Year: 2016
Type: article
Abstract: The classification of ependymomas, tumors with a variety of histological subtypes and anatomic locations, is beginning to yield to molecular classification based upon genomic analyses. Specifically, posterior fossa ependymomas can be sorted into two subtypes based upon genetic and methylation profiles: EPN_PFA and EPN_PFB (also known as Group A and Group B). In contrast to Group B, Group A ependymomas tend to occur in young children, to be laterally located, to be more invasive with higher metastatic potential, and to have increased DNA methylation and worse prognosis. Historically, standard therapy for pediatric posterior fossa ependymomas includes maximal safe tumor resection followed by involved field radiation therapy. Extent of resection is recognized as the most important prognostic factor. To determine the relative contributions to prognosis of molecular tumor classification, extent of resection, patient age and radiation therapy, an international collaboration amassed a large retrospective cohort of these rare tumors. Taylor and colleagues collected clinical data and performed unsupervised hierarchical clustering of genome-wide methylation arrays on 4 cohorts comprising a total of 820 posterior fossa ependymomas. Eighty-three percent of cases were Group A and 17% Group B; children younger than five years were almost exclusively Group A, adults were 89% Group B, and children aged 10–17 were evenly divided between the two groups. Multivariate modeling demonstrated that molecular subgroup was by far the strongest predictor of overall and progression-free survival (OS and PFS), with extent of resection and adjuvant radiation therapy also being significant factors. Group A patients had a much worse outcome independent of age (OS-10 yr approximately 50%) and did especially poorly with subtotal resection; postoperative radiation in this setting was of limited benefit. In contrast, Group B patients with gross total resection did very well in general, and some may be cured even without adjuvant radiation. Moreover, some Group B patients who recurred following gross total resection were apparently salvaged with re-resection followed by radiation. Subtotally resected Group B patients had a PFS-10 yr of 50%. These results highlight the prognostic importance of molecular subtyping and suggest this analysis should be performed in patients older than five years. As extent of resection is prognostic in both molecular subgroups, gross total resection should be attempted whenever possible. The authors suggest that equipoise exists to justify a trial randomizing Group B patients who have undergone gross total tumor resection to either adjuvant radiation therapy versus observation, although the logistical challenges for such a trial in a rare tumor are daunting. GSK3β is a ubiquitous multi-functional protein kinase that regulates a broad range of cellular processes and has been investigated as a therapeutic target in diabetes and Alzheimer’s disease (AD).1 In medulloblastoma, GSK3β plays a tumor-suppressive role via phosphorylation and degradation of N-myc.2 GSK3β is also well known for regulation of β-catenin stability via phosphorylation, which promotes its ubiquitylation and subsequent degradation in the absence of Wnt signaling. In contrast, GSK3β may have an oncogenic role in some tumor types and has also been implicated in glioblastoma migration and invasion.3 The mechanisms involved in the oncogenic properties of GSK3β remain poorly defined. A recent study by Zhou et al.4 identifies a pro-tumorigenic role of GSK3β in glioblastoma via regulation of the levels of the histone demethylase KDM1A (lysine-specific histone demethylase – LSD1). The authors initially showed that nuclear GSK3β and KDM1A expression are directly correlated in a panel of glioma cell lines and also in Gsk3β-/-- MEFs. Interestingly, in contrast with the role of GSK3β in causing the degradation of substrates such as β-catenin, the authors show that GSK3β phosphorylates KDM1A at serine 683, leading to its deubiquitylation by USP22 and subsequent stabilization. Increased levels of KDM1A cause demethylation of histone H3K4 and increased glioblastoma stem cell self-renewal. Demethylated H3K4 reduces transcription of genes involved in stem cell differentiation, including BMP2, CDKN1A, and GATA6, thereby promoting stemness. Thus, nuclear GSK3β promotes stemness via a transcriptional program mediated by KDM1A. The importance of this pathway is supported by studies on patient specimens showing that KDM1A expression correlates with nuclear GSK-3β and USP22 levels. In vivo, KDM1A and GSK3β shRNA slowed the growth of intracranial xenografts, and this was rescued by coexpression of shRNA-resistant KDM1A and S683D-mutated KDM1A, respectively. Finally, and perhaps most significantly in terms of translational impact, the authors used tideglusib, a GSK-3 inhibitor recently investigated in clinical trials for Alzheimer’s disease (5), to treat mice bearing patient-derived glioblastoma stem-like cell xenografts. This led to downregulation of KDM1A, inhibition of tumor growth, and sensitization to temozolomide and also significantly improved survival. These data point towards novel and translatable strategies for future clinical trials in glioblastoma. Management of brain metastases (BM) remains a clinically significant challenge notwithstanding the palliative utility of surgery and radiotherapy. Lung cancer comprises the largest cohort of patients with BM, accounting for 50–60% overall.1 Twenty to thirty percent of patients with stage IV non-small cell lung cancer (NSCLC) develop BM. BM occur synchronously with initial diagnosis of NSCLC in 20% of patients. Multiple subsets of NSCLC are currently recognized and are defined by oncogenic driver mutations. ALK (anaplastic lymphoma kinase) rearrangements, most commonly EML4-ALK fusion, occur in approximately 2% to 7% of patients with NSCLC, and recent studies have demonstrated the success of therapeutic approaches targeting the ALK protein in this patient group.2,3 Two recent trials suggest not only successful treatment of systemic disease but as well BM with ALK inhibitors.3,4 Crizotinib, an oral, small-molecule inhibitor of ALK, MET, and ROS1 receptor tyrosine kinases, has demonstrated significant antitumor efficacy in patients with advanced ALK-positive NSCLC. The phase III study PROFILE 1014 recently showed superior progression-free survival (PFS), objective response rate (ORR), and patient-reported outcomes (PRO) for crizotinib versus pemetrexed/platinum combination chemotherapy in patients with previously untreated advanced ALK-positive NSCLC, establishing crizotinib as standard first-line therapy in this population.2 An analysis by Solomon of patients with BM (all treated and stable as required by protocol) in the PROFILE 1014 study (23% of 343 patients enrolled) demonstrated a significantly higher intracranial disease control rate in patients with BM treated with crizotinib compared with combination chemotherapy.3 Intracranial efficacy was prospectively assessed using protocol-specified brain imaging at baseline and then every 6 or 12 weeks in patients with or without known BM. Nonetheless, the CNS remains a frequent site of initial crizotinib failure in 20–30% of patients, likely due in part to limited brain penetration and development of resistance (for example, the ALK kinase domain gatekeeper mutation L1196 or amplification of the ALK gene).5 When CNS progression occurs with ongoing extracranial disease control, it is rational to use locally ablative therapy with radiotherapy to control CNS disease, allowing effective systemic crizotinib treatment to be continued beyond intracranial disease progression. In patients with disease progression both in the CNS and systemically, second-generation ALK inhibitors such as ceritinib, alectinib or brigatinib may be employed, all with improved brain penetration. Crino conducted a phase II study, ASCEND-2, of ceritinib in patients with ALK-rearranged NSCLC previously treated with chemotherapy and crizotinib.4 One hundred forty patients were enrolled, of which 100 had baseline BM, 20 of those with untreated measurable target lesions. In patients with BM, the intracranial ORR was 45%. These data, together with emerging evidence for intracranial efficacy with other second-generation ALK inhibitors, suggest that patients who experience progression in the brain during crizotinib treatment may benefit from subsequent treatment with an alternative ALK-targeted therapeutic. The use of successive ALK inhibitors in conjunction with brain-directed radiotherapy has significantly improved the survival (median overall survival 26–49.5 months) and treatment paradigm for this molecularly defined subset of patients with NSCLC and BM. As the most common intracranial neoplasm in adults,1 meningiomas are not well understood. What we do know are that most cases are sporadic whereas others are familial. Patients who have undergone radiation, especially to the brain or scalp, or those who have experienced traumatic brain injury are more at risk. Moreover, patients with NF2 have an increased risk of developing meningiomas. Fortunately, the majority of meningiomas are benign; however, a small percentage are either atypical or malignant. So, with the incidence of meningiomas on the rise, we ask ourselves: is this simply a reflection of increasing exposure to environmental risk factors (i.e. nurture), underlying genetics (i.e. nature), or perhaps just more sensitive methods of detection? Current research shows a waxing interest in meningiomas, highlighting new questions and challenges for this clinically diverse group of tumors as they relate to pathology, genetics, and treatment. In order to begin to parse out some of these questions, Clark and colleagues analyzed 775 meningiomas.2 Interestingly, they identified recurrent somatic mutations in POLR2A, a gene encoding the largest subunit of RNA polymerase II, which is responsible for synthesizing messenger RNA in eukaryotic cells. These mutations essentially hijack this enzyme and drive tumorigenesis. Additionally, POLR2A mutations were exclusively detected in WHO grade 1 meningiomas whereas none were found in 159 atypical or malignant meningiomas screened for these mutations, identifying a discreet subset of benign meningiomas. POLR2A-mutant meningiomas were also shown to preferentially arise in the midline from the region of the tuberculum sellae. Clark’s work, along with that of other researchers, supports the occurrence of somatic alterations in SMARCB1 within meningiomas,3,4 but its co-occurrence with somatic NF2 mutations and observation that the majority of these NF2/SMARCB1 tumors tend to originate from the midline dura of the falx, unlike those NF2 meningiomas lacking SMARCB1, are new. The authors also identified meningioma subtype-specific super-enhancers, most notably a super-enhancer covering the WNT6 and WNT10A gene locus with WNT6 significantly overexpressed, whereas the super-enhancer encompassing ZIC1 and another member of this protein family found in close genomic configuration ZIC4 were lost in POLR2A-mutant meningiomas. Notably, both WNT and ZIC genes play critical roles in cortical development through regulating the proliferation of meningeal cells. Further analysis of meningioma drivers revealed involvement of other key embryonic signaling mechanisms fundamental to formation of the meninges. Perhaps benign meningiomas do, in fact, arise from early progenitors as a result of dysregulated embryogenesis or regression of specialized mature cells to a simpler state reminiscent of stem cells, supporting the notion of “nature” over “nurture” in this veritable argument of whether genes or environment matter more. With the incidence of meningiomas on the rise, however, an interplay between the two is more likely at work. Further research is needed to answer this question as it relates to benign, malignant, and atypical meningiomas, but the authors provide an intriguing start to the conversation. Many tumors upregulate the non-stereotypical cyclin-dependent kinase 5 (Cdk5), which has been shown to play diverse roles in promoting cancer, making it an attractive therapeutic target. However, the role of Cdk5 in cancer immune surveillance is not well established. Using medulloblastoma (MB) as a prototype, Dorand et al. show that Cdk5 regulates and promotes antitumor immunity. The authors first examined MB cell lines and clinical samples and identified high expression of Cdk5 and its co-activators, p35 and p39. Disrupting Cdk5 in MB cell lines using shRNA or CRSPR-Cas9 approaches caused a significant reduction of tumor volume and increased tumor free survival in mice. Depletion of CD4+ T cells using a blocking antibody alone or in combination with CD8 blocking antibodies completely rescued tumor formation in these mice, suggesting that this effect was dictated by T cell infiltration. Mice re-challenged with a lethal dose of control MM cells were tumor free, confirming that Cdk5 plays a role in T cell-dependent immune surveillance and memory. To explore the mechanism by which this occurs, the authors examined the correlation of Cdk5 with PD-L1, an immune inhibitory ligand known to cause immune suppression. A direct association was observed between Cdk5 and PD-L1 in clinical specimens, and the expression of both baseline and IFN-γ-induced PD-L1 were reduced in Cdk5-deficient cells. Chemical inhibition of Cdk5 using roscovitine also led to the same response. The authors next examined the expression of IRF1 and IRF2 regulators of the IFN-γ signaling pathway. They found elevated expression of both proteins, which also act as co-repressors for PD-L1. A phosphoproteomic screen identified phosphorylation of IRF2BP2 as a top hit in Cdk5-deficient cells, which explains the reduction of PD-L1 in these cells. Intracranial administration of these cells also confirmed the dependency on Cdk5 for overall anti-immune responses, microglia, and monocyte infiltration. In summary, while the exact mechanistic link between Cdk5 and IRF2BP2 phosphorylation is unclear, this study uncovers a novel role for Cdk5 in anti-cancer immunity via regulation of PD-L1. With broad-spectrum Cdk inhibitors such as dinaciclib undergoing clinical trials, a subset of MB patients with high Cdk5 expression can be a target group for these agents.
Source: Neuro-Oncology
Language: English
Subfield: Genetics
Field: Medicine
Domain: Health Sciences
Open Access status: bronze
OpenAlex ID: W4254018199