MethPed: a DNA methylation classifier tool for the identification of pediatric brain tumor subtypes

Long-read sequencing for brain tumors

Brain tumors and genomics have a long-standing history given that glioblastoma was the first cancer studied by the cancer genome atlas. The numerous and continuous advances through the decades in sequencing technologies have aided in the advanced molecular characterization of brain tumors for diagnosis, prognosis, and treatment. Since the implementation of molecular biomarkers by the WHO CNS in 2016, the genomics of brain tumors has been integrated into diagnostic criteria. Long-read sequencing, also known as third generation sequencing, is an emerging technique that allows for the sequencing of longer DNA segments leading to improved detection of structural variants and epigenetics. These capabilities are opening a way for better characterization of brain tumors. Here, we present a comprehensive summary of the state of the art of third-generation sequencing in the application for brain tumor diagnosis, prognosis, and treatment. We discuss the advantages and potential new implementations of long-read sequencing into clinical paradigms for neuro-oncology patients.

Clear cell chondrosarcoma: a review of clinicopathologic characteristics, differential diagnoses, and patient management

Clear cell chondrosarcoma (CCC), an extremely rare primary bone tumor, is currently classified by the World Health Organization as a low-grade malignant cartilaginous neoplasm. Clinically, CCC occurs primarily in males with a peak incidence in the third to fifth decades of life, and occasionally, it presents in skeletally immature patients. Unlike conventional chondrosarcoma, CCC has a predilection for the epiphysis of long bones and often displays radiologic features reminiscent of chondroblastoma. The recommended treatment is wide operative resection. CCC has a local recurrence rate of approximately 30%, and nearly 20% cases metastasize mainly to bone and lung often a decade after surgical intervention. Incomplete excision or curettage is associated with a high rate of recurrence. Histologically, the process is characterized by infiltrative lobules and sheets of round to oval cells with abundant cleared cytoplasm and well-defined cell borders associated with trabecula of osteoid and woven bone, scattered osteoclasts, and foci of conventional low-grade chondrosarcoma in about one-half of cases. Correlation with clinical and radiologic characteristics, such as epiphyseal location and young patient age, assists in establishing a correct diagnosis. Pathologic diagnosis of CCC is complicated by the low diagnostic accuracy of core needle biopsy, overlapping histologic features with other matrix-rich primary bone tumors, and a lack of a specific immunohistochemical and molecular profile. DNA methylation-based profiling classifier (sarcoma classifier) is one recent technologic advancement that may help to confirm the histopathological diagnosis of CCC or indicate the need for thorough reassessment in cases where results contradict previous conventional findings.

Sex differences in methylation profiles are apparent in medulloblastoma, particularly among SHH tumors

Background

Medulloblastoma, the most common malignant pediatric brain tumor, displays marked sex differences in prevalence of the four main molecular subgroups: SHH, WNT, Group 3 and Group 4. Males are more frequently diagnosed with SHH, Group 3 and 4 tumors, which have worse prognoses than WNT tumors. Little is known about sex differences in methylation profiles within subgroups.

Methods

Using publicly available methylation data (Illumina HumanMethylation450K array), we compared beta values for males versus females. Differentially methylated positions (DMP) by sex within medulloblastoma subgroups were identified on the autosomes. DMPs were mapped to genes and Reactome pathway analysis was run by subgroup. Kaplan-Meier survival curves (Log-Rank p-values) were assessed for each sex within subgroup. MethylCIBERSORT was used to investigate the tumor microenvironment using deconvolution to estimate the abundances of immune cell types using DNA methylation data.

Results

There were statistically significant differences in sex by medulloblastoma subgroups (chi-squared p-value=0.00004): Group 3 (n=144; 65% male), Group 4 (n=326; 67% male), SHH (n=223; 57% male) and WNT (n=70; 41% male). Females had worse survival than males for SHH (p-value=0.02). DMPs by sex were identified within subgroups: SHH (n=131), Group 4 (n=29), Group 3 (n=19), and WNT (n=16) and validated in an independent dataset. Unsupervised hierarchical clustering showed that sex-DMPs in SHH did not correlate with other tumor attributes. Ten genes with sex DMPs (RFTN1, C1orf103, FKBP1B, COL25A1, NPDC1, B3GNT1, FOXN3, RNASEH2C, TLE1, and PHF17) were shared across subgroups. Significant pathways (p<0.05) associated with DMPs were identified for SHH (n=22) and Group 4 (n=4) and included signaling pathways for RET proto-oncogene, advanced glycosylation end product receptor, regulation of KIT, neurotrophic receptors, NOTCH, and TGF-β. In SHH, we identified DMPs in four genes (CDK6, COL25A1, MMP16, PRIM2) that encode proteins which are the target of therapies in clinical trials for other cancers. There were few sex differences in immune cell composition within tumor subgroups.

Conclusion

There are sexually dimorphic methylation profiles for SHH medulloblastoma where survival differences were observed. Sex-specific therapies in medulloblastoma may impact outcomes.

Methylation Profiling in Diffuse Gliomas: Diagnostic Value and Considerations

Diffuse gliomas cause significant morbidity across all age groups, despite decades of intensive research efforts. Here, we review the differences in diffuse gliomas in adults and children, as well as the World Health Organisation (WHO) 2021 classification of these tumours. We explain how DNA methylation-based classification works and list the methylation-based tumour types and subclasses for adult and paediatric diffuse gliomas. The benefits and utility of methylation-based classification in diffuse gliomas demonstrated to date are described. This entails the identification of novel tumour types/subclasses, patient stratification and targeted treatment/clinical management, and alterations in the clinical diagnosis in favour of the methylation-based over the histopathological diagnosis. Finally, we address several considerations regarding the use of DNA methylation profiling as a diagnostic tool, e.g., the threshold of the classifier, the calibrated score, tumour cell content and intratumour heterogeneity.

Machine Learning in the Classification of Pediatric Posterior Fossa Tumors: A Systematic Review

Background: Posterior fossa tumors (PFTs) are a morbid group of central nervous system tumors that most often present in childhood. While early diagnosis is critical to drive appropriate treatment, definitive diagnosis is currently only achievable through invasive tissue collection and histopathological analyses. Machine learning has been investigated as an alternative means of diagnosis. In this systematic review and meta-analysis, we evaluated the primary literature to identify all machine learning algorithms developed to classify and diagnose pediatric PFTs using imaging or molecular data. Methods: Of the 433 primary papers identified in PubMed, EMBASE, and Web of Science, 25 ultimately met the inclusion criteria. The included papers were extracted for algorithm architecture, study parameters, performance, strengths, and limitations. Results: The algorithms exhibited variable performance based on sample size, classifier(s) used, and individual tumor types being investigated. Ependymoma, medulloblastoma, and pilocytic astrocytoma were the most studied tumors with algorithm accuracies ranging from 37.5% to 94.5%. A minority of studies compared the developed algorithm to a trained neuroradiologist, with three imaging-based algorithms yielding superior performance. Common algorithm and study limitations included small sample sizes, uneven representation of individual tumor types, inconsistent performance reporting, and a lack of application in the clinical environment. Conclusions: Artificial intelligence has the potential to improve the speed and accuracy of diagnosis in this field if the right algorithm is applied to the right scenario. Work is needed to standardize outcome reporting and facilitate additional trials to allow for clinical uptake.

Spatial heterogeneity in DNA methylation and chromosomal alterations in diffuse gliomas and meningiomas

Adult-type diffuse gliomas and meningiomas are the most common primary intracranial tumors of the central nervous system. DNA methylation profiling is a novel diagnostic technique increasingly used also in the clinic. Although molecular heterogeneity is well described in these tumors, DNA methylation heterogeneity is less studied. We therefore investigated the intratumor genetic and epigenetic heterogeneity in diffuse gliomas and meningiomas, with focus on potential clinical implications. We further investigated tumor purity as a source for heterogeneity in the tumors. We analyzed genome-wide DNA methylation profiles generated from 126 spatially separated tumor biopsies from 39 diffuse gliomas and meningiomas. Moreover, we evaluated five methods for measurement of tumor purity and investigated intratumor heterogeneity by assessing DNA methylation-based classification, chromosomal copy number alterations and molecular markers. Our results demonstrated homogeneous methylation-based classification of IDH-mutant gliomas and further corroborates subtype heterogeneity in glioblastoma IDH-wildtype and high-grade meningioma patients after excluding samples with low tumor purity. We detected a large number of differentially methylated CpG sites within diffuse gliomas and meningiomas, particularly in tumors of higher grades. The presence of CDKN2A/B homozygous deletion differed in one out of two patients with IDH-mutant astrocytomas, CNS WHO grade 4. We conclude that diffuse gliomas and high-grade meningiomas are characterized by intratumor heterogeneity, which should be considered in clinical diagnostics and in the assessment of methylation-based and molecular markers.

Recent Advances in Pediatric Cancer Research

Over the past few years, the field of pediatric cancer has experienced a shift in momentum, and this has led to new and exciting findings that have relevance beyond pediatric malignancies. Here we present the current status of key aspects of pediatric cancer research. We have focused on genetic and epigenetic drivers of disease, cellular origins of different pediatric cancers, disease models, the tumor microenvironment, and cellular immunotherapies.

DNA 5-hydroxymethylcytosine in pediatric central nervous system tumors may impact tumor classification and is a positive prognostic marker

Background

Nucleotide-specific 5-hydroxymethylcytosine (5hmC) remains understudied in pediatric central nervous system (CNS) tumors. 5hmC is abundant in the brain, and alterations to 5hmC in adult CNS tumors have been reported. However, traditional approaches to measure DNA methylation do not distinguish between 5-methylcytosine (5mC) and its oxidized counterpart 5hmC, including those used to build CNS tumor DNA methylation classification systems. We measured 5hmC and 5mC epigenome-wide at nucleotide resolution in glioma, ependymoma, and embryonal tumors from children, as well as control pediatric brain tissues using tandem bisulfite and oxidative bisulfite treatments followed by hybridization to the Illumina Methylation EPIC Array that interrogates over 860,000 CpG loci.

Results

Linear mixed effects models adjusted for age and sex tested the CpG-specific differences in 5hmC between tumor and non-tumor samples, as well as between tumor subtypes. Results from model-based clustering of tumors was used to test the relation of cluster membership with patient survival through multivariable Cox proportional hazards regression. We also assessed the robustness of multiple epigenetic CNS tumor classification methods to 5mC-specific data in both pediatric and adult CNS tumors. Compared to non-tumor samples, tumors were hypohydroxymethylated across the epigenome and tumor 5hmC localized to regulatory elements crucial to cell identity, including transcription factor binding sites and super-enhancers. Differentially hydroxymethylated loci among tumor subtypes tended to be hypermethylated and disproportionally found in CTCF binding sites and genes related to posttranscriptional RNA regulation, such as DICER1. Model-based clustering results indicated that patients with low 5hmC patterns have poorer overall survival and increased risk of recurrence. Our results suggest 5mC-specific data from OxBS-treated samples impacts methylation-based tumor classification systems giving new opportunities for further refinement of classifiers for both pediatric and adult tumors.

Conclusions

We identified that 5hmC localizes to super-enhancers, and genes commonly implicated in pediatric CNS tumors were differentially hypohydroxymethylated. We demonstrated that distinguishing methylation and hydroxymethylation is critical in identifying tumor-related epigenetic changes. These results have implications for patient prognostication, considerations of epigenetic therapy in CNS tumors, and for emerging molecular neuropathology classification approaches.

Supplementary Information

The online version contains supplementary material available at 10.1186/s13148-021-01156-9.

MethylSPWNet and MethylCapsNet: Biologically Motivated Organization of DNAm Neural Networks, Inspired by Capsule Networks

DNA methylation (DNAm) alterations have been heavily implicated in carcinogenesis and the pathophysiology of diseases through upstream regulation of gene expression. DNAm deep-learning approaches are able to capture features associated with aging, cell type, and disease progression, but lack incorporation of prior biological knowledge. Here, we present modular, user-friendly deep-learning methodology and software, MethylCapsNet and MethylSPWNet, that group CpGs into biologically relevant capsules-such as gene promoter context, CpG island relationship, or user-defined groupings-and relate them to diagnostic and prognostic outcomes. We demonstrate these models' utility on 3,897 individuals in the classification of central nervous system (CNS) tumors. MethylCapsNet and MethylSPWNet provide an opportunity to increase DNAm deep-learning analyses' interpretability by enabling a flexible organization of DNAm data into biologically relevant capsules.

DNA methylation profiling for molecular classification of adult diffuse lower-grade gliomas

BACKGROUND

DNA methylation profiling has facilitated and improved the classification of a wide variety of tumors of the central nervous system. In this study, we investigated the potential utility of DNA methylation profiling to achieve molecular diagnosis in adult primary diffuse lower-grade glioma (dLGG) according to WHO 2016 classification system. We also evaluated whether methylation profiling could provide improved molecular characterization and identify prognostic differences beyond the classical histological WHO grade together with IDH mutation status and 1p/19q codeletion status. All patients diagnosed with dLGG in the period 2007-2016 from the Västra Götaland region in Sweden were assessed for inclusion in the study.

RESULTS

A total of 166 dLGG cases were subjected for genome-wide DNA methylation analysis. Of these, 126 (76%) were assigned a defined diagnostic methylation class with a class prediction score ≥ 0.84 and subclass score ≥ 0.50. The assigned methylation classes were highly associated with their IDH mutation status and 1p/19q codeletion status. IDH-wildtype gliomas were further divided into subgroups with distinct molecular features.

CONCLUSION

The stratification of the patients by methylation profiling was as effective as the integrated WHO 2016 molecular reclassification at predicting the clinical outcome of the patients. Our study shows that DNA methylation profiling is a reliable and robust approach for the classification of dLGG into molecular defined subgroups, providing accurate detection of molecular markers according to WHO 2016 classification.

Epigenomics and immunotherapeutic advances in pediatric brain tumors

Brain tumors are the leading cause of childhood cancer-related deaths. Similar to adult brain tumors, pediatric brain tumors are classified based on histopathological evaluations. However, pediatric brain tumors are often histologically inconsistent with adult brain tumors. Recent research findings from molecular genetic analyses have revealed molecular and genetic changes in pediatric tumors that are necessary for appropriate classification to avoid misdiagnosis, the development of treatment modalities, and the clinical management of tumors. As many of the molecular-based therapies developed from clinical trials on adults are not always effective against pediatric brain tumors, recent advances have improved our understanding of the molecular profiles of pediatric brain tumors and have led to novel epigenetic and immunotherapeutic treatment approaches currently being evaluated in clinical trials. In this review, we focus on primary malignant brain tumors in children and genetic, epigenetic, and molecular characteristics that differentiate them from brain tumors in adults. The comparison of pediatric and adult brain tumors highlights the need for treatments designed specifically for pediatric brain tumors. We also discuss the advancements in novel molecularly targeted drugs and how they are being integrated with standard therapy to improve the classification and outcomes of pediatric brain tumors in the future.

DNA methylation-based age estimation in pediatric healthy tissues and brain tumors

Several DNA methylation clocks have been developed to reflect chronological age of human tissues, but most clocks have been trained on adult samples. The rapid methylome changes in children and the role of epigenetics in pediatric tumors calls for tools accurately estimating methylation age in children. We aimed to evaluate seven methylation clocks in multiple tissues from healthy children to inform future studies on the optimal clock for pediatric cohorts, and analyzed the methylation age in brain tumors. We found that clocks trained on pediatric samples were the best in all tested tissues, highlighting the need for dedicated clocks. For blood samples, the Skin and blood clock had the best correlation with chronological age, while PedBE was the most accurate for saliva and buccal samples, and Horvath for brain tissue. Horvath methylation age was accelerated in pediatric brain tumors and the acceleration was subtype-specific for atypical teratoid rhabdoid tumor (ATRT), ependymoma, medulloblastoma and glioma. The subtypes with the highest acceleration corresponded to the worst prognostic categories in ATRT, ependymoma and glioma, whereas the relationship was reversed in medulloblastoma. This suggests that methylation age has potential as a prognostic biomarker in pediatric brain tumors and should be further explored.

Intratumor DNA methylation heterogeneity in glioblastoma: implications for DNA methylation-based classification

BACKGROUND

A feature of glioblastoma (GBM) is cellular and molecular heterogeneity, both within and between tumors. This variability causes a risk for sampling bias and potential tumor escape from future targeted therapy. Heterogeneous intratumor gene expression in GBM is well documented, but little is known regarding the epigenetic heterogeneity. Variability in DNA methylation within tumors would have implications for diagnostics, as methylation can be used for tumor classification, subtyping, and determination of the clinically used biomarker O6-methylguanine-DNA methyltransferase (MGMT) promoter methylation. We therefore aimed to profile the intratumor DNA methylation heterogeneity in GBM and its effect on diagnostic properties.

METHODS

Three to 4 spatially separated biopsies per tumor were collected from 12 GBM patients. We performed genome-wide DNA methylation analysis and investigated intratumor variation.

RESULTS

All samples were classified as GBM isocitrate dehydrogenase (IDH) wild type (wt)/mutated by methylation profiling, but the subclass differed within 5 tumors. Some GBM samples exhibited higher DNA methylation differences within tumors than between, and many cytosine-phosphate-guanine (CpG) sites (mean: 17 000) had different methylation levels within the tumors. MGMT methylation status differed in IDH mutated patients (1/1).

CONCLUSIONS

We demonstrated that intratumor DNA methylation heterogeneity is a feature of GBM. Although all biopsies were classified as GBM IDH wt/mutated by methylation analysis, the assigned subclass differed in samples from the same patient. The observed heterogeneity within tumors is important to consider for methylation-based biomarkers and future improvements in stratification of GBM patients.

Batch-normalization of cerebellar and medulloblastoma gene expression datasets utilizing empirically defined negative control genes

Motivation

Medulloblastoma (MB) is a brain cancer predominantly arising in children. Roughly 70% of patients are cured today, but survivors often suffer from severe sequelae. MB has been extensively studied by molecular profiling, but often in small and scattered cohorts. To improve cure rates and reduce treatment side effects, accurate integration of such data to increase analytical power will be important, if not essential.

Results

We have integrated 23 transcription datasets, spanning 1350 MB and 291 normal brain samples. To remove batch effects, we combined the Removal of Unwanted Variation (RUV) method with a novel pipeline for determining empirical negative control genes and a panel of metrics to evaluate normalization performance. The documented approach enabled the removal of a majority of batch effects, producing a large-scale, integrative dataset of MB and cerebellar expression data. The proposed strategy will be broadly applicable for accurate integration of data and incorporation of normal reference samples for studies of various diseases. We hope that the integrated dataset will improve current research in the field of MB by allowing more large-scale gene expression analyses.

Availability and implementation

The RUV-normalized expression data is available through the Gene Expression Omnibus (GEO; https://www.ncbi.nlm.nih.gov/geo/) and can be accessed via the GSE series number GSE124814.

Supplementary information

Supplementary data are available at Bioinformatics online.

Brain tumour diagnostics using a DNA methylation-based classifier as a diagnostic support tool

AIMS

Methylation profiling (MP) is increasingly incorporated in the diagnostic process of central nervous system (CNS) tumours at our centres in The Netherlands and Scandinavia. We aimed to identify the benefits and challenges of MP as a support tool for CNS tumour diagnostics.

METHODS

About 502 CNS tumour samples were analysed using (850 k) MP. Profiles were matched with the DKFZ/Heidelberg CNS Tumour Classifier. For each case, the final pathological diagnosis was compared to the diagnosis before MP.

RESULTS

In 54.4% (273/502) of all analysed cases, the suggested methylation class (calibrated score ≥0.9) corresponded with the initial pathological diagnosis. The diagnosis of 24.5% of these cases (67/273) was more refined after incorporation of the MP result. In 9.8% of cases (49/502), the MP result led to a new diagnosis, resulting in an altered WHO grade in 71.4% of these cases (35/49). In 1% of cases (5/502), the suggested class based on MP was initially disregarded/interpreted as misleading, but in retrospect, the MP result predicted the right diagnosis for three of these cases. In six cases, the suggested class was interpreted as 'discrepant but noncontributory'. The remaining 33.7% of cases (169/502) had a calibrated score <0.9, including 7.8% (39/502) for which no class indication was given at all (calibrated score <0.3).

CONCLUSIONS

MP is a powerful tool to confirm and fine-tune the pathological diagnosis of CNS tumours, and to avoid misdiagnoses. However, it is crucial to interpret the results in the context of clinical, radiological, histopathological and other molecular information.

Methylation-based algorithms for diagnosis: experience from neuro-oncology

Brain tumours are the most common tumour-related cause of death in young people. Survivors are at risk of significant disability, at least in part related to the effects of treatment. Therefore, there is a need for a precise diagnosis that stratifies patients for the most suitable treatment, matched to the underlying biology of their tumour. Although traditional histopathology has been accurate in predicting treatment responses in many cases, molecular profiling has revealed a remarkable, previously unappreciated, level of biological complexity in the classification of these tumours. Among different molecular technologies, DNA methylation profiling has had the most pronounced impact on brain tumour classification. Furthermore, using machine learning-based algorithms, DNA methylation profiling is changing diagnostic practice. This can be regarded as an exemplar for how molecular pathology can influence diagnostic practice and illustrates some of the unanticipated benefits and risks. © 2020 Pathological Society of Great Britain and Ireland. Published by John Wiley & Sons, Ltd.

Invited Review: DNA methylation-based classification of paediatric brain tumours

DNA methylation-based machine learning algorithms represent powerful diagnostic tools that are currently emerging for several fields of tumour classification. For various reasons, paediatric brain tumours have been the main driving forces behind this rapid development and brain tumour classification tools are likely further advanced than in any other field of cancer diagnostics. In this review, we will discuss the main characteristics that were important for this rapid advance, namely the high clinical need for improvement of paediatric brain tumour diagnostics, the robustness of methylated DNA and the consequential possibility to generate high-quality molecular data from archival formalin-fixed paraffin-embedded pathology specimens, the implementation of a single array platform by most laboratories allowing data exchange and data pooling to an unprecedented extent, as well as the high suitability of the data format for machine learning. We will further discuss the four most central output qualities of DNA methylation profiling in a diagnostic setting (tumour classification, tumour sub-classification, copy number analysis and guidance for additional molecular testing) individually for the most frequent types of paediatric brain tumours. Lastly, we will discuss DNA methylation profiling as a tool for the detection of new paediatric brain tumour classes and will give an overview of the rapidly growing family of new tumours identified with the aid of this technique.

ZBTB16: A new biomarker for primitive neuroectodermal tumor element / Ewing sarcoma

Primitive neuroectodermal tumor (PNET) traditionally encompasses two different classes of tumors with similar morphology - PNET of the peripheral nervous system (pPNET) and PNET of the central nervous system (cPNET). The latter also includes germ cell tumor-derived PNET (gPNET). There are currently no specific markers for gPNET. This study seeks to investigate the expression of ZBTB16 in PNET and other small round blue cell tumors as well as its potential diagnostic utility. Immunohistochemical expression of the ZBTB16 was studied in a total of 27 PNETs (12 pPNETs, 8 cPNETs, 3 primary testicular gPNETs, and 4 metastatic gPNETs) and 38 small round blue cell tumors. Positive expression for ZBTB16 was seen diffusely in 9/12 (75%), moderately in 2/12 (17%) and focally in 1/12 (8%) of pPNETs, diffusely in 3/7 (43%) and moderately in 4/7 (57%) of gPNETs, and diffusely in 2/8 (25%), moderately in 2/8 (25%) and focally in 4/8 (50%) of cPNETs. Whereas, all of the 38 non-PNET small round blue cell tumors were nonreactive. The results suggest that ZBTB16 is a highly sensitive and specific biomarker for both pPNET and gPNET/cPNET. ZBTB16 effectively differentiates PNETs from other small round blue cell tumor mimics, including the two most common germ cell tumor-derived somatic malignancies - rhabdomyosarcoma and nephroblastoma. Of note, compared to the expression of ZBTB16 in pPNET/Ewing sarcoma and gPNET, the expression of ZBTB16 in cPNET was more variable, which appears consistent with the heterogeneity of cPNET. The close proximity of ZBTB16 and FLI-1 genes on chromosome 11q may explain the overexpression of ZBTB16 in PNET, especially in pPNET with t(1122) translocation.

Pontine tumor in a neonate: case report and analysis of the current literature

Tumors of the central nervous system represent the largest group of solid tumors found in pediatric patients. Pilocytic astrocytoma is the most common pediatric glioma, mostly located in the posterior fossa. The majority of brainstem tumors, however, are classified as highly aggressive diffuse intrinsic pontine gliomas (DIPGs) and their prognosis is dismal.The authors report on the case of a neonate in whom MRI and neuropathological assessment were used to diagnose DIPG. Before initiation of the planned chemotherapy, the tumor regressed spontaneously, and the newborn exhibited a normal neurological development. Meanwhile, Illumina Human Methylation450 BeadChip analysis reclassified the tumor as pilocytic astrocytoma of the posterior fossa.In conclusion, the authors advocate not initiating immediate intensive therapy in newborns with brain tumors, even with classical appearance of a DIPG; rather, they would like to encourage a biopsy to define the best individual therapeutic approach and avoid ineffective chemotherapy.

Cell line-based xenograft mouse model of paediatric glioma stem cells mirrors the clinical course of the patient

The leading cause of cancer-related mortality among children is brain tumour, and glioblastoma multiforme (GBM) has the worst prognosis. New treatments are urgently needed, but with few cases and clinical trials in children, pre-clinical models such as patient-derived tumour xenografts (PDTX) are important. To generate these, tumour tissue is transplanted into mice, but this yields highly variable results and requires serial passaging in mice, which is time-consuming and expensive. We therefore aimed to establish a cell line-based orthotopic mouse model representative of the patient tumour. Glioma stem cell (GSC) lines derived from paediatric GBM were orthotopically transplanted into immunodeficient mice. Overall survival data were collected and histological analysis of the resulting neoplasias was performed. Genome-wide DNA methylation arrays were used for methylation and copy-number alterations (CNA) profiling. All GSC lines initiated tumours on transplantation and the survival of the mice correlated well with the survival of the patients. Xenograft tumours presented histological hallmarks of GBM, and were also classified as GBM by methylation profiling. Each xenograft tumour clustered together with its respective injected GSC line and patient tumour based on the methylation data. We have established a robust and reproducible cell line-based xenograft paediatric GBM model. The xenograft tumours accurately reflected the patient tumours and mirrored the clinical course of the patient. This model can therefore be used to assess patient response in pre-clinical studies.

Computational and Statistical Analysis of Array-Based DNA Methylation Data

The characterization of aberrant DNA methylation is emerging as a key part of the study of cancer development and phenotype. The technical advancements and decreasing costs of methods for high-throughput profiling of DNA methylation have brought about a high interest in the use of such methods in disease association studies. Here we discuss the principles for DNA methylation analysis using data from the Infinium DNA methylation BeadChip assays and describe the computational steps and statistical considerations going from processing of the raw array data to analysis of differential methylation. Moreover, we provide detailed guidelines on how to perform tumor subtype classification based on DNA methylation signatures.

Privacy-Preserving Similar Patient Queries for Combined Biomedical Data

The decreasing costs of molecular profiling have fueled the biomedical research community with a plethora of new types of biomedical data, enabling a breakthrough towards more precise and personalized medicine. Naturally, the increasing availability of data also enables physicians to compare patients’ data and treatments easily and to find similar patients in order to propose the optimal therapy. Such similar patient queries (SPQs) are of utmost importance to medical practice and will be relied upon in future health information exchange systems. While privacy-preserving solutions have been previously studied, those are limited to genomic data, ignoring the different newly available types of biomedical data.

In this paper, we propose new cryptographic techniques for finding similar patients in a privacy-preserving manner with various types of biomedical data, including genomic, epigenomic and transcriptomic data as well as their combination. We design protocols for two of the most common similarity metrics in biomedicine: the Euclidean distance and Pearson correlation coefficient. Moreover, unlike previous approaches, we account for the fact that certain locations contribute differently to a given disease or phenotype by allowing to limit the query to the relevant locations and to assign them different weights. Our protocols are specifically designed to be highly efficient in terms of communication and bandwidth, requiring only one or two rounds of communication and thus enabling scalable parallel queries. We rigorously prove our protocols to be secure based on cryptographic games and instantiate our technique with three of the most important types of biomedical data – namely DNA, microRNA expression, and DNA methylation. Our experimental results show that our protocols can compute a similarity query over a typical number of positions against a database of 1,000 patients in a few seconds. Finally, we propose and formalize strategies to mitigate the threat of malicious users or hospitals.

Stem cell cultures derived from pediatric brain tumors accurately model the originating tumors

Brain tumors are the leading cause of cancer-related death in children but high-grade gliomas in children and adolescents have remained a relatively under-investigated disease despite this. A better understanding of the cellular and molecular pathogenesis of the diseases is required in order to improve the outcome for these children. In vitro-cultured primary tumor cells from patients are indispensable tools for this purpose by enabling functional analyses and development of new therapies. However, relevant well-characterized in vitro cultures from pediatric gliomas cultured under serum-free conditions have been lacking. We have therefore established patient-derived in vitro cultures and performed thorough characterization of the cells using large-scale analyses of DNA methylation, copy-number alterations and investigated their stability during prolonged time in culture. We show that the cells were stable during prolonged culture in serum-free stem cell media without apparent alterations in morphology or growth rate. The cells were proliferative, positive for stem cell markers, able to respond to differentiation cues and initiated tumors in zebrafish and mice suggesting that the cells are cancer stem cells or progenitor cells. The cells accurately mirrored the tumor they were derived from in terms of methylation pattern, copy number alterations and DNA mutations. These unique primary in vitro cultures can thus be used as a relevant and robust model system for functional studies on pediatric brain tumors.

Insights From Molecular Profiling of Adult Glioma

The comprehensive molecular profiling of cancer has resulted in new insights into the biology and classification of numerous tumor types. In the case of primary brain tumors that commonly affect adults, an emerging set of disease-defining biomarker sets is reshaping existing diagnostic entities that had previously been defined on the basis of their microscopic appearance. Substantial progress has been made in this regard for common primary brain tumors in adults, especially diffuse gliomas, where large-scale profiling efforts have led to the incorporation of highly prevalent molecular alterations that promote a biologically based classification as an adjunct to the traditional histopathologic approach. The growing awareness that histologically indistinguishable tumors can be divided into more precise and biologically relevant subgroups has demanded a more global routine approach to biomarker assessment. These considerations have begun to intersect with the decreasing costs and availability of genome-wide analysis tools and, thus, incorporation into routine practice. We review how molecular profiling already has led to an evolution in the classification of brain tumors. In addition, we discuss the likely trajectory of incorporation of global molecular profiling platforms into the routine clinical classification of adult brain tumors.

Validation of the MethylationEPIC BeadChip for fresh-frozen and formalin-fixed paraffin-embedded tumours

DNA methylation is the most studied epigenetic modification due to its role in regulating gene expression, and its involvement in the pathogenesis of cancer and several diseases upon aberrations in methylation. The method of choice to evaluate genome-wide methylation has been the Illumina HumanMethylation450 BeadChip (450K), but it was recently replaced with the MethylationEPIC BeadChip (EPIC). We therefore sought to validate the EPIC array in comparison to the 450K array for both fresh-frozen (FF) and formalin-fixed paraffin-embedded (FFPE) tumours. We also performed analysis on the EPIC array with paired FF and FFPE samples to adapt to a clinical setting where FFPE is routinely used. Further, we compared two restoration methods, REPLI-g and Infinium, for FFPE-derived DNA on the EPIC array.

The Pearson correlation of β values for common probes on the 450K and EPIC array was high for both FF (mean: 0.992) and FFPE (mean: 0.984) samples. The β values generated from the EPIC array for FFPE samples correlated well with the paired FF tumours, but varied between 0.901 and 0.987. We did note that sample pairs with lower correlation had less bimodal density distributions of β values and displayed higher noise in the copy number alteration plots (generated from the methylation array data) in the FFPE sample. Both REPLI-g and the Infinium restoration for FFPE samples performed well on the EPIC array and generated equivalent correlation scores to the paired FF sample.

Preoperative systemic levels of VEGFA, IL-7, IL-17A, and TNF-β delineate two distinct groups of children with brain tumors

BACKGROUND

Primary brain tumors are the most common solid tumors in children. Increasing evidence demonstrates diverse intratumoral immune signatures, which are tentatively reflected in peripheral blood.

PROCEDURE

Twenty cytokines were analyzed in preoperative plasma samples from five healthy children and 45 children with brain tumors, using a multiplex platform (MesoScale Discovery V-PLEX^® ). Tumor types included medulloblastoma (MB), ependymoma, sarcoma, high-grade glioma, pilocytic astrocytoma, and other low-grade gliomas.

RESULTS

A panel of four cytokines [VEGFA, interleukin (IL)-7, IL-17A, and tumor necrosis factor (TNF)-β] delineated two distinct patient groups, identified as VEGFA^high IL-7^high IL-17A^low TNF-β^low (Group A) and VEGFA^low IL-7^low IL-17A^high TNF-β^high (Group B). Healthy controls and the vast majority of patients with MB were found within Group A, whereas patients with other tumor types were equally distributed between the two groups. Unrelated to A/B affiliation, we detected trends toward increased IL-10 and decreased IL-12/23 and TNF-α in several tumor types. Finally, a small number of patients displayed evidence of enhanced systemic immune activation, including elevated levels of interferon-γ, granulocyte monocyte colony-stimulating factor, IL-6, IL-12/23, and TNF-α. Following tumor resection, cytokine levels in a MB patient approached the levels of healthy controls.

CONCLUSIONS

We identify common features and individual differences in the systemic immune profiles of children with brain tumors. Overall, patients with MB displayed a uniform cytokine profile, whereas other tumor diagnoses did not predict systemic immunological status in single patients. Future characterization and monitoring of systemic immune responses in children with brain tumors will have important implications for the development and implementation of immunotherapy.

MethPed: an R package for the identification of pediatric brain tumor subtypes

Background

DNA methylation profiling of pediatric brain tumors offers a new way of diagnosing and subgrouping these tumors which improves current clinical diagnostics based on histopathology. We have therefore developed the MethPed classifier, which is a multiclass random forest algorithm, based on DNA methylation profiles from many subgroups of pediatric brain tumors.

Results

We developed an R package that implements the MethPed classifier, making it easily available and accessible. The package can be used for estimating the probability that an unknown sample belongs to each of nine pediatric brain tumor diagnoses/subgroups.

Conclusions

The MethPed R package efficiently classifies pediatric brain tumors using the developed MethPed classifier. MethPed is available via Bioconductor: http://bioconductor.org/packages/MethPed/

New Brain Tumor Entities Emerge from Molecular Classification of CNS-PNETs

Primitive neuroectodermal tumors of the central nervous system (CNS-PNETs) are highly aggressive, poorly differentiated embryonal tumors occurring predominantly in young children but also affecting adolescents and adults. Herein, we demonstrate that a significant proportion of institutionally diagnosed CNS-PNETs display molecular profiles indistinguishable from those of various other well-defined CNS tumor entities, facilitating diagnosis and appropriate therapy for patients with these tumors. From the remaining fraction of CNS-PNETs, we identify four new CNS tumor entities, each associated with a recurrent genetic alteration and distinct histopathological and clinical features. These new molecular entities, designated "CNS neuroblastoma with FOXR2 activation (CNS NB-FOXR2)," "CNS Ewing sarcoma family tumor with CIC alteration (CNS EFT-CIC)," "CNS high-grade neuroepithelial tumor with MN1 alteration (CNS HGNET-MN1)," and "CNS high-grade neuroepithelial tumor with BCOR alteration (CNS HGNET-BCOR)," will enable meaningful clinical trials and the development of therapeutic strategies for patients affected by poorly differentiated CNS tumors.