondrosarcomas.
Between April 2004 and August 2008, 11 patients (median age, 40 years) with chordomas (n=6) and chondrosarcomas (n=5) underwent proton therapy at Shizuoka Cancer Centre. The median tumor dose was 67 GyE. Late toxicity was assessed using the NCI-CTCAE grading system. The median follow-up time was 29.1 months (range, 3–55.8 months).
Actuarial 2-year local control rates were 80% for chordoma and 100% for chondrosarcoma. No regional failure or distant metastasis were observed, except for one patient with dedifferentiated chordoma. Four patients had pharyngitis, otitis media, and skin trouble of grade 1–2 as acute adverse events. No patients presented with posttherapeutic brainstem or optic pathways necrosis or dysfunction.
The results compare favorably to other irradiation series. Observed toxicity was acceptable. These preliminary results are encouraging, but should be confirmed with a long-term follow-up.
Temozolomide (TMZ) induces prolonged G2M arrest of glioma cells, and then nonapoptotic cell death associated with senescence-like phenomenon (in p53 wild-type cells) or mitotic catastrophe (in p53 nonfunctional cells). To clarify the mechanism of TMZ resistance and to develop more effective TMZ-based regimen, we established U87MG-derived TMZ-resistant (TR) clones by serial treatment with TMZ. We found that TR clones underwent only transient or minimal G2 arrest following TMZ treatment even though G2 checkpoint was clearly activated as in their parental cells and that G2 checkpoint inhibitor resensitized these clones to TMZ. Since it has been suggested that G2 checkpoint system might suppress the linkage between DNA damage and cell death, we tested the effect of a cdk inhibitor flavopiridol that could target cdc2 (cdk1), a key protein in G2 checkpoint pathway, on TMZ-treated cells. Flavopiridol (<10 nM) potentiated the cytotoxicity of TMZ, and suppressed the expression of key proteins at G2M transition including polo-like kinase 1, aurora kinases and Pin1. FACS analysis revealed flavopiridol induced accumulation of TMZ-treated cells, but not untreated cells, exclusively at G2 (4N DNA content). This “complete” G2 arrest was associated not only with increased expression of γ-H2AX, a DNA double-strand break marker, but also with increased release of cytochrome C in cytoplasm, which suggests that flavopiridol promoted apoptotic cell death signal in TMZ-treated glioma cells. Flavopiridol also enhanced cytotoxicity of TMZ to the cells with overactivated Akt, which has been previously shown to promote TMZ resistance, and resensitized TR clones to TMZ. Our results suggest that TMZ resistance could be promoted by enhanced DNA repair activity in G2M transition following G2 checkpoint activation, and that cdk inhibitor enhanced TMZ cytotoxicity by suppression of this activity. We conclude that cdk1-targeted compound might be useful as chemosensitization agent for gliomas.
Temozolomide (TMZ), an active alkylating agent in treating malignant gliomas, kills tumor cells with ill-defined molecular mechanism. To understand TMZ genotoxicity, we analyzed temporal regulation of DNA repair genes with levels of DNA damage, checkpoint activation and cellular status of O6-methylguanine-DNA methyltransferase (MGMT). During the first postdrug treatment cell cycle (PTC), rapid DNA double-strand breaks (DSBs) were detected (1 h) in MGMT-deficient cells, followed by ATM phosphorylation, p53 accumulation, transient G1/S arrest, and significant reduction of S-phase progression. Concurrently, global transcription repression was discerned, followed by 2- to 5-fold upregulation of nine nucleotide/base excision repair (BER/NER) and five DSB repair genes. During the second PTC, prominent G2/M arrest was accompanied by more than 5-fold upregulation of 29 genes, which were members of multiple DNA repair pathways. In MGMT-proficient cells, significantly lower levels of DSBs (less than 30% of MGMT-deficient cells) and lack of NER gene upregulation were identified in the first 6 h when substantial level of MGMT was observed. Association of NER and rapid DSB formation was further demonstrated by overlapping nuclear staining of the major NER endonuclease XPG and the DSB marker g-H2AX protein. In conclusion, our data suggest that distinct transcriptional regulation of DNA repair genes contributes to the different TMZ genotoxicity at the first two PTCs. To our knowledge, this is the first study analyzing temporal regulation of global DNA repair and corresponding genotoxicity in different MGMT context for SN1-type alkylating agent. Our findings will also have significant implications for improving O6-alkylguanine–based cancer chemotherapy.
Anaplastic oligodendroglial tumors (AOD) are chemo-sensitive tumors. Two large randomized trials did not see improved survival in patients treated with adjuvant pivotal trial on chemoirradiation with temozolomide suggested that benefit from the addition of temozolomide to radiotherapy was basically confined to patients with a methylated MGMT gene promoter and thus no alkyltransferase expression. In the present study we assessed the impact of MGMT promoter methylation on progression-free survival (PFS) and overall survival (OS) using MLPA in the prospective randomized EORTC study 26951 on adjuvant PCV on 368 patients with AOD. The control arm received only radiotherapy (RT), the experimental arm RT plus PCV.
MGMT methylation was assessed using a methylation-specific MLPA (MRC-Holland) based on methylation sensitive restriction analysis. To estimate the fraction of methylated MGMT promoter DNA, normalized values obtained with each MGMT probe of digested DNA samples (in which only methylated DNA will remain undigested and produce a PCR product) were divided by normalized values of corresponding undigested DNAs. For the MGMT gene promoter CpG methylation in three regions was assessed, from which an average score was calculated; a ratio above 0.25 is considered indicative of methylation. Statistical analysis was done using this cut-off and with the MGMT average score expressed as a continuous variable, both for PFS and OS.
In 152 of the 165 patients with sufficient material available for MGMT promoter analysis a result was obtained. Thirty-nine of these tumors showed 1p/19q codeletion. In 121 of the 152 cases (80%) an MLPA average consistent with methylation was observed, which included 38 of the 39 1p/19q codeleted samples (97%, p<0.001). MGMT promoter methylation was found to be equally prognostic in both the RT and the RT/PCV treated patients, for both PFS and OS. In multivariate analysis using MLPA average as a continuous variable, MGMT promoter methylation and 1p/19q were independent prognostic factors. The PFS hazard ratio reduction of MGMT promoter methylation was 0.303 (95% confidence interval [CI] 0.135–0.677), for 1p/19q codeletion 0.423 (95% CI, 0.232–0.768). The cut-off as proposed for MS-MPLA was not optimal to separate patients in two groups of different prognosis.
1p/19q codeletion was strongly associated with MGMT gene promoter methylation. In this homogeneously treated group of AOD patients, the independent prognostic favorable effect of MGMT promoter methylation on both PFS and OS was equally strong in the RT group compared to the RT/PCV group. There is currently no mechanistic explanation for the improved PFS in MGMT gene promoter methylated tumors after RT only, and other possible explanations need to be investigated.
The objective of this study was to assess the safety and tolerability of cediranib (Recentin), an oral, pan-VEGF receptor tyrosine kinase inhibitor, in combination with oral lomustine and to determine a dose for further studies with this combination in recurrent glioblastoma patients.
Patients at least 18 years old with pathological confirmation of recurrent glioblastoma were eligible. Patients could not have received previous anti-VEGF therapy and could have had no more than two prior chemotherapy regimens. Patients completing the first cycle (6 weeks) or having a dose limiting toxicity (DLT) before 6 weeks were evaluable. A total of 12 patients were enrolled in the study. Six patients were enrolled in the first cohort at a dose of cediranib 30 mg in combination with lomustine 130 mg/m2. Another six patients were enrolled in the second cohort at a dose of cediranib 30 mg in combination with lomustine 110 mg/m2.
There were 2/6 DLTs reported in the first cohort (one grade 3 fatigue, one hypertensive crisis due to untreated hypertension). Five of six patients developed grade 3 or 4 hematological toxicities attributed to lomustine. The lomustine dose was reduced to 110 mg/m2 for the second cohort. There was 1/6 DLT (grade 3 fatigue) in the second cohort.
The results of this phase I study demonstrate that the combination of cediranib 30 mg with lomustine 110 mg/m2 is safe and tolerable. An international, multicenter, randomized phase III trial has been initiated.
Single agent signal transduction modulators generally have been ineffective in most cancers including glioblastoma. Therefore, inhibiting multiple signaling pathways may be more effective. However, performing individual phase I/II trials for each combination is inefficient. This study used a sequential accrual design to test three combinations of sorafenib, with erlotinib, temsirolimus, or tipifarnib.
The NABTC conducted phase I studies of sorafenib (VEGFR/PDGFR/Raf inhibitor) in combination with erlotinib (EGFR inhibitor), temsirolimus (mTOR inhibitor), or tipifarnib (farnesyltransferase inhibitor) in recurrent GBM. Accrual was sequential, decreasing study pauses for maturation of toxicity data.
Histologically proven GBM, radiologic progression, >18 years old, KPS>60, adequate bone marrow and organ function. No limit on the number of prior therapies. No enzyme- inducing antiepileptic drugs were allowed. Dose-finding used a standard 3+3 design. MTD was defined as the dose with DLTs in 1/6 or fewer patients. Serum pharmacokinetic (PK) studies were performed.
Overall 49 patients were enrolled onto phase I. The MTD was sorafenib 400 mg BID and erlotinib 100 mg qD. The MTD was temsirolimus 25 mg qweek and sorafenib 400 mg BID. The MTD was sorafenib 200 mg BID and tipifarnib 100 mg qD×21 days, but the tipifarnib is below therapeutic levels. PK studies showed no drug-drug interactions for sorafenib with tipifarnib or temsirolimus. However, there was no accumulation of erlotinib, suggesting that sorafenib alters erlotinib metabolism or clearance.
This study demonstrated that sorafenib can be combined with other targeted agents in and a sequential accrual design decreases logistical problems of phase I studies in multicenter trials. Phase II doses were successfully determined for sorafenib with temsirolimus and with erlotinib, permitting enrollment onto the phase II component. The new phase I study using an alternating week schedule of tipifarnib will be performed to permit a dose more likely to reach therapeutic concentrations.
Recurrent malignant glioma (MG) patients treated with bevacizumab (BV), a neutralizing monoclonal antibody to vascular endothelial growth factor (VEGF) with or without chemotherapy, have noted significant therapeutic benefit. In this study, we evaluate the efficacy of BV plus etoposide (E), a topoisomerase inhibitor, among recurrent MG patients.
Recurrent patients with no more than three prior episodes of recurrence are eligible, while those with prior BV treatment or prior intracranial hemorrhage are excluded. The primary outcome measure is 6-month progression-free survival (6PFS). BV is dosed at 10 mg/kg intravenously every other week. Etoposide is orally administered daily (50 mg/m2) for days 1–21 of each 28-day cycle.
Fifty-nine patients (GBM, n=27; grade 3 MG, n=32) with a median of two prior progressions have enrolled. With a median follow-up of 45 weeks, median overall survival (OS) for GBM and grade 3 MG patients was 46 and 47 weeks, and 6PFS was 44% and 40.6%, respectively. The most common toxicities were neutropenia (41%), fatigue (22%), and infection (20%) and were grade 2 in most cases. One patient developed grade 1 intracranial hemorrhage and one patient had a grade 4 GI perforation.
Combination of bevacizumab and etoposide is well tolerated in recurrent MG patients and is associated with encouraging antitumor benefit. Accrual is complete and an update of further treatment and follow-up will be presented.
Bevacizumab (BV) has shown a promising activity in recurrent malignant gliomas (MG) in combination with irinotecan. Few data are available on the combination of bevacizumab and nitrosoureas.
In this ongoing phase II study patients with MG recurrent after surgery, radiation therapy and temozolomide are eligible. The treatment consists of an induction phase with BV at 10 mg/kg intravenously on day 1 and 15 and fotemustine (FTM) (a nitrosourea with elevated lipophilic properties) at 75 mg/m2 intravenously on day 1 and 8, followed after a 3-week interval by a maintenance phase with BV at 10 mg/kg i.v. and FTM 75 mg/m2 i.v. every 3 weeks until tumor progression or unacceptable toxicity. Patients undergo clinical and MRI assessment. The coprimary end points are objective response rate (ORR) and progression-free survival at 6 months (PFS6), with secondary end points of safety time to tumor progression (TTP) and overall survival.
From April 2008 to December 2008, 34 patients were enrolled, and 31 (22 glioblastomas and 9 anaplastic gliomas) were evaluable for response. Overall response rate (2 CR and 9 PR) was 35% (glioblastomas 33%, anaplastic gliomas 41.5%). Steroids were reduced in 50% of patients. Sixteen of 31 patients progressed with a TTP of 2.6 months (1–8.5). Progression was local in 10/16, local + leptomeningeal spread in 3/16 and gliomatosis in 3/16. Fifteen patients are free of tumor progression (from 2 to 8 months). Toxicities included grade III–IV neutropenia in three patients, grade III–IV piastrinopenia in five, and grade III thrombosis in two. Seventeen patients developed mild to moderate fatigue, six arterial hypertension, and three grade I intratumoral hemorrhage.
Combination of bevacizumab and fotemustine in recurrent malignant gliomas is safe and promising. Updated results, monitoring of CBV with perfusion MRI and correlations between MGMT promoter methylation and response/outcome will be presented.
Bevacizumab (B), a neutralizing VEGF monoclonal antibody, has antiglioma activity as single agent and in combination with cytotoxic therapy. Erlotinib (E), an EGFR tyrosine kinase inhibitor, may exhibit antitumor activity in some high-grade glioma (HGG) patients. B plus E was associated with clinical benefit in several cancers. We performed a single-arm, phase II trial to assess the efficacy and safety of B and E in patients with recurrent HGG.
The primary end point was 6-month progression-free survival (PFS-6). Radiographic response, pharmacokinetics and correlative biomarkers were secondary end points. E was orally administered daily at 200 mg/day for patients not on enzyme-inducing anticonvulsants (EIAC) and 500 mg/day for patients on EIAC. All patients received 10 mg/kg of B intravenously every 2 weeks.
Fifty-six patients with recurrent HGG (n=24 for glioblastoma multiforme [GBM] and n=32 for anaplastic gliomas [AGs]) were assessable for outcome. The PFS-6 rates were 25% for GBM and 50% for AGs. There was no survival difference between EIAC and non-EIAC groups. Rash (54% grade 1–2 and 38% grade 3) was the most common side effect. Fatigue, nausea and diarrhea were also common but mostly grade 1–2. Serious side effects were rare and included two patients with pulmonary embolism, single patients with either intestinal perforation, ischemic stroke, gastric bleeding or nasal septal perforation. Pharmacokinetic and tissue biomarker profiles are in preparation.
Combination of bevacizumab and erlotinib is tolerated and associated with antitumor benefit among heavily pretreated recurrent high-grade glioma patients.
We have established highly invasive nonangiogenic and less invasive highly angiogenic GBM xenografts by passaging human glioblastoma biopsies in nude rats. The animal model shows that GBM progression is characterized by two distinct tumor phenotypes; one stem-like phenotype that relies on proinvasive programs and one less infiltrative angiogenesis-dependent phenotype that shows loss of stem cell markers. Recent clinical information indicates that bevacizumab treatment has a dramatic effect on the contrast enhancing tumor compartment whereas the invasive part of the tumor is affected to a lesser extent. At present it is unclear if the remaining tumor cells represent a particular stem cell like subpopulation within the GBM. To address this question we assessed the treatment response of GBM xenografts to bevacizumab. In particular we asked the following: (1) To what extent does antiangiogenic treatment effectively reduce tumor volume by decreasing proliferation and/or inducing tumor cell death, as opposed to simply inducing a normalization of blood vessels. (2) To what extent does it induce the survival and proliferation of the cancer stem-like cell leading to increased tumor invasiveness and the development of secondary tumor foci. (3) Alternatively, does it trigger a conversion of the tumor cell metabolism? Finally we address whether the phenotypic changes induced by antiangiogenic treatment are sustained after reimplantation into the nude rat brain. Animals were treated with bevacizumab 10 mg/kg i.v. once weekly. Before sacrifice tumors were evaluated by T1- and T2-weighted MRI, and dynamic contrast enhancement MRI (DCE-MRI). Magnetic resonance spectroscopy (MRS) was applied to assess the metabolic tumor profile after treatment. Immunohistochemical analysis showed a normalization of vascular elements after treatment; these observations were correlated with the MRI observations and with the histological phenotype. In addition, tumor cell proliferation and apoptosis, and the behavior of the invasive, cancer stem-like cell population was quantified.
Accurate determination of response and progression is crucial in the evaluation of new treatments for brain tumor and in the care of individual patients. Macdonald’s criteria (Macdonald DR, et al. J Clin Oncol. 1990;8:1277–1280) are based on measurement of cross-sectional area of enhancing tumor, taking into account changes in steroid dose and neurological status. These criteria have been widely used in clinical trials of brain tumor therapy. Limitations include difficulty in measuring tumors with complex shapes and components (i.e., cystic or postoperative tumors) or indistinct borders, nontumor factors that can produce imaging changes (i.e., enhancement due to postoperative inflammation, infarction, hemorrhage, infection, epilepsy), reaction to local therapies (convection-enhanced delivery, chemotherapy wafers, radiosurgery, etc.), posttreatment changes that may mimic tumor (i.e., pseudoprogression, radiation necrosis), and lack of applicability to nonenhancing tumors (i.e., low-grade gliomas). Antiangiogenic therapies, which may reduce MRI enhancement by restoring the blood-brain barrier, while nonenhancing T2/FLAIR hyperintense tumor may continue to enlarge, highlight the difficulty of assessing response to novel treatments. The RECIST criteria (Therasse P, et al. J Natl Cancer Inst. 2000;92:205–216) use unidimensional tumor measurements and have all the limitations of Macdonald’s criteria. The Response Assessment in Neuro-Oncology (RANO) group is an ongoing unofficial international multidisciplinary consensus-building effort to develop new response criteria to address these issues. Three working committees (high-grade glioma, low-grade glioma, surgical therapies) are outlining strengths and limitations of current response criteria, evaluating new imaging modalities, exploring clinical, neurocognitive, and quality of life end points, and will propose standardized response guidelines. The input of clinicians, investigators, industry, regulatory agencies, and funding bodies will be obtained. Preliminary guidelines were presented at ASCO 2009 in Orlando. This WFNO presentation summarizes the current status of the RANO effort.
While neurosurgeons and radiation oncologists caring for patients with brain tumors carefully preplan interventions using image-based guidance systems and simulators, oncologists choose systemic administration with no quantitative estimates of drug distribution. Clinicians generally assume that small lipid-soluble agents penetrate an intact blood-brain barrier and water-soluble agents enter brain tumors in a manner similar to available radiographic contrast agents. Important drug related factors governing drug entry into brain and brain tumors include molecular weight, charge, protein binding, serum pharmacokinetics, and whether the agent is a substrate for efflux pumps. The integrity of the blood-brain barrier varies in different regions of the tumor and may be influenced by external factors such as radiation, glucocorticoids, and anti-VEGF targeted therapies. Other tumor related variables may include necrosis, cyst, hemorrhage, and local edema, mass effect and pressure.
We have created (patent pending) a Drug Entry Simulator that displays a contrast enhanced MRI of the patient, a serum concentrations of drug over time, and an overlay of drug concentrations over time superimposed on the MRI. The area under the concentration time curve can be assessed for any region of interest. The accuracy of the model is continually refined using results from clinical trials employing microdialysis or surgical biopsies that measure concentrations of systemically administered agents within brain tumors. Examples will be presented to illustrate the information this simulator can provide using water-soluble (methotrexate), more lipid-soluble (temozolomide) chemotherapy, and targeted agents (imatinib).
Pretreatment estimates of drug entry, maximal concentration, time over effective concentration, and area under the concentration time curve should result in more informed choices regarding the use of systemically administered agents in patients with primary and metastatic brain tumors and the best timing of radiation therapy.
The most common mutation of the epidermal growth factor receptor (EGFR) in GBM is EGFRvIII. HGF/SF is the ligand for the receptor tyrosine kinase c-Met, both of which are often coexpressed in GBM. The presence of the HGF/c-Met axis or expression of EGFRvIII each independently enhances GBM growth and invasiveness. Using tyrosine kinase array technology, we show that expression of EGFRvIII in U87MG GBM cells (U87MG.de2-7 cells) leads to coactivation of several receptor tyrosine kinases, including PDGFRβ; and c-Met. A fully human neutralizing antibody directed to HGF (AMG 102) did not inhibit this EGFRvIII mediated activation of c-Met, demonstrating that it is ligand independent. Treatment of parental U87MG xenografts with low doses of AMG 102 resulted in significant inhibition of tumor growth, while U87MG.de2-7 xenografts were profoundly resistant to AMG 102 treatment. Treatment with panitumumab, a fully human antibody directed to the EGFR and EGFRvIII, was able to inhibit the EGFRvIII mediated activation of c-Met and significantly reversed the resistance to AMG 102. Indeed, combination therapy with panitumumab and AMG 102 resulted in a marked increase in tumor cell apoptosis. An EGFRvIII molecule with an active kinase but incapable of autophosphorylation at the five major autophosphorylation sites involved in signal transduction was also able to mediate resistance to AMG 102, suggesting that the activation of c-Met by EGFRvIII occurs via a direct interaction rather than through these docking sites. Thus, expression of EGFRvIII leads to the promiscuous activation of several kinases, causing resistance to ligand inhibitory-based strategies. Since panitumumab can reverse this process, the combination of this antibody with AMG 102 may be an effective treatment for GBM patients.
Clinical experience in malignant gliomas suggests that agents targeting single molecular abnormalities do not control tumor growth. This investigation attempts to improve this situation by testing the efficacy the cytotoxic agent etoposide added to the mTOR inhibitor temsirolimus in an intracranial tumor model. Additionally it evaluates peripheral blood monocyte (PBMC) p70S6 kinase production and activity as a surrogate for brain tumor changes after mTOR inhibitor treatment.
Nude mice with intracranial U87 tumors treated with etoposide or etoposide plus temsirolimus therapy survived significantly longer than a control vehicle group. The addition of temsirolimus to etoposide increased survival over etoposide alone. A substantial in vitro decline in U87 cell phospho-p70S6 kinase protein expression occurred in a dose-related manner to temsirolimus. U87 cell 70S6 kinase activity as measured by a P32-based p70s6 kinase assay declined to 22% of baseline after 30 min in 1 nM temsirolimus and 5% of baseline after 30 min in 10 nM temsirolimus. P70S6 kinase activity in PBMCs, and concurrent intracranial and flank U87 tumors after temsirolimus (10 mg/kg) treatment was 29%, 39%, and 34%, respectively, after 24 h. These declined to 20%, 28%, and 27%, respectively, after 72 h.
In the U87 model used, temsirolimus slightly improves the survival advantage provided by etoposide. Temsirolimus decreases phospho-p70S6 kinase protein and its activity in a dose-dependent manner in vitro. P70S6 kinase activity impairment in U87 intracranial and flank subcutaneous U87 tumors is paralleled by that in those same animals PBMCs. Thus, this assay can be used to monitor biologic effect of this drug in this model. These results warrant (1) assessment of the combination of etoposide and temsirolimus clinically, and (2) correlative confirmation of the use of PBMCs as a surrogate monitor of the management of these patients.
The purpose of this study was to achieve a preliminary assessment of whether glioblastomas (GBMs) with p16 homozygous gene deletions (approximately two-thirds of all GBM) are responsive to cyclin-dependent kinase 4 (CDK-4) and CDK-6 small molecule inhibition, in the context of an orthotopic GBM xenograft therapy response model.
Luciferase-modified U87 GBM cells, which lack p16 function, were injected into the brains of athymic nu/nu mice that were randomized to CDK4/6 inhibitor PD-0332991 versus vehicle only treatment groups, with oral administration of inhibitor 1× daily (150 mg/kg) for 4 consecutive weeks, beginning at day 13 subsequent to tumor cell injection. All animals were monitored 1–2× weekly using bioluminescence imaging (BLI) to assess tumor response to therapy, and each animal was followed until presentation of neurological symptoms indicative of excessive tumor burden that requires euthanasia.
BLI monitoring revealed sustained antitumor activity for the entire period of therapy administration, which was followed by rapid tumor growth after completing the 4-week CDK4/6 inhibitor administration regimen. CDK4 inhibitor extended median survival by 18 days (60%) relative to vehicle treatment, with mean survival extended by greater than 19 days. One CDK4/6 inhibitor treatment group mouse showed no evidence of intracranial tumor following completion of treatment, and was euthanized at day 100 following tumor cell injection. Comparison of MIB-1 staining in tumor from the brain of one mouse that was sacrificed while on therapy versus the MIB-1 staining in tumor from a control (untreated) group mouse revealed a greater than 8-fold reduction in tumor MIB-1 positivity resulting from CDK4/6 inhibitor treatment.
Our results indicate that oral administration of CDK4/6 inhibitor PD-0332991 results in substantial antiproliferative activity against an orthotopic GBM xenograft lacking p16 function, and support further investigation of this therapeutic for treating patients with GBM.
High-grade gliomas (HGG) show an overexpression of TGF-β 2. AP 12009 was developed for targeted suppression of TGF-β 2. Aim of the phase IIb study was to evaluate the efficacy and safety of AP 12009 compared to standard chemotherapy in recurrent or refractory HGG patients.
The phase IIb study G004 was an international, open-label, randomized and active-controlled dose-finding study. Main objective was to compare two doses of AP 12009 (10 μM or 80 μM) and standard chemotherapy (TMZ or PCV) with regard to response rate, survival, and safety. Patients with recurrent or refractory HGG (AA WHO grade III and GBM WHO grade IV, n=145) were randomized into the three treatment groups. AP 12009 was administered intratumorally by convection-enhanced delivery with up to 11 treatment cycles (7 days on, 7 days off/cycle).
In both AP 12009 treatment groups, long-lasting tumor responses were observed in AA as well as in GBM patients. For 10 μM AP 12009–treated AA patients, a significantly better overall response rate (CR+PR) at 14 months compared to control was observed (p=0.034). Also the tumor control rate (CR+PR+SD) differed significantly compared to control (p=0.003). The observed tumor responses correlate with superior overall survival. A median overall survival benefit of 17.4 months was noted for AP 12009–treated AA patients compared to the chemotherapy control. In the GBM subgroup, AP 12009 was as efficacious as standard chemotherapy regarding short-term survival and superior in terms of long-term survival.
The phase IIb study revealed the superior long-term survival of recurrent or refractory HGG patients treated with AP 12009. A pivotal phase III study in AA patients has started and a pivotal study in GBM patients is being planned. Clinical studies with AP 12009 are also ongoing in pancreatic and colorectal carcinoma and malignant melanoma.
Terameprocol is a global transcription inhibitor that affects cell division (Cdc2, G2/M), apoptosis (survivin pathway), drug resistance (MDR1 and PGp), radiation resistance in hypoxia (HIF-1a), and hypoxia responsive genes (VEGF).
A dose escalation study was conducted in heavily pretreated, recurrent, measurable, high-grade gliomas. Terameprocol was administered intravenously for 5 consecutive days each month and discontinued for toxicity or progression. Patients taking and not taking enzyme inducing anticonvulsants (EIAC) were escalated independently.
35 patients were accrued with a median age of 46 (2971), KPS of 80 (60–100), number of prior chemotherapy regimens of 2 (1–6). Forty-three percent had glioblastoma, and 57% had anaplastic gliomas. Pharmacology revealed no differences with coadministration of EIAC. Treatment related grade 3 or 4 toxicities began at 1,700 mg, with one low phosphorus. In eight patients treated at 2,200 mg, there was one metabolic acidosis (secondary to PEG formulation) and one ileus. The drug was reformulated to avoid acidosis and was well tolerated at 1,700 mg, but hypoxia and interstitial nephritis were seen at 2,200 mg. Thus, the recommended daily dose for future studies is 1,700 mg. No radiologic responses were seen but stable disease was noted in 9 of 32 (28%) evaluable patients; five patients (13%) remained on treatment for more than 6 months (6+, 8, 10, 10, and 21+ months). The median survival of this patient population was 5.9 months.
This study established the toxicity profile of terameprocol, determined that EIASD do not affect its pharmacology, and identified the dose to be used in future studies. Encouraging long-term stability was noted in this phase I trial of heavily pretreated recurrent high-grade gliomas. The toxicity profile and observed disease stability suggest that terameprocol could be safely combined with temozolomide and radiation and has the potential to impact survival in newly diagnosed glioblastoma.
Autophagy is organelles and proteins degradation system through the lysosomal machinery in eukaryotic cells. Accumulating evidences show that irradiation or some chemotherapy-agents cause autophagy in various types of tumor cells. However, the autophagy may have opposite effects, that is, cell-killing or cell-protective effect on tumor depending on type of tumor, of anticancer therapies. Therefore, appropriate modification of autophagy, i.e., inhibition of cell-protective autophagy or promotion of cell-killing autophagy, could augment cytotoxicity caused by anticancer therapy in tumor cells. We previously reported temozolomide (TMZ) caused not apoptosis but autophagy in malignant glioma cells. The purpose of this study is to clarify which type of autophagy is induced by TMZ in glioma cells and to determine the appropriate modification of TMZ-induced autophagy. Autophagy assays showed that TMZ induced cell-protective autophagy in glioma cells. When glioma cells were treated with combination of TMZ and RTA203 (RTA), a specific inhibitor of vacuolar type H+-ATPase, TMZ-induced autophagy was blocked at the late stage by abrogating function of lysosome, resulting in augmentation of cytotoxicity. The combination of TMZ and RTA represented more than additive effect in cell viability assay. Apoptosis assays showed RTA alone induces a few apoptosis, while combination of TMZ and RTA significantly increased apoptosis. In animal study, the combination treatment shows significantly longer survival time compared with each single treatment. In conclusion, TMZ induce cell-protective autophagy in malignant glioma cells and inhibition of TMZ-induced autophagy by RTA enhanced cytotoxicity through increase of apoptosis.
Previously, we demonstrated that interferon-β (IFN-β) markedly enhanced chemosensitivity to temozolomide (TMZ) in an in vitro study (Natsume, Cancer Res, 2005); this suggested that one of its major mechanisms is the downregulation of MGMT transcription via p53 induction. This effect was also observed in an experimental animal model (Natsume, Cancer Chemother Pharmacol. 2007). Following those pre-clinical studies, we conducted the Integrated Japanese Multicenter Clinical Trial of IFN-β and TMZ for Glioma in Combination with Radiotherapy (INTEGRA Study).
Twenty-two patients with malignant gliomas (newly diagnosed and recurrent) were enrolled in this study. All the newly diagnosed patients received radiotherapy (fractionated focal irradiation in daily fractions of 2 Gy administered 5 days per week for 6 weeks, i.e., a total of 60 Gy) plus TMZ (75 mg/m2/day, daily from the first to the last day of radiotherapy); and IFN-β (3 MIU/body, administered IV on alternate days during radiotherapy), followed by six cycles of adjuvant TMZ (200 mg/m2/day, on days 1–5, administered every 28 days) and IFN-β (3 MIU/body, administered once every 28 days). Recurrent tumors were treated with TMZ (200 mg/m2/day, on days 1–5, administered every 28 days) and IFN-β (3 MIU/body, administered once every 28 days). The preliminary results reveal that this combination therapy causes minimal toxicity. The most frequent toxic effect is the inhibition of hematopoiesis (leukopenia and neutropenia) that, in most cases, recovered within a month after the cessation of drug administration. In 15 patients with measurable malignant tumors, three had complete response, three had partial response, five had stable disease, and four had progressive disease.
A combination of IFN-β and TMZ tolerated well, and has great potential of enhancing the efficacy of TMZ. Further clinical trials will be warranted.
Concurrent chemoradiotherapy using temozolomide (TMZ) demonstrated a clinical efficacy as primary therapy for the patients with newly diagnosed glioblastoma (GBL), and its effect was influenced by the molecular genetic factors such as MGMT status. The authors postulated a possible racial difference in the response to this therapy and performed a meta-analysis of 103 Korean patients with primary GBL treated from June 2004 to November 2007 in four major institutions (Yonsei University, Ulsan University, Seoul National University, The Catholic University of Korea). The patients received radiotherapy for a total 60 Gy plus continuous daily TMZ (75 mg/m2/day), followed by six cycles of adjuvant TMZ (150–200 mg/m2/day). The median age was 57 years, and 77% of the patients underwent debulking surgery (more than subtotal resection). At a median follow-up of 12 months, the progression-free survival (PFS) was 9 months, and 6-month PFS rate was 66.3%. The median OS was 19 months. The 12-month OS rate was 78.7%, and 24-month OS rate was 48.9%. The response rate was 73% (21 CR and 32 PR of 73 evaluable patients). Median OS (19 months vs. 14 months) as well as 6-month PFS (71% vs. 47%) was significantly improved in the debulking surgery group. No grade 3 or 4 hematologic toxicity was noticed and grade 1 or 2 toxicity developed in only 8.3% of the patients. Concomitant chemoradiotherapy with TMZ appeared to prolong the survival of Korean patients with newly diagnosed GBL and was safe in terms of toxicities and tolerability. However, further investigations are mandatory in larger number of patients, longer term follow-up, and molecular genetic analysis, which have already been undertaken.
Few studies have compared the effects of low-dose frequent administration of chemotherapeutic agents, metronomic therapy, versus bolus chemotherapies on tumor bed vasculature and other tumor characteristics. In this study, we evaluated tumor control, antiangiogenic effects, and tumor cell characteristics in a rodent glioma model in response to these two methods of drug administration.
Rat 9L gliosarcoma cells were stably transduced with the luciferase gene, allowing cells to be monitored using Bioluminescent Imaging (BLI). BLI admits tumor response to treatment to be observed in animals both immediately and repeatedly imaged over time. Fifty thousand tumor cells were inoculated into the right striatums of 15 Fisher 344 rats. Ten days after inoculation, animals were randomly divided into three groups. Group 1 received no treatment. Group 2 received 10 mg of carboplatin via bolus intraperitoneal (i.p.) administration every 2 weeks. Group 3 received 2 mg of carboplatin i.p. five times per week. Tumor growth was measured every 2–3 days. To observe tumor vascularity and reaction to hypoxia, immunohistochemical assessments were carried out on brain specimens excised from rats 4 weeks after initial chemotherapeutic treatment.
Metronomic and bolus administration of carboplatin both controlled tumor growth and vascularity. However, levels of hypoxic reaction and malignant mutation were significantly lower in tumors treated with metronomic therapy; represented by hypoxia-inducible factor-1 alpha (HIF-1α) and cMet protein, respectively.
Recent evidence has shown that tumor cells in a hypoxic condition may survive with malignant changes that appear to be triggered by cMet protein activation. Our results suggest bolus administration of drug is more likely to result in cells becoming hypoxic and acquiring malignant mutations. Metronomic therapy may provide a better alternative for treatment, as it appears to produce lower levels of hypoxia and malignant mutations.
We performed a phase II trial of low-dose continuous (metronomic) treatment using temozolomide for recurrent GBMs.
Temozolomide-refractory patients with GBM who experienced disease recurrence or progression during or after cyclic treatment schedule of temozolomide after surgery and standard radiotherapy were eligible. This phase II trial included two cohorts of patients. The initial cohort, comprising 10 patients, received temozolomide at 40 mg/m2 every day. After this regimen seemed safe and effective, the metronomic schedule was changed to 50 mg/m2 every day. The second cohort, comprising 28 patients, received temozolomide at 50 mg/m2 every day.
The 6-month progression-free survival in all 38 patients was 32.5% (95% CI, 29.3–35.8%) and the 6-month overall survival was 56.0% (95% CI, 36.2–75.8%). One patient developed a grade 3 neutropenia, grade 2 thrombocytopenia in three patients, and grade 2 increase of liver enzyme (GOT/GPT) in three patients. Of all patients included in this study, four patients were withdrawn because of side effects, including sustained hematological disorders, cryptococcal infection, and cellulitis. In a response group, quality of life measured with SF-36 was well preserved, compared with the pretreatment status.
Metronomic treatment of temozolomide is an effective treatment for recurrent GBM that is even refractory to conventional treatment of temozolomide and has acceptable toxicity.
Most patients with newly diagnosed GBM progress after standard RT with concomitant, and adjuvant temozolomide (TMZ) and there is no standard second-line regimen. Continuous dosing and dose intensification of TMZ are thought to reduce levels of O6-methylguanine-DNA methyltransferase (MGMT), which has been associated with TMZ resistance. Altering the schedule of TMZ may be one strategy to reinduce response. A prospective phase II study was conducted to investigate this approach; we report the survival analysis and MGMT data.
Patients with GBM (n=90) and anaplastic glioma (n=30) who failed standard TMZ (5/28-day adjuvant regimen) received continuous dose-intense TMZ 50 mg/m2 for 28/28 days for up to 1 year. The primary end point was 6-month progression-free survival (PFS).
120 patients were enrolled at 11 centers. Patients were divided into four cohorts: recurrence during the first 6 months of adjuvant therapy (early group); after 6 months of therapy (extended group); after stopping adjuvant TMZ (completed group); and anaplastic glioma. No enrollment was permitted in the first 3 months after RT in order to minimize the influence of pseudoprogression. Six-month PFS was 28% (early), 13% (extended), 29% (completed), and 33% (anaplastic). Nonhematological toxicity was mild and easily managed. Progressive lymphopenia was seen in 40% but no serious opportunistic infection was observed. MGMT promoter methylation was tested in 71/120 patients of which 51 could be evaluated. Twenty-two (43%) and 29 (57%) had a methylated and unmethylated promoter, respectively. These two groups were similar in age and surgical resection but more nonmethylated patients were ECOG 0 (45% vs. 32%). They received a median of 6 cycles of continuous TMZ with a median TTP of 4.6 (meth) versus 5.4 (nonmeth) months and median time to death on study of 8.8 (meth) versus 9.4 (nonmeth) months. Five of 17 (29%) MGMT promoter methylated GBM patients reached >6-month PFS compared to 7/22 (32%) nonmethylated.
Continuous dose-intense TMZ 50 mg/m2 administered on a 28/28 day schedule is active and well tolerated after failure of the conventional 5/28 day regimen. Efficacy compares favorably to other commonly used second-line agents. This analysis revealed similar on-treatment time, 6-month PFS, TTP, and survival for GBM patients regardless of MGMT promoter methylation. Continuous TMZ given at first recurrence of GBM may overcome the disadvantage of high MGMT expression. Further studies of this extended TMZ schedule are warranted.
Medulloblastoma is the most common malignant pediatric brain tumor and mechanisms underlying its development are poorly understood. We identified recurrent amplification of the miR-17/92 polycistron proto-oncogene in 6% of pediatric medulloblastomas by high-resolution SNP genotyping arrays and subsequent interphase FISH on a human medulloblastoma tissue microarray. Profiling the expression of 548 mature microRNAs in a series of 90 primary human medulloblastomas revealed that components of the miR-17/92 polycistron are the most highly upregulated microRNAs in medulloblastoma. Expression of miR-17/92 was highest in the subgroup of medulloblastomas associated with activation of the Sonic Hedgehog (Shh) signaling pathway compared to other subgroups of medulloblastoma. Medulloblastomas in which miR-17/92 was upregulated also had elevated levels of MYC/MYCN expression. Consistent with its regulation by Shh, we observed that Shh treatment of primary cerebellar granule neuron precursors (CGNPs), proposed cells-of-origin for the Shh-associated medulloblastomas, resulted in increased miR-17/92 expression. In CGNPs, the Shh effector N-myc, but not Gli1, induced miR-17/92 expression. Ectopic miR-17/92 expression in CGNPs synergized with exogenous Shh and enabled them to proliferate in the absence of Shh. We conclude that Hedgehog signaling promotes the transformation of cerebellar neural precursor cells, driven at least in part by overexpression of the miR-17/92 polycistron. Moreover, in a subset of medulloblastomas, excessive Sonic Hedgehog signaling drives oncogene addiction to miR-17/92, and selects for subsequent genomic amplification of the miR-17/92 polycistron.
The genetic and epigenetic events involved in the initiation and progression of medulloblastoma (MB) are poorly described. To gain a better understanding of MB pathogenesis, we have analyzed 201 primary MBs and 11 MB cell lines using a high-density single nucleotide polymorphism (SNP) array platform. We identified a discrete area of homozygous deletion on 9q31 in a primary MB sample containing nine known genes. As loss of heterozygosity (LOH) at this locus is relatively frequent in MBs, we focused on one of the genes, Kruppel-like factor 4 (KLF4) as a possible tumor suppressor gene in MB. KLF4 is a zinc-finger transcription factor that regulates the expression of genes involved in differentiation and cell-cycle arrest and has recently been shown to undergo promoter methylation and LOH in gastrointestinal cancer. To determine if KLF4 is similarly inactivated by epigenetic mechanisms in MB, we profiled its epigenetic status using multiple complementary techniques. KLF4 expression was restored in MB cell lines after treatment with 5-aza-deoxycytidine, as shown by semiquantitative RT-PCR. Bisulfite sequencing analysis showed methylation of the promoter region of KLF4 in MB cell lines and not in normal cerebellum. We examined the DNA methylation status of KLF4 in 45 primary MBs by methylation-specific PCR and determined that 18% of the tumors showed methylation of the KLF4 promoter region. We reexpressed KLF4 in KLF4-negative human MB cell lines and observed a decreased growth rate of MB cells in vitro, and extended survival of mice harboring intracranial xenotransplanted human MB cells in vivo. Our results suggest that KLF4, which appears to be inactivated frequently in MBs through genetic and/or epigenetic mechanisms, may be a novel factor in MB carcinogenesis.
To review state-of-the-art and relevant advances in the management of ependymomas.
Ependymomas are uncommon neoplasms of the CNS, and may occur either in the brain or the spinal cord. Compared with intracranial ependymomas, spinal ependymomas are less common. Studies performed on genetic changes in ependymoma provide some insight into the pathogenesis and prognostic markers and yield new therapeutic targets. Almost all clinical studies have shown a major impact of extent of resection; thus, a complete resection must be attempted, whenever possible, at first surgery or at reoperation. Involved field radiotherapy is the recommended adjuvant therapy. The lowest limit in age when radiotherapy can be used safely and the need for radiotherapy for completely resected supratentorial nonanaplastic ependymomas are controversial issues that are examined in clinical trials. Craniospinal irradiation is used occasionally for metastatic disease. Chemotherapy is reserved for young infants in order to avoid or delay the use of radiotherapy or in an attempt to facilitate second-look surgery. Data from relevant studies will be presented. Reoperation can provide improvement in progression-free survival. Small series without long-term toxicity data have shown that reirradiation following relapse can occasionally result in long-term survival in relapsed patients. Chemotherapy regimens have been used in relapsed ependymoma with only occasional responses without providing benefit in overall survival. New agents are desperately needed for this rare disease. Molecular biology studies suggest some potential new therapeutic targets.
Infants with malignant embryonal brain tumors have poorer survival compared with older children. Surveillance, Epidemiology and End Results (SEER) Program data on the survival of all children with medulloblastoma/primitive neuroectodermal tumor (PNET) confirm this pattern of poor survival rates in the youngest group of patients, less than 3 years of age, with an overall survival rate of 39% at 5 years and median survival of 1.8 years. Previous studies in large cooperative study groups consistently report this same trend, with collective survival rates ranging from 22% to 47%. There have been divergent strategies across international cooperative groups, strategies such as prolonged postoperative chemotherapy, intraventricular administration of chemotherapy, intensive frontline chemotherapy with further intensification with stem cell rescue, and adjuvant focal conformal radiation. Metastatic disease at diagnosis remains problematic but there is evidence to suggest that intensive therapy with stem cell rescue is encouraging. The incidence of the histologic subtype, desmoplastic medulloblastoma, among the infant population is notable and their potential improved outcome compared to the classic variety is of great interest. The ability to identify infants with medulloblastoma from those with atypical teratoid rhabdoid tumor must be considered a confounding factor. Recent molecular profiling data on medulloblastoma will guide therapy in the future, thus impacting the approach to treating young infants. The predominant issue that surrounds therapy for infants is the well-established risks of radiation therapy in terms of neurocognitive development, far outweighing the benefits of therapeutic response. Focused developmental assessments should be a required aspect in determining the “success” of any future therapy. An international collaborative effort investigating the varying treatment strategies in concert with measuring long-term neurocognitive outcomes is currently under way.
Medulloblastoma is the commonest malignant brain tumor in childhood with an annual incidence of around 400 cases per year in the United States. In recent years significant strides have been made in the treatment of this tumor with long-term survival of 85% in children with average risk disease and up to 70% in those with disseminated tumor. However, the outcomes in infants and those with recurrent disease remain poor. While high-dose chemotherapy with stem cell rescue has been the main stay of salvage therapy at recurrence with or without radiation therapy, it is far from clear if this approach can be successfully employed in all children at recurrence. A profusion of studies have focused on the pathology and molecular characteristics of this tumor. Identification of histologic subtypes, alterations in the Sonic Hedgehog (SHH) and WNT signaling pathways, overexpression of ERBB2, OTX-2 gene amplification, and the recent efforts to identify the putative tumor initiating cell in medulloblastoma have helped in understanding the biologic basis of this disease, assessing prognosis, and to begin devising a more biology-based therapy of children with this tumor. A significant proportion of children suffer from leptomeningeal disease (LMD) at relapse. Therapeutic options are currently limited for this group and more effective strategies are needed. This discussion will focus on the recent advances in the treatment of children with recurrent medulloblastoma with specific emphasis on HDC with stem cell rescue, treatment of LMD, molecular targeted therapies, immunotherapy, and other innovative strategies that are already being tested in the clinic or actively being explored in the laboratory.
Measuring children has proven to be a challenge. We have previously described the FMH, a questionnaire for pediatric brain tumor protocols. Here we report the experience in the treatment protocol HIT-GBM-C, which included intensive simultaneous radiochemotherapy followed by further chemotherapy for high-grade glioma and diffuse intrinsic pontine glioma.
The FMH (“Fertigkeitenskala Muenster/Heidelberg (FMH)”) is a quantitative measure of health status, which includes 56 simple questions between 2 and 15 words to be answered with yes or no. The number of positive answers is translated in age-dependent percentiles (FMH%). Physicians were also asked to rank the handicap in a 5-point scale: normal, mild handicap, age normal activity severely reduced but not in bed, in bed but not in ICU, and ICU. No financial compensation was offered.
Questionnaires were available from 50 of 97 eligible patients. Five patients scored initially more than 40 on FMH%, four of which survived. Sixteen patients scored less than initially 39%, and 15 of them died. The physicians judged the patients as normal in 20.4% initially, which decreases to 16.3% in the middle of the protocol. In reverse only 16.3% had activity severely reduced, which increased to 30.6% after receiving the half of the treatment. The FMH% correlated well with the physicians judgments (p < 0.005).
The FMH scale is valid as it correlated with the physician assessment. Patients with poor performance at treatment start had a poor prognosis. The number of questionnaires sent back was insufficient for answering a treatment related question. If a treatment comparison is to be answered, an incentive needs to be provided.
The Pediatric Quality of Life Inventory, PedsQL, is a widely used modular instrument for measuring health-related quality of life in children aged 2–18. The PedsQL Brain Tumor Module comprises six subscales: cognitive problems, CP; pain and hurt, PH; movement and balance, MB; procedural anxiety, PA; nausea, NA; and worry, WO. Here, we developed the Japanese version of the PedsQL Brain Tumor Module.
Translation equivalence was confirmed using the standard back-translation method, and content validity was assured by cognitive debriefing tests of both children and parents separately. Self-reported or interviewer-administered questionnaires including both the PedsQL Generic Core Scales and the State-Trait Anxiety Inventory for Children, STAIC, were completed by 137 children with brain tumors, 90 self-reported, 47 interviewer-administered, and 166 parents (all self-reported). Average child age was 9.8 years. Forty-seven children had embryonal tumors (29.0%), 36 had germ cell tumors (22.2%), 25 had high-grade glioma (15.4%), and 39 had low-grade glioma (24.1%). Sixty-five children (39.2%) were under treatment.
High internal consistency (Cronbach’s α=0.75–0.84 for children, 0.80–0.96 for parents) and sufficient test-retest reliability (intraclass correlation coefficient [ICC]=0.67–0.77 for children, 0.74–0.95 for parents) were demonstrated for all subscales except child-reported PH (α=0.50, ICC=0.45). Factorial validity was supported through exploratory factor analysis (factor-item correlation = 0.33–0.96 for children, 0.55–1.00 for parents). Known-groups validity confirmed that CP was sensitive for developmental disorders, MB for paresis, and NA for currently undergoing chemotherapy. Convergent and discriminant validity with the PedsQL Generic Core Scales and STAIC were acceptable. Child-parent correlation was 0.41–0.65 for five of the six subscales and 0.18 for WO.
The Japanese version of the PedsQL Brain Tumor Module is suitable for the measurement of health-related quality of life of children with brain tumors in clinical trials and research.
The NIH Toolbox, part of the NIH Blueprint for Neuroscience Research, seeks to develop brief, comprehensive assessment tools measuring motor, cognitive, sensory and emotional health. The ultimate goals of the Toolbox include the provision of (1) uniformity among measures across disease groups, and (2) measures that demonstrate stable psychometric properties across the life span. Upon completion, the Toolbox will be available for use in longitudinal epidemiologic studies and clinical trials for people ages 3–85. In this paper, we describe the Toolbox developing process using motor and cognitive function as examples.
We first conducted both online surveys and in-depth interviews with clinical research experts in each domain. Specifically, experts were asked to nominate essential constructs to be assessed for motor and cognitive function and rank them based on conceptual and clinical relevance. A follow-up consensus meeting was held to finalize the constructs to be included in the Toolbox.
147 responded to an online survey; additional 12 and 8, respectively, completed in-depth interviews for motor and cognition. The results were reviewed by the NIH project team and external advisory group. The follow-up consensus meeting led to the inclusion of locomotion, balance, dexterity, strength, and endurance for the motor domain, and attention, episodic memory, executive function, language, processing speed and working memory for the cognition domain.
We are currently revising existing instruments and developing additional instruments to assess the selected constructs, which will be completed by spring 2009. Normative data from the U.S. general population will be completed in 2010, and the final Toolbox will be available in 2011. The final Toolbox will be a valuable research resource, improving the ability to compare results from different studies. Use of Toolbox assessments holds the promise of improving our understanding of the overall health of neuro-oncology patients.
We report prospective neuropsychological and neuroendocrine function in young patients with residual/recurrent craniopharyngioma treated with stereotactic conformal radiotherapy (SCRT).
Eighteen patients (median age 14 years) treated with SCRT underwent prospective detailed serial evaluation of neurocognitive, Barthel index (BI), LOTCA and neuroendocrine function.
Among 18 patients (14 males, 4 females; mean follow-up 28.4 months), 16 (89%) were controlled and two (11%) patients died with disease progression. At pre-RT, 16 (89%) patients had full-scale IQ (FSIQ) below normal. Pre-RT mean verbal IQ (VQ), performance IQ (PQ) and FSIQ were 74.12, 82.42 and 76.53, respectively. Mean FSIQ at 6 months and 2 years were 80.5 and 77.7; PQ score 87.5 and 86.6; VQ score 86.33 and 82.88, respectively. Mean trait anxiety (C2) score was poor at baseline (29.67) and improved to 21.83 at 2-year follow-up. Mean BI score at baseline, 6 months, and 2 years were 95.64, 99.81, and 100, respectively. LOTCA scores were maintained at 2 years in orientation, spatial-perception, thinking, and attention-concentration domain. At baseline, 83.3% (15) patients had hormone deficiency in at least one axis. Growth hormone, corticosteroid, thyroid and sex hormone axis impairment were in 67% (12), 61% (11), 33.3% (6), and 5% (1) patients, respectively. Pre-RT BI score was significantly lower in visually handicapped (p=0.007), low KPS score (0.004), poor neurological function status (NPS) (p=0.014) and in patients with severe hydrocephalus (p=0.031). FSIQ score was significantly lower in patients with thyroid replacement (p=0.05). LOTCA scores were lower in visually impaired (p=0.071) and in patients with poor neurological function (p=0.021).
Craniopharyngioma patients have poor neuropsychological and endocrinal function even before starting RT. However, there is no further decline up to 2 years follow-up. Factors other than RT such as tumor, patient-related factors and surgery may influence baseline neurocognitive function.
Between 17% and 50% of people with brain tumor (BT) display some degree of challenging behavior (CB) including executive impairment, behavioral disturbance, or social/emotional dysfunction. However, little information on management strategies is available for patients and/or their caregivers to assist in coping with these problems. A Web-based review identified 17 key English language information resources (Web-based or hard copy) on BT from across Australasia, North America, and the United Kingdom. Although 47% contained lists of cognitive/behavioral impairments, 71% provided either no or minimal information (1–2 sentences) on management strategies. This project aimed to develop patient-caregiver-oriented fact sheets to address this information gap.
A literature review regarding behavioral/cognitive impairments after BT was conducted. Resources developed for other neurologic groups were reviewed. Combining data from these sources with the clinical experience of the project team, draft fact sheets were devised and presented to a focus group of patients and caregivers for feedback. Drafts were also circulated to a state multidisciplinary health provider BT network convened by the New South Wales Cancer Institute. Readability statistics were also calculated to ensure that the resources met public health standards for comprehension.
Sixteen fact sheets (Web-based and hard copy) were developed addressing apathy, anger, stress/anxiety, fatigue, high-level thinking problems (executive impairment), inappropriate social/sexual behavior, lability, perseveration, disorientation/confusion, egocentricity, impulsiveness, depression, neglecting personal care, memory difficulties, attention/concentration, and communication. Each sheet contains a definition of the problem, prevalence statistics, symptom list, deidentified case study, management strategies for the person with BT and caregiver(s), key questions to ask health professionals and Web-links to further information. After only 3 months of release, substantive requests for these resources from across Australia have been received.
Results suggest that these fact sheets fill an important information gap, providing an important resource for patients, caregivers, and health care providers.
Few studies have documented behavioral disturbance, impairments in executive functioning as well as emotional/social dysfunction (as distinct from mood disturbance) associated with brain tumor (BT). This study investigated (1) the prevalence of such challenging behaviors (CBs) and (2) potential correlates among patient, tumor, treatment variables and functional status.
A consecutive series of 54 patients with BT recruited from two Sydney-based hospitals completed a series of self-report surveys documenting patient impairment. Thirty-seven caregivers derived from this series also participated. Measures of CBs included the Frontal Systems Behavior Rating Scale (FRSBE), Overt Behavior Scale (OBS), and Emotional/Social Dysfunction Questionnaire (ESDQ). Functional measures included the Sydney Psychosocial Reintegration Scale (SPRS; self-report scale documenting function in occupational, relationship, and independent living domains) and clinician-rated KPS (31% at 90–100; 34% at 80; 35% at 50–70).
Median patient age was 51 years (range, 18–91) and time since diagnosis was 4 months (range, 1–82). Patients had high-grade (39%), low-grade (22%), or benign tumors (39%), and 60% underwent radiotherapy. Patient self-report found prevalence rates of CBs ranging from 17% to 50% across various domains including verbal aggression, physical aggression, lack of initiation, disinhibition, apathy and executive impairments. No consistent pattern of correlates was found between CBs (patient or caregiver report) and patient demographics, tumor features (benign vs. malignant diagnosis, grade, location), treatment timing or modality (surgery, radiotherapy or chemotherapy). However, functional status (KPS and SPRS scores) demonstrated significant correlations (0.30–0.50 range) with various caregiver-rated CB variables including FRSBE apathy and executive impairment subscales; ESDQ emotional dysfunction, fatigue, inappropriate behavior, and poor insight subscales, and OBS total score.
Patients’ functional status may be the most useful indicator of challenging behaviors after brain tumor. KPS could be linked to further screening questions to help identify and subsequently manage challenging behaviors in brain tumor patients.
CNS embryonal tumors are devastating cancers in children, consisting of medulloblastomas, supratentorial primitive neuroectodermal tumors (sPNETs), and atypical teratoid/rhabdoid tumors. Based on the genetic alterations found in humans, multiple models of human CNS embryonal tumors have been developed in genetically engineered mice. The objectives of this study were to create novel mouse models for discovering efficient therapy and to understand brain tumor biology.
The RCAS/tv-a system was used for modeling brain tumors in p53 null mice. The experimental retrovirus RCAS that has an inserted oncogene c-Myc or β-catenin was injected into neonatal mouse brains. Transgenic mice expressing RCAS receptor tv-a under the control of the GFAP promoter (Gtv-a mice) allow cell-type-specific RCAS infection and gene transfer in vivo. Symptomatic mice were sacrificed and examined histologically. We also reviewed published reports of the molecular and cytogenetic abnormalities in human CNS embryonal tumors.
sPNETs were generated in Gtv-a p53/− mice by forced c-Myc expression with 34% incidence. Combined β-catenin activation with c-Myc promoted tumor progression and induced divergent differentiation. This sPNET was histologically similar to human large cell/anaplastic variant of medulloblastomas and when injected into cerebellum, large cell/anaplastic medulloblastomas formed at higher incidence. By reviewing published reports, we found that human sPNETs seemed to share the several genetic abnormalities with other CNS embryonal tumors. The strongest similarity of the molecular/cytogenetic profile was observed between sPNETs and large cell/anaplastic medulloblastomas in terms of chromosome 1q gain, c-Myc and N-Myc amplification, β-catenin stabilization, and p53 inactivation.
These findings indicate that common genetic abnormalities are seen in variants of human CNS embryonal tumors, and multiple histologic variants of these tumors can be generated from a single set of genetic abnormalities in mice. These data provide insight into the biology and classification of CNS embryonal tumors.
Genotyping of 212 medulloblastomas (MBs) using high-resolution SNP arrays identified novel amplifications and homozygous deletions, including recurrent, mutually exclusive, highly focal genetic events in genes targeting histone lysine methylation, particularly histone 3, lysine 9 (H3K9). Proper control of histone lysine methylation is critical for transcriptional regulation and plays an essential role in the differentiation of stem and progenitor cell populations. We identified two tumors with focal homozygous deletions of EHMT1 on 9q34, whose protein product is an H3K9 methyltransferase. Similarly, we found focal deletion of SMYD4 on 17p13, another histone lysine methyltransferase (3 tumors). Conversely, we found amplification of JMJD2C (7 tumors) and JMJD2B (5 tumors) through the combined analysis of our SNP array data and FISH probing of an MB tissue microarray. The JMJD2 family genes encode H3K9 demethylases, whose effect is opposite to that of EHMT1 on H3K9. The ability of EHMT1 to methylate H3K9 is blocked by acetylation. We found amplification of MYST3 (four tumors), a histone 3 lysine acetyltransferase. We also found deletions in polycomb genes that interpret the state of H3K9 methylation including L3MBTL3 (one tumor), L3MBTL2 (four tumors), and SCML2 (three tumors). Reexpression of L3MBTL3 in an MB cell line where it is homozygously deleted resulted in decreased growth, and cell cycle redistribution. Reexpression of L3MBTL3 also increased H3K9 dimethylation of E2F6 target gene promoters as demonstrated by chromatin immunoprecipitation. Additionally, retroviral-mediated overexpression of JMJD2C in cerebellar granule cell progenitors caused a reduction in the levels of H3K9 dimethylation. Copy number aberrations of genes with critical roles in writing, reading, removing, and blocking the state of histone lysine methylation, particularly H3K9, suggest that defective control of the histone code contributes to the pathogenesis of medulloblastoma.
We modified the Sleeping Beauty (SB) system so that SB is mobilized in the EGL under the influence of the Math1 enhancer. Nearly 100% of Math1SB11, SB transposon donor, Ptch± mice develop medulloblastoma with leptomeningeal metastases by 3 months, where-as 10–20% of Ptch± mice develop localized medulloblastoma by 8 months. Mice carrying Math1SB11, or transposon donor, or both do not develop cancer. We sequenced more than one million insertions mapping to more than 158,000 sites from more than 140 SB-induced primary medulloblastomas and leptomeningeal mets using 454 technology. Statistical analysis identified 29 common insertion sites (CISs) in primaries and 44 CISs in metastases including well-known medulloblastoma genes such as CREBBP and PTEN, as well as other cancer genes such as FHIT, CACNA2D1, and ERAS. CISs also frequently targeted genes involved in neural stem cell biology such as NFIA and NFIB. All primary/met pairs arose from a common ancestor as they share common clonal insertions. Some clonal events present in the mets are present in only a small subclone of the primary, supporting the clonal selection theory of metastasis and suggesting that some of these events affect metastasis virulence genes. Similarly, some highly clonal events seen in the primary tumor are not seen in the metastases. These likely represent events that have arisen postmetastatic dispersion, demonstrating that parallel evolution also occurs. Most important, comparison of insertions from primary tumors and matched spinal and frontal lobe leptomeningeal metastases reveals that in some cases the mets share a clonal insertion that is only present in a very minor subclone of the primary tumor. This highly suggests that leptomeningeal dissemination only occurs once. Our results suggest reasons for the failure of medulloblastoma therapies developed against targets discovered in the primary tumor. Further understanding of the clonal hierarchy of medulloblastoma metastases will help design future effective therapies.
Primitive neuroectodermal tumors (PNET) represent one of the largest groups of malignant pediatric brain tumors and include medulloblastoma as well as CNS-PNET arising in the cerebral hemispheres. Despite histologic similarities, CNS-PNET exhibits worse outcomes than medulloblastoma. In contrasts to medulloblastoma, CNS-PNET remains poorly characterized with limited studies of less than 10 cases reported to date. Prior studies indicate CNS-PNET differs genetically from medulloblastoma, however, specific biological features of CNS-PNET remain to be defined. In this study, we performed high-resolution gene expression and copy number analyses on 40 primary CNS-PNET using the Illumina Human-6 V2 BeadChip expression and the Affymetrix 500-K GeneChip Mapping arrays. Data analysis revealed absence of isochromosome 17q and MYCC amplification, and only one case of MYCN amplification in 40 CNS-PNET. One of five of CNS-PNET had chr22q loss that did not involve the rhabdoid tumor IN1/SMARCB1 locus. These data support the distinct identity of CNS-PNET. Copy number analysis respectively identified 28 and 13 regions of focal deletions and amplification in CNS-PNET involving known and novel tumor suppressor and oncogenic loci. Remarkably although CNS-PNET exhibited heterogeneity in copy number profiles, integrated gene expression and DNA copy number analysis demonstrated segregation of CNS-PNET into two major subgroups with specific molecular and clinicopathologic characteristics. Notably, a distinct subgroup, characterized by recurrent copy number alterations of chr 2, 3, and 19 was associated with younger age at diagnosis (<4 years, p=0.026) and significantly poorer survival (mean of 4±1.3 vs. 44±12.8 months; p=0.0001). Interestingly, gene expression profiles of these aggressive CNS-PNET indicated significant enrichment (p=0.00001) of genes associated with pluripotency and downregulation of proapoptotic loci. In summary, through integrated genomic analysis we have identified genetic features that define subgroups of CNS-PNET. Our data lead us to propose that aggressive CNS-PNET are characterized by early developmental lineages and prosurvival phenotypes.
Intracranial germ cell tumors (GCTs) are relatively rare brain tumors affecting primarily adolescents. Although the survival rate of pure germinoma is excellent (>90%), optimal volume and dose of radiation or the role of chemotherapy is not clear, and the risk factors of relapse is unknown. In contrast, the outcome of nongerminomatous germ cell tumors (NGGCTs) is poor (30–40% survival) and the standard treatment for NGGCTs remains controversial. The lack of progress in improving the outcome of these patients is due to our poor understanding of the biology of intracranial GCTs. To better characterize underlying genetic alteration of intracranial GCTs, copy number aberration (CNA) analysis using genomewide genotyping microarray was performed. We have analyzed the DNA samples of 55 intracranial GCTs by GeneChip Human Mapping 100K Array. After applying stringent quality control criteria, 30 cases with both tumor and matched blood samples (18 pure germinomas and 12 NGGCTs) were selected for further analysis using the Copy Number Analyzer for GeneChip (CNAG). Overall, copy number gains are observed more frequently than losses. The most frequently observed CNAs include gain of chromosomes 1q, 2, 7, 8, 12p, 21, and X as well as loss of chromosomes 11q and 13. Validation of these results is ongoing using real-time quantitative PCR method. The CNAs that are specific to histologic subtypes were also identified. Loss of chromosomes 11q and 13 were specific to germinomas. Gain of chromosome 2 was observed more frequently in germinomas, whereas gain of chromosomes 1q and 12p were more frequently observed in NGGCTs. We identified multiple frequently altered loci of intracranial GCTs genome. Similar CNAs have been observed in intracranial GCTs as well as extracranial counterparts. These findings will serve as the focus of further search for the genes involved in tumorigenesis of intracranial GCTs.
Temozolomide is now widely used as the first-line chemotherapy for gliomas. However, efficacy of temozolomide is virtually based on reports on adult gliomas, and recent results for pediatric gliomas have not necessarily demonstrated similar efficacy of temozolomide shown in adult gliomas. Pediatric gliomas are characterized by distinct chromosomal abnormality from adult gliomas, and it is of interest whether any genetic or epigenetic characteristics of pediatric gliomas cause distinct sensitivity to temozolomide.
We examined several factors that affect sensitivity to temozolomide in pediatric gliomas with comparison to adult gliomas. We studied expression of MGMT (O6-methylguanine methyltransferase) and DNA mismatch repair gene MSH 6 (mutS homolog 6) by immunohistochemistry, presence or absence of promoter methylation of the MGMT gene by methylation-specific PCR, and genetic aberrations by comparative genomic hybridization.
Expression of MGMT in pediatric astrocytic tumors was significantly higher than those of adult tumors by immunohistochemistry. Moreover promoter region of MGMT gene was unmethylated in almost all pediatric astrocytic tumors by methylation-specific PCR. However, there was no significant difference in expression of MSH6 between pediatric and adult tumors. Codeletion of 1p and 19q was not observed in pediatric gliomas, and the most frequent aberration in those tumors was gain on 1q, with 10 of 23 cases showing the chromosomal abnormality. Gain on 1q showed tendency toward poor survival and aggressive behavior of the tumors regardless of grade.
Pediatric gliomas appear to have distinct genetic profiling associated with temozolomide resistance. Higher expression of MGMT, absence of codeletion of 1p and 19q, and frequent gain on 1q in pediatric gliomas might be related to differential sensitivity to temozolomide from their adult counterparts.
Pilocytic astrocytomas (PAs) are the most frequently occurring brain tumor in 5–19 year-olds. Little is known about the genetic alterations underlying their development. We carried out a genomewide genetic analysis in 44 PAs to elucidate the molecular mechanisms underlying the development of these tumors. Array-CGH analyses using a combination of a 1-Mb and a chromosome 7 tile path BAC array as well as a custom-designed oligonucleotide array identified two types of tandem duplication, one being very common and the other rare. The former spanned approximately 2 Mb at 7q34 and was found in 29 out of 44 (66%) PAs. This rearrangement was not observed in any of 244 higher-grade astrocytic tumors. PCR confirmed that the duplication resulted in a novel fusion gene between KIAA1549 and BRAF. The fusion gene encodes a novel, in-frame fusion protein that lacks the BRAF regulatory domain, has constitutive kinase activity, and confers anchorage-independent growth on transfected NIH3T3 cells. The other tandem duplication was found at 3p25 in a single tumor without the BRAF fusion gene. This resulted in an in-frame oncogenic fusion between SRGAP3 and RAF1, which lacks the RAF1 autoregulatory amino-terminus but retains the kinase domain. It shows elevated kinase activity compared with wild-type RAF1 and transforms NIH3T3 cells. Among the remaining 14 PAs without the BRAF or RAF1 fusion genes, three tumors had BRAF mutations, including one with a novel 3-bp insertion. Three other patients without RAF gene alteration had clinical features of NF1. Mutations of the NF1 gene in this syndrome lead to hyperactive Ras signaling and RAF activation. Our results thus show that the great majority of PAs (36 out of 44, 82%) have alterations of the MAPK pathway. This highlights the central importance of this pathway in PA tumorigenesis and underlines its potential as a therapeutic target.
Patient-specific virtual controls (VCs) are created for 26 patients with histologically diagnosed glioblastoma multiforme allowing testing and validation of a biomathematical model for glioma response to radiation therapy (RT). Individualized glioma growth kinetics, described by the mathematical model parameters quantifying net rates of invasion and proliferation of glioma cells, are determined from serial MRI prior to treatment for each patient.1 Knowledge of these patient-specific growth kinetics allows for simulation of untreated VCs for each patient against which therapeutic response can be assessed. Serial posttreatment imaging is used to assess response to RT. The net proliferation rate is strongly correlated (r=0.84, p<0.01) with the radiation efficacy (α from the classical linear-quadratic radiobiology model). This correlation was tested as a means of predicting radiation efficacy from the pretreatment proliferation rate using a leave-one-out-cross-validation analysis yielding a mean error between simulated and actual post-RT tumor radius of 3 mm. In a survival analysis, pretreatment growth kinetics and radiation response metrics was a significant predictor of survival even when controlling for the prognostic value of recursive partitioning (RPA) class established by the radiation therapy oncology group (RTOG). The fundamental result from the investigation is the ability to predict response to therapy prior to treatment using a variety of clinically established and novel biomathematical model-based metrics. Ultimately, directing treatment with knowledge of VCs allows for improvement in the design of individualized patient therapies, such as RT, as well as a novel technique to assess response for any therapies since an untreated VC provides comparison for the actual tumor behavior.
To prospectively analyze neurological status (NS) and quality of life (QL) of patients affected with BM treated with WBI.
Patients with prognostic RPA classes I, II, IIIa-b (Lutterbach, 02), any age, KPS 50–100, no chemotherapy during WBI, and informed consent were included. Exclusion criteria were class IIIc, poor medical or psychosocial conditions, or leptomeningeal dissemination. NS was based on modified cerebral stroke scale with the following items: headache, seizures, speech, visual, motor, brainstem, sensitive, gaiting disorders, and MMSE. EORTC QLQ C-30 and BN-20 were used for each patient. WBI consisted of a total dose of 30 Gy/10 days. Changes of NS were defined as follows: Clinical improvement was the reduction of 25% of scoring and equal or less dose of DXM; clinical progression was the increase of 25% of scoring or more dose of DXM or both, and clinical stability was defined as other situations.
278 patients were registered; 103 (37%) accomplished inclusion criteria. Most patients had lung (57%) or breast cancers (23%). Ninety-eight of 103 patients finalized WBI. At the first month, 80.6% of intention-to-analyze cases (79 patients) fulfilled neurological evaluation. NS improvement or stabilization at the first month was 64%. Maintained good MMSE (29–30) influenced NS (p=0.04); none of the following variables did: age, gender, histology, number of BM, BM location, RPA class, radiological response, dose of DXM, or baseline NS. Median progression to NS deterioration was 2.3 m. QL improved in emotional (significant) and cognitive fields (trend), and worsened in physical and fatigue fields (both significant).
(1) 63% of the entire cohort were not available for the study. (2) Only 80.6% were able to be scored at the first month. (3) NS improvement or stabilization at the first month was 64%. (4) Neurocognitive function could determine NS. (5) QL did not decrease after WBI.
To establish feasibility of doing intensity-modulated radiation therapy (IMRT) for craniospinal irradiation (CSI) using Helical Tomo-Therapy (IMRT_Tomo) and report challenges in clinical implementation. A dosimetric comparison with conventional IMRT (IMRT_LA) and three-dimensional conformal radiotherapy (3DCRT) is also reported.
Initially, computed tomography (CT) data sets of four previously treated patients of medulloblastoma were used to generate 3DCRT; IMRT_LA; and IMRT_Tomo plans for CSI. Standard CSI dose of 35 Gy/21 fractions was prescribed to planning target volume (PTV). Subsequently, five patients were treated on an ongoing prospective protocol of IMRT for CSI using Helical TomoTherapy.
The mean volume of PTV receiving at least 95% of prescribed dose (V95%) was more than 98% in all techniques. All techniques resulted in comparable dose homogeneity index (DHI) for PTV_brain. For PTV_spine, IMRT_Tomo achieved highest mean DHI of 0.96 compared to 0.91 for IMRT_LA and 0.84 for 3DCRT. The best dose conformity index was achieved by IMRT_Tomo for PTV_brain (0.96) and IMRT_LA for PTV_spine (0.83). IMRT_Tomo was superior in terms of reduction of maximum and mean doses to almost all OARs. The average beam-on time was significantly longer in IMRT_Tomo. Issues in implementation included whole-body immobilization, areas to be imaged daily, coregistration efficiency, intrafraction motion, and impact of differential shifts, which were handled using appropriate methodology.
Helical TomoTherapy seems to be ideally suited for CSI avoiding junctions, field-matching and abutment dosimetry. It is favorable in terms of target coverage, dose homogeneity, conformity, and OAR sparing. Challenges in successful implementation of IMRT_Tomo for CSI can be circumvented provided they are preempted during the planning phase. The longer beam-on time in IMRT_Tomo raises concerns about intrafraction motion and secondary carcinogenesis.
To report the effect of age and RT doses to volumes of different normal structures on neurocognitive outcomes in young patients with benign and low-grade brain tumors treated prospectively with stereotactic conformal radiation therapy (SCRT).
Twenty-eight patients (median age 13 years; 23 male and 5 female) with low-grade and benign residual/progressive brain tumors (10 craniopharyngioma, 8 cerebellar astrocytoma, 8 cerebral low-grade glioma) were treated with SCRT to a dose of 54 Gy/30 fractions/6 weeks with at least 2 years of follow-up were analyzed. Prospective neuropsychological assessment was done at baseline preradiotherapy (RT) and at subsequent follow-up with an age-appropriate neuropsychological battery of tests. The change in intelligent quotient (IQ) scores was correlated with the age and dose-volumes to normal structures using logistic regression analysis.
While the overall mean full-scale IQ (FSIQ) at baseline pre-RT was 80 and remained unchanged at 2-year follow-up, a third of patients had >10% drop in FSIQ over baseline. Patients <15 years of age had a significantly higher chance of developing >10% drop in FSIQ (53% vs. 10%, p=0.03). Comparison of dosimetric data in patients showing a significant drop in IQ with patients with maintained IQ revealed that patients receiving >42.5 Gy (80% of the prescribed dose) to >13% of volume (p=0.048) and >27 Gy (50% of the prescribed dose) to >50% of the volume of left temporal lobe were the ones to show significant drop in FSIQ (p=0.06). Calculation of the RT doses to other normal structures including supratentorial brain and right temporal lobe did not reveal any significant correlation.
Our detailed dosimetric data show younger age and radiotherapy doses to left temporal lobe to be predictors of neurocognitive decline, and could be used as possible dose constraints for high-precision RT planning.
Pseudoprogression has been recognized and widely accepted in the treatments for malignant gliomas, as transient increase in volume of enhanced area just after the chemoradiotherapy, especially using temozolomide. We experienced the similar phenomenon in the treatments of malignant gliomas and meningiomas using boron neutron capture therapy (BNCT). BNCT is high linear energy transfer, cell-selective particle radiation. Here we introduce the representative cases and analyze this pathogenesis. Fifty-two cases of malignant glioma and 13 cases of malignant meningioma who were treated by BNCT were reviewed retrospectively mainly in MRI. Twelve out of 52 malignant gliomas and 3 out of 13 malignant meningiomas showed transient increase of enhanced volume in MRI within 3 months after BNCT, without any chemotherapy. For these cases peak tumor doses were given by BNCT between 234.7 and 1774.2 Gy-Equivalent as given in daily 2 Gy by fractionated XRT. In these cases, six glioma cases underwent operation for the suspicion of relapse. In histology, most parts of the specimen showed necrosis with small amount of tumor cell residual. Ki-67 labeling index showed decreased positivity compared with previous samples of the individuals. Fluoride-labeled boronophenylalanine positron emission tomography (PET) was applied in five and two cases of malignant gliomas and meningiomas, respectively at the time of transient increase of lesions. These PETs showed decreased lesion/normal brain ratio in all cases compared with those obtained prior to BNCT. With surgery or without surgery, all lesions were decreased or stable in size during observation. Transient increase in enhanced volume in malignant gliomas and meningiomas immediately after BNCT seemed to be pesudo-progression. This pathogenesis was considered as treatment-related intra-tumoral necrosis in subacute phase after BNCT. Pseudoprogression has been recognized as good prognosis landmark with intensive treatments. Therefore, pseudoprogression in BNCT suggested potential therapeutic effects of this particle radiation methodology.
Recently, new radiation therapies, such as IMRT, BNCT and other particle radiation modalities using high absorbed doses to tumor tissue have been developed for patients with malignant gliomas, mainly glioblastomas (GB), with good clinical results. Meanwhile, clinically, the differential diagnosis of the tumor progression (TP) of GB and radiation necrosis (RN) has become important, as has the assessment of tumor activity after adjuvant therapy. We determined how to evaluate tumor activity before and/or after radiation therapy using positron emission tomography (PET) with boronophenylalanine (BPA) as a tracer. F-BPA-PET images were obtained from histologically verified 38 GB, 8 RN, and 5 RN cases with partial residual tumors. The lesion/normal (L/N) ratios for these groups were 4.2±1.4, 1.5±0.3, and 2.0±0.3, respectively. Also, repeat PET imaging was found to be useful for evaluating changes in GB-associated tumor activity with respect to the treatment received. For the treatment, we experienced 20 RN cases in these 3 years. Anticoagulants and vitamin E alone improved or stabilized RN in one-third of the RN cases, especially for relatively small lesions in gadolinium (Gd) enhancement. Also this treatment was effective for prevention of RN. In 11 RN cases, lesionectomy was applied with 5-ALA–guided microsurgery. Four out of 11 cases showed improved KPS 1 month after surgery, while two cases showed aggravation and the remaining five were unchanged and stable in KPS. Steroids could be decreased in amount in all 11 cases. 5-ALA–guided resection showed a clear boundary between the necrotic tissue and surrounding normal brain in surgery. Recently we experienced the treatment of three medically refractory RN in eloquent area using humanized anti-VEGF antibody, bevacisumab. This treatment showed rapid shrinkage of perilesional edema and enhanced lesion in all cases only by single intravenous administration of bevacisumab at 5 mg/kg. Let me show the representative cases in my talk.
Our data suggested that deletion of chromosome locus 6q22–23 was a characteristic of poor prognosis PCNSL patients. In this report we describe a pilot study of aCGH in combination with FISH to confirm deletion status and assess copy member changes. Frozen tumor samples were retrieved from Mayo Tumor Registry (IRB 08-001933). All cases were reconfirmed as PCNSL; as having diagnostic tissue without hemorrhage or necrosis; and as EBV-encoded RNA (EBER) negative. Each case was interrogated using well-validated and published protocols for FISH, aCGH; and immunohistochemistry (IHC) for protein tyrosine phosphatase κ (PTPRK), the gene product of 6q22–23. The cohort comprised two male and three female immunocompetent patients. Two of the four patients with sufficient frozen tumor samples had 6q deletion by aCGH, one complete and one partial. The patient with complete deletion had homozygous deletion by FISH. The patient with partial 6q deletion by array CGH had no FISH abnormalities. Other recurrent abnormalities included partial 1p and partial 1q loss, partial 6p loss, and partial or complete gains on chromosomes 7, 12, 13, 18, and 19q; as well as deletions affecting well-characterized tumor suppressor genes p16 and Rb1. This combination of complimentary techniques was chosen on the premise that it would better ascertain whole genome chromosomal imbalances, ploidy, and chromosome integrity. This combination strategy is the most logical next step in the determination of genomic signatures consistent with chromosomal instability mechanisms in PCNSL. This approach may provide important insights concerning the mechanisms responsible for generating complex genomes. The resulting phenotypic diversity can generate tumors with a propensity for an aggressive disease course. A better understanding of the underlying mechanisms leading to PCNSL development could result in the identification of prognostic markers and therapeutic targets.
Although most primary CNS lymphomas (PCNSL) are diffuse large B-cell lymphomas (DLBCL), biological characteristics of PCNSL are not fully understood. We have attempted to investigate the origin of PCNSL with respect to the stage of B-cell differentiation and challenged the hypothesis that the tumor’s stage of differentiation could impact the patient’s prognosis. Clinical data of 33 patients treated under the diagnosis of DLBCL of the CNS were retrospectively collected. In 32 patients, immunohistochemical study for the following antigen was performed for classifying B-cell differentiation: CD10, BCL-6, MUM1, and CD138. The median PFS of all 33 patients was 17.2 months and median OS was 63.3 months. Forty percent of the cases showed CD10/BCL-6+/MUM1+, indicating that the tumor was originating from the transition stage from germinal center B-cell (GCB) to post-GCB. Twenty-five percent of cases were CD10/BCL-6/MUM1+, suggesting post-GCB origin. Twenty-five percent of cases were CD10+ or CD10/BCL-6+/MUM1, suggesting GCB origin. When B-cell stages of the tumor were compared to clinical outcomes, late B-cell origins (post-GCB) showed a statistically significant better prognosis than early B-cell origins (pure GCB type and transition stage from GCB to post-GCB type) with a median PFS of 42.4 and 9.6 months, respectively. We suggest that PCNSL often derive from B-cell in a differentiation stage from GCB to post-GCB and that classifying PCNSL as above could be beneficial for predicting the patient’s outcome and provide more insights into the biology of PCNSL.
Neurolymphomatosis (NL) is defined as a clinical neuropathy induced by infiltration of peripheral nerves, nerve roots, plexus, or cranial nerves by any hematologic malignancies. Given the rarity of this condition, a retrospective review with collaboration across many centers is under way.
To describe the demographics, diagnostic details, management, and outcome of NL.
40 patients (diagnosed 1993–2008) were retrospectively assembled from 10 centers in five countries.
Median age was 55.5 years and 60% were males. NL was related to systemic lymphoma in 80%, to acute lymphoblastic lymphoma in 12.5% and to primary CNS lymphoma in 7.5%. It occurred as the initial manifestation of the malignancy in 30% of cases. The most common manifestations were related to peripheral nerves involvement observed in 55%, followed by spinal nerve roots (42.5%), cranial nerves (37.5%) and plexus (35%) with multiple site symptomatology in 50% and painful neuropathy in 78%. Diagnostic evaluations included imaging, CSF studies (36 patients) and biopsies (19 patients). Imaging studies suggested the diagnosis in 78% of cases with 70% (26/37) of MRI and 82% (14/17) of CT-PET being positive. CSF cytology was positive in 36% and nerve biopsy confirmed the diagnosis in 16/19 (84%). Therapeutic approaches in 39 patients included high-dose methotrexate and/or Ara-c in 61.5%, intra-CSF chemotherapy in 50% and radiotherapy in 36%. Response to treatment by MRI in 21/39 cases revealed 13 (62%) CR, 3 (14%) PR, 1 (5%) SD, and 4 (19%) PD. Response by neurological score in 36/39 (92%) patients showed improvement/stabilization in 27 (75%). Median survival was 15 months with 31% alive at 56 months.
NL is a challenging diagnosis associated with an overall dismal prognosis. Contemporary imaging techniques frequently detect the relevant neural invasion. An aggressive multimodality therapy can prevent neurological deterioration and is associated with a prolonged survival in 30% of the affected patients.
High-dose methotrexate (HD-MTX) has been considered as the most effective agent in the initial treatment of primary CNS lymphoma (PCNSL). However, some patients have refractory lymphoma to initial treatment, and other patients relapse easily. Since treatment options for patients with progressive or recurrent PCNSL are limited, their prognoses are very poor. Temozolomide is a well-tolerated oral alkylating agent and easy to permeate the blood-brain barrier. The effectiveness of temozolomide is reported on not only malignant glioma, but also recurrent PCNSL. In this study, we evaluated the results of treatment with temozolomide for the patients with refractory or relapsed PCNSL.
Immunocompetent patients with refractory or relapsed PCNSL were eligible after the initial treatment with HD-MTX (3.5 g/m2) with or without irradiation. All of the patients were treated with temozolomide 150–200 mg/m2, for 5 days in 28 days. This treatment was continued until disease progression.
Seventeen patients with a median age of 68 years (ranged from 49 to 88 years) were included in this study. Four patients were used as second line treatment, 10 as third, and 3 as fourth. Five complete remissions (median treatment 12 times; ranged from 4 to 35 times), five partial response and stable disease and seven progressions were observed. No major toxicities were observed, apart from grade 3 thrombocytopenia in one patient.
Although some patients were elderly and had been heavily pretreated, temozolomide yielded 29% objective responses and was well tolerated without any major toxicity.
High-grade glioma (HGG) continues to have a dismal prognosis with limited survival despite multimodal therapy. This study assesses the impact of the delivery of the HSV-tk gene into the healthy tumor bed of resected HGG using a serotype 5 adenoviral vector (sitim-agene ceradenovec). This gene induces the expression of thymidine kinase, which phosphorylates the systemically administered ganciclovir to be effective locally against malignant cells.
Two hundred fifty patients with operable HGG were recruited from 38 centers in Europe and Israel between November 2005 and April 2007. Patients were randomized to receive standard therapy, with or without sitimagene ceradenovec injections. On days 5–19 postoperatively, the active group received ganciclovir (5 mg/kg i.v. twice daily). Adjuvant treatment with temozolomide was permitted, but not obligatory, depending on local treatment practice. Patients are currently being followed for reintervention-free survival, overall survival (OS) and MRI progression, as well as standard safety monitoring. The primary outcome measure is time from surgery until reintervention or death. The principal secondary outcome was time to all-cause mortality—OS.
The study was completed in May 2008. Updated final results from the OS outcome will be available in early Q1 2009 and will be presented at the meeting, along with demographics of the study population, efficacy of the therapy (as Kaplan-Meier survival plots and statistical analyses) and the overall time-based safety profile.
Conclusions will be derived from the results to provide an overall risk and benefit assessment of this form of gene therapy in operable HGG.
The cancer stem cell hypothesis implies that intratumor heterogeneity is created by rare cancer stem cells. These cells have the capacity to differentiate into various tumor cell types as well as to self renew. In brain tumors, CD133 has been proposed as a marker for cancer stem cells, although this has been challenged by a series of studies demonstrating tumor formation from CD133-negative glioma cells as well.
To further elucidate the heterogeneity seen in Glioblastoma Multiforme (GBM), we created lineage specific lentiviral reporter constructs that expressed GFP under the Nestin, glial fibrillary acidic protein (GFAP), or neuron specific enolase (NSE) promoter. The U373 and U87GBM cell lines were infected with these constructs, and separated with simultaneous removal of CD133-positive cells using fluorescence-activated cell sorting (FACS).
This way, we obtained glioma subpopulations expressing different cell type markers, and subsequently implanted these cells in NOD-SCID mice. Finally, we repeated the experiments using two primary GBM biopsies instead of cell lines. In the cell lines, all glioma cell populations gave rise to tumors at similar rates. For the biopsies, both nestin + and nestin − cells grew tumors, while none of the lineage specific cells (GFAP and NSE) produced tumors.
We conclude that different glioma subpopulations exhibit tumor initiating ability, although glioma cells expressing mature markers may be relatively less tumorigenic than other glioma cell populations.
We successfully apply the monolayer culture method to establish glioma stem cell line from surgical sample (paper in submission). We present our novel technique for establishment of GBM stem cell line and some particular feature of its xenografts.
We used three glioblastoma surgical samples. We cultured cells from each sample on monolayer methods with EGF/FGF without serum. After conformation every cell line fulfilled the criteria of glioma stem cell, we injected cells into NOD/SCID mouse brain and observed the pattern of invasion and differentiation of tumor cells pathologically on its xenograft.
We established three GBM stem cell lines from three samples successfully. Every line fulfill the criteria of glioma stem cell. They express stem cell markers such as Sox2, Nestin and have ability of multilineage differentiation. All lines have tumor initiating ability in vivo. Although ratio of CD44 expression is almost 100% in all lines, interestingly the ratio of CD133 expression varied as follows: 96.5%, 1.4%, 43.6%. Xenografts have glioblastoma specific feature such as necrosis/pseudopallisading. The undifferentiated GFAP(–)Nestin(+) tumor cell showed strong invasive capacity, invaded contralaterally even in the early stage. Undifferentiated GFAP(–)Nestin(+) tumor cell favored invasion to subventricular zone (stem cell niche). In all xenografts, invasive front is formed by undifferentiated GFAP(–)Nestin(+) tumor cell. The higher CD133 expressed line has the stronger invasive capacity. Lower CD133 line formed somewhat well-margined mass. Differentiated GFAP(+) tumor cell remained at injection site and expanded and formed tumor bulk.
The novel monolayer culture method is useful to establish glioma stem cell line and provides enough chances to study glioma stem cell from surgical sample. The xenografts of GBM stem cell lines recaptured the pathological feature of original sample patient. Undifferentiated stem-like tumor cells have strong invasion ability and formed invasive front of tumor.
Disseminating disease of high-grade gliomas is difficult to treat. We examined, in this study, the therapeutic effect of intrathecal administration of mesenchymal stem cells transduced with herpes simplex virus-thymidine kinase gene (MSCtk) in experimental leptomeningeal glioma model.
Leptomeningeal glioma model was produced by injecting C6 glioma cells into the cisterna magna in Sprague-Dawley rats. Mesenchymal stem cells were isolated from 6-week-old rats and MSCtk cells were established using retrovirus transduction system. First, to examine in vivo bystander effect, rats were intrathecally coinjected with a mixture of MSCtk and C6 cells and then, intraperitoneally administered with ganciclovir (GCV) or saline for 10 days (coinjection model). Next, to examine the therapeutic effect of MSCtk/GCV therapy, MSCtk cells were intrathecally administered 1 day after C6 injection and then, GCV or saline was administered (treatment model). Tumor volume was measured on day 14 in one group and survival time was measured in the other group for each model.
GCV administration significantly reduced tumor sizes both in the coinjection model (0.41±0.22 vs. 3.10±0.97 mm2, p<0.01) and in the treatment model (0.73±0.29 vs. 2.84±0.82 mm2, p<0.01). Survival was also significantly prolonged in GCV group both in the coinjection model (29.17±3.27 vs. 18.83±0.79 days, p<0.01) and in the treatment model (21.50±1.48 vs. 17.17±0.48 days, p<0.01).
This study provided a novel treatment strategy for leptomeningeal glioma dissemination using intrathecal MSCtk injection followed by systemic GCV administration.
The purpose of this study was to evaluate the effects of chloroquine on glioma cell viability in vitro and in vivo, and to explore the therapeutic potential of chloroquine as an adjunct to glioma therapy.
Human glioma cell lines with known p53 status or primary cultures derived from surgical specimens of human glioblastomas were propagated in the presence of serum or under serum-free conditions as neurospheres. Radioresistant populations were selected from neurosphere cultures of glioma stem-like cells by repetitive exposures to clinically relevant doses of radiation. The tumorigenic potential was evaluated in an orthotopic glioma mouse model. Assessment of cell death was performed by trypan blue exclusion, by immunofluorescence detection of cells positive for activated caspase-3 or TUNEL (TdT-mediated dUTP-nick end labeling). The transcriptional activity of p53 was assessed by Western blot.
ClQ potently induces death response in glioma cells in vitro and inhibits growth of experimental gliomas in an orthotopic glioma mouse model. We demonstrate that accumulation of the p53 protein, activation of p53 transcriptional response and induction of p53-dependent apoptosis are the mechanisms underlying induction of cell death by ClQ in cells with wild-type p53. Unlike DNA damage, chloroquine does not induce p53 posttranslational modifications contributing to p53 stabilization through phosphorylation. In addition to its p53-activating effects, chloroquine is also able to suppress glioma cell growth via p53-independent mechanisms. We found that chloroquine inhibits growth of glioma cells that express mutant p53, possess stem-like properties and are capable of establishing highly invasive tumors. Furthermore, chloroquine treatment renders radioresistant cell populations selected from glioma stem-like cultures susceptible to radiation, and chemotherapy with BCNU or temozolomide.
Chloroquine possesses therapeutic potential and can be envisaged as an adjunct for glioma therapy.
A distinct subpopulation of tumorigenic cells exhibiting properties of neural stem cells, generally termed as brain tumor initiating cells (BTICs), has recently emerged as the major cellular component determining the invasive and radioresistant phenotype of malignant brain tumors. There is an urgent need for experimental models that would recapitulate the hallmark features of human glioblastoma multiforme (GBM) such as histological heterogeneity and a highly invasive, radioresistant phenotype. The purpose of this study was to develop an invasive orthotopic GBM mouse model using human glioma cells with defined genetic background and the hallmark features of stem-like BTICs.
Stem-like BTICs were selected from the human glioma cell line G112 and propagated in the presence of serum or under serum-free conditions as floating neurospheres. Cell viability was assessed by using a neurosphere assay and trypan blue exclusion. The BTICs stem-like phenotype was assessed in vitro by immunofluorescence staining of cultured cells and in vivo by immunohistochemical examinations.
The human glioma cell line G112 shows a stem-like BTIC phenotype under serum-free culture conditions. Genetic analyses indicate that G112BTIC derive from a single genotype. The G112BTIC neurosphere cultures are capable of long-term self-renewal, show immunoreactivity for neural stem cell markers nestin and CD133 and display a radioresistant phenotype in vitro and in vivo. Tumors formed by G112BTIC cells show an extremely aggressive growth pattern, exhibit the hallmark features of human GBM and show an impaired radioresponse. In contrast, non-BTIC tumors show significant reduction upon treatment with identical radiation doses.
The G112BTIC glioma model provides an adequate experimental platform for the testing of potential therapeutic efficacy of existing treatments and developing new strategies to target radio- and chemoresistant glioma cell populations.
Surgically induced supplementary motor area (SMA) deficiency syndrome has been reported, and reversibility of the SMA deficiency syndrome has been emphasized. In this report, we aimed to confirm functional results after the resection of gliomas involving the SMA.
Twelve patients with gliomas of the SMA treated by surgery between 2002 and 2007 were studied. The nondominant side was affected in seven cases and the dominant side in five. Degree and duration of postoperative deficits were evaluated and correlated with the extent of tumor resection on the postoperative MRI and intraoperative neuromonitoring findings.
Postoperatively, motor deficits were evident in 11 of 12 patients and speech deficits in four of five. Recovery of motor function began between a few hours and 7 days after the surgery, and significantly improved within 2 months. Five patients in whom the posterior resection margin was at a distance of less than 0.4 mm from the precentral sulcus experienced a persistent drop in skilled motor function. Two patients who complained of cognitive dysfunction subjectively after the surgery were assessed with WAIS-III, and a diminishing in processing speed was recognized in both patients.
Preserved muscle tone and recovery from the distal part of the limbs were characteristic of SMA deficiency syndrome compared with damage to the pyramidal tract. A good correlation existed between the severity and duration of postoperative motor deficits and the distance of the resection margin from the precentral sulcus. The follow-up of cognitive dysfunction, characterized by a diminishing in processing speed, is necessary.
The extent of resection in low-grade glioma (LGG) still remains controversial to be concerned with patient outcome. A part of LGGs have progressive growth and malignant transformation; however, there are no diagnostic tool to distinguish their clinical natures. In this study, we assess the diagnostic role of radical resection in the management of LGG.
A series of 40 patients has been diagnosed as LGG, World Health Organization (WHO) grade II, by modern diagnostic imaging since we introduced awake surgery into LGG resection in 2004. The high-intensity area in FLAIR images was defined as LGG. The tumor was radically resected within the limits avoiding neurological deficits. The widespread histopathological analyses were performed at the various loci of resected LGGs.
Two patients were diagnosed with stereotactic biopsy because of difficulties of tumor resection. Although the diagnostic images show characteristic features of LGG, the widespread histopathological studies demonstrated local malignancy in a part of LGGs. In the series of 40 LGGs, there are 9 diffuse astrocytomas, 10 diffuse astrocytomas with mitosis, 5 anaplastic astrocytomas, 4 dysembryoplastic neuroepithelial tumors, 5 oligodendrogliomas, 2 oligoastrocytomas, and 1 cortical dysplasia.
The widespread histopathological study with radical resection of LGGs helps us to detect local malignancy. This finding suggests that extensive surgical resection and widespread histopathological study in large specimens are recommended to decide the management of LGG.