Explore the vast range of titles available.

Recently, we described a machine learning approach for classification of central nervous system tumors based on the analysis of genome-wide DNA methylation patterns [ 6 ]. Here, we report on DNA methylation-based central nervous system (CNS) tumor diagnostics conducted in our institution between the years 2015 and 2018. In this period, more than 1000 tumors from the neurosurgical departments in Heidelberg and Mannheim and more than 1000 tumors referred from external institutions were subjected to DNA methylation analysis for diagnostic purposes. We describe our current approach to the integrated diagnosis of CNS tumors with a focus on constellations with conflicts between morphological and molecular genetic findings. We further describe the benefit of integrating DNA copy-number alterations into diagnostic considerations and provide a catalog of copy-number changes for individual DNA methylation classes. We also point to several pitfalls accompanying the diagnostic implementation of DNA methylation profiling and give practical suggestions for recurring diagnostic scenarios.

The WHO classification of brain tumours describes 15 subtypes of meningioma. Nine of these subtypes are allotted to WHO grade I, and three each to grade II and grade III. Grading is based solely on histology, with an absence of molecular markers. Although the existing classification and grading approach is of prognostic value, it harbours shortcomings such as ill-defined parameters for subtypes and grading criteria prone to arbitrary judgment. In this study, we aimed for a comprehensive characterisation of the entire molecular genetic landscape of meningioma to identify biologically and clinically relevant subgroups. In this multicentre, retrospective analysis, we investigated genome-wide DNA methylation patterns of meningiomas from ten European academic neuro-oncology centres to identify distinct methylation classes of meningiomas. The methylation classes were further characterised by DNA copy number analysis, mutational profiling, and RNA sequencing. Methylation classes were analysed for progression-free survival outcomes by the Kaplan-Meier method. The DNA methylation-based and WHO classification schema were compared using the Brier prediction score, analysed in an independent cohort with WHO grading, progression-free survival, and disease-specific survival data available, collected at the Medical University Vienna (Vienna, Austria), assessing methylation patterns with an alternative methylation chip. We retrospectively collected 497 meningiomas along with 309 samples of other extra-axial skull tumours that might histologically mimic meningioma variants. Unsupervised clustering of DNA methylation data clearly segregated all meningiomas from other skull tumours. We generated genome-wide DNA methylation profiles from all 497 meningioma samples. DNA methylation profiling distinguished six distinct clinically relevant methylation classes associated with typical mutational, cytogenetic, and gene expression patterns. Compared with WHO grading, classification by individual and combined methylation classes more accurately identifies patients at high risk of disease progression in tumours with WHO grade I histology, and patients at lower risk of recurrence among WHO grade II tumours (p=0·0096) from the Brier prediction test). We validated this finding in our independent cohort of 140 patients with meningioma. DNA methylation-based meningioma classification captures clinically more homogenous groups and has a higher power for predicting tumour recurrence and prognosis than the WHO classification. The approach presented here is potentially very useful for stratifying meningioma patients to observation-only or adjuvant treatment groups. We consider methylation-based tumour classification highly relevant for the future diagnosis and treatment of meningioma. German Cancer Aid, Else Kröner-Fresenius Foundation, and DKFZ/Heidelberg Institute of Personalized Oncology/Precision Oncology Program.

Bulk-tissue DNA methylomes represent an average over many different cell types, hampering our understanding of cell-type-specific contributions to disease development. As single-cell methylomics is not scalable to large cohorts of individuals, cost-effective computational solutions are needed, yet current methods are limited to tissues such as blood. Here we leverage the high-resolution nature of tissue-specific single-cell RNA-sequencing datasets to construct a DNA methylation atlas defined for 13 solid tissue types and 40 cell types. We comprehensively validate this atlas in independent bulk and single-nucleus DNA methylation datasets. We demonstrate that it correctly predicts the cell of origin of diverse cancer types and discovers new prognostic associations in olfactory neuroblastoma and stage 2 melanoma. In brain, the atlas predicts a neuronal origin for schizophrenia, with neuron-specific differential DNA methylation enriched for corresponding genome-wide association study risk loci. In summary, the DNA methylation atlas enables the decomposition of 13 different human tissue types at a high cellular resolution, paving the way for an improved interpretation of epigenetic data. This resource presents an in silico generated DNA methylation atlas that can be used for cell-type deconvolution of human tissues.

DNA methylation data-based precision cancer diagnostics is emerging as the state of the art for molecular tumor classification. Standards for choosing statistical methods with regard to well-calibrated probability estimates for these typically highly multiclass classification tasks are still lacking. To support this choice, we evaluated well-established machine learning (ML) classifiers including random forests (RFs), elastic net (ELNET), support vector machines (SVMs) and boosted trees in combination with post-processing algorithms and developed ML workflows that allow for unbiased class probability (CP) estimation. Calibrators included ridge-penalized multinomial logistic regression (MR) and Platt scaling by fitting logistic regression (LR) and Firth’s penalized LR. We compared these workflows on a recently published brain tumor 450k DNA methylation cohort of 2,801 samples with 91 diagnostic categories using a 5 × 5-fold nested cross-validation scheme and demonstrated their generalizability on external data from The Cancer Genome Atlas. ELNET was the top stand-alone classifier with the best calibration profiles. The best overall two-stage workflow was MR-calibrated SVM with linear kernels closely followed by ridge-calibrated tuned RF. For calibration, MR was the most effective regardless of the primary classifier. The protocols developed as a result of these comparisons provide valuable guidance on choosing ML workflows and their tuning to generate well-calibrated CP estimates for precision diagnostics using DNA methylation data. Computation times vary depending on the ML algorithm from <15 min to 5 d using multi-core desktop PCs. Detailed scripts in the open-source R language are freely available on GitHub, targeting users with intermediate experience in bioinformatics and statistics and using R with Bioconductor extensions. This work compares several ML and calibration algorithms for classifying tumor DNA methylation profiles. The resulting protocol provides workflows for selecting, training and calibrating ML algorithms to generate well-calibrated multiclass probability estimates.

Background Craniopharyngiomas are rare, mostly benign brain tumors, and their management remains challenging due to the limited data from large cohorts. This study evaluates the practice patterns and outcomes for craniopharyngioma patients managed at a tertiary care center. Methods This retrospective cohort study included patients with histologically confirmed craniopharyngioma treated between 1996 and 2022. Patient, tumor, and treatment variables were analyzed for their association with local control (LC), progression-free survival (PFS), and overall survival (OS) using multivariable Cox regression models. Results A total of 88 patients were analyzed. The median clinical and radiographic follow-up periods were 62.0 and 42.5 months, respectively. Fifty-three recurrences and twelve deaths were observed. After primary treatment, the 2-, 4, and 6-year LC and PFS rates were 69.1, 50.7, 37.7% and 71.5, 55.4, and 47.3%, respectively. For patients undergoing primary treatment, multivariable Cox regression showed an association between the extent of resection, i.e., gross total resection (GTR), and PFS (hazard ratio (HR): 0.36, p  = 0.01) with weaker evidence for LC (HR: 0.40, p  = 0.053). Age was the only variable associated with OS (HR: 1.05, p  = 0.01). Seventeen patients received radiotherapy, which was not formally associated with LC, PFS, and OS. The majority of patients required hormone replacement therapy after treatment. Conclusions This study underlines the role of GTR in delaying disease progression and the need for hormone replacement after treatment. While radiotherapy was not formally associated with any benefit in this series, its use might be helpful in candidates after subtotal resection and for treating recurrences. Further prospective research is needed to refine treatment algorithms, improve long-term outcomes, and optimize the quality of life of affected patients.

We have recently developed a machine learning classifier that enables fast, accurate, and affordable classification of brain tumors based on genome-wide DNA methylation profiles that is widely employed in the clinic. Neuro-oncology research would benefit greatly from understanding the underlying artificial intelligence decision process, which currently remains unclear. Here, we describe an interpretable framework to explain the classifier’s decisions. We show that functional genomic regions of various sizes are predominantly employed to distinguish between different tumor classes, ranging from enhancers and CpG islands to large-scale heterochromatic domains. We detect a high degree of genomic redundancy, with many genes distinguishing individual tumor classes, explaining the robustness of the classifier and revealing potential targets for further therapeutic investigation. We anticipate that our resource will build up trust in machine learning in clinical settings, foster biomarker discovery and development of compact point-of-care assays, and enable further epigenome research of brain tumors. Our interpretable framework is accessible to the research community via an interactive web application ( https://hovestadtlab.shinyapps.io/shinyMNP/ ). AI-guided epigenetic classification is integral for accurate brain tumour diagnosis. Here, the authors introduce an interpretable framework to explain the AI process, revealing biological insights and potential therapeutic targets.

Oncogene-induced senescence (OIS) is crucial for tumour suppression. Senescent cells implement a complex pro-inflammatory response termed the senescence-associated secretory phenotype (SASP). The SASP reinforces senescence, activates immune surveillance and paradoxically also has pro-tumorigenic properties. Here, we present evidence that the SASP can also induce paracrine senescence in normal cells both in culture and in human and mouse models of OIS in vivo . Coupling quantitative proteomics with small-molecule screens, we identified multiple SASP components mediating paracrine senescence, including TGF-β family ligands, VEGF, CCL2 and CCL20. Amongst them, TGF-β ligands play a major role by regulating p15 INK4b and p21 CIP1 . Expression of the SASP is controlled by inflammasome-mediated IL-1 signalling. The inflammasome and IL-1 signalling are activated in senescent cells and IL-1α expression can reproduce SASP activation, resulting in senescence. Our results demonstrate that the SASP can cause paracrine senescence and impact on tumour suppression and senescence in vivo . A property of oncogene-induced senescence (OIS) is the induction of a secretory phenotype, termed the senescence-associated secretome (SASP). Gil and colleagues now provide evidence that senescence can be transmitted in a paracrine manner, by showing that induction of the SASP in cells undergoing OIS by inflammasome-mediated interleukin-1 signalling can promote senescence of normal neighbouring cells.

The WHO 2007 classification of tumors of the CNS distinguishes between diffuse astrocytoma WHO grade II (A II WHO2007 ) and anaplastic astrocytoma WHO grade III (AA III WHO2007 ). Patients with A II WHO2007 are significantly younger and survive significantly longer than those with AA III WHO2007 . So far, classification and grading relies on morphological grounds only and does not yet take into account IDH status, a molecular marker of prognostic relevance. We here demonstrate that WHO 2007 grading performs poorly in predicting prognosis when applied to astrocytoma carrying IDH mutations. Three independent series including a total of 1360 adult diffuse astrocytic gliomas with IDH mutation containing 683 A II IDHmut , 562 AA III IDHmut and 115 GBM IDHmut have been examined for age distribution and survival. In all three series patients with A II IDHmut and AA III IDHmut were of identical age at presentation of disease (36–37 years) and the difference in survival between grades was much less (10.9 years for A II IDHmut , 9.3 years for AA III IDHmut ) than that reported for A II WHO2007 versus AA III WHO2007 . Our analyses imply that the differences in age and survival between A II WHO2007 and AA III WHO2007 predominantly depend on the fraction of IDH -non-mutant astrocytomas in the cohort. This data poses a substantial challenge for the current practice of astrocytoma grading and risk stratification and is likely to have far-reaching consequences on the management of patients with IDH -mutant astrocytoma.

Missense mutations of the V600E type constitute the vast majority of tumor-associated somatic alterations in the v-RAF murine sarcoma viral oncogene homolog B1 ( BRAF ) gene. Initially described in melanoma, colon and papillary thyroid carcinoma, these alterations have also been observed in primary nervous system tumors albeit at a low frequency. We analyzed exon 15 of BRAF spanning the V600 locus by direct sequencing in 1,320 adult and pediatric tumors of the nervous system including various types of glial, embryonal, neuronal and glioneuronal, meningeal, adenohypophyseal/sellar, and peripheral nervous system tumors. A total of 96 BRAF mutations were detected; 93 of the V600E type and 3 cases with a three base pair insertion between codons 599 and 600. The highest frequencies of BRAF V600E mutations were found in WHO grade II pleomorphic xanthoastrocytomas (42/64; 66%) and pleomorphic xanthoastrocytomas with anaplasia (15/23; 65%), as well as WHO grade I gangliogliomas (14/77; 18%), WHO grade III anaplastic gangliogliomas (3/6) and pilocytic astrocytomas (9/97; 9%). In pilocytic astrocytomas BRAF V600E mutation was strongly associated with extra-cerebellar location ( p  = 0.009) and was most frequent in diencephalic tumors (4/12; 33%). Glioblastomas and other gliomas were characterized by a low frequency or absence of mutations. No mutations were detected in non-glial tumors, including embryonal tumors, meningiomas, nerve sheath tumors and pituitary adenomas. The high mutation frequencies in pleomorphic xanthoastrocytomas, gangliogliomas and extra-cerebellar pilocytic astrocytomas implicate BRAF V600E mutation as a valuable diagnostic marker for these rare tumor entities. Future clinical trials should address whether BRAF V600E mutant brain tumor patients will benefit from BRAF V600E -directed targeted therapies.

With the number of prognostic and predictive genetic markers in neuro-oncology steadily growing, the need for comprehensive molecular analysis of neuropathology samples has vastly increased. We therefore developed a customized enrichment/hybrid-capture-based next-generation sequencing (NGS) gene panel comprising the entire coding and selected intronic and promoter regions of 130 genes recurrently altered in brain tumors, allowing for the detection of single nucleotide variations, fusions, and copy number aberrations. Optimization of probe design, library generation and sequencing conditions on 150 samples resulted in a 5-workday routine workflow from the formalin-fixed paraffin-embedded sample to neuropathological report. This protocol was applied to 79 retrospective cases with established molecular aberrations for validation and 71 prospective cases for discovery of potential therapeutic targets. Concordance of NGS compared to established, single biomarker methods was 98.0 %, with discrepancies resulting from one case where a TERT promoter mutation was not called by NGS and three ATRX mutations not being detected by Sanger sequencing. Importantly, in samples with low tumor cell content, NGS was able to identify mutant alleles that were not detectable by traditional methods. Information derived from NGS data identified potential targets for experimental therapy in 37/47 (79 %) glioblastomas, 9/10 (90 %) pilocytic astrocytomas, and 5/14 (36 %) medulloblastomas in the prospective target discovery cohort. In conclusion, we present the settings for high-throughput, adaptive next-generation sequencing in routine neuropathology diagnostics. Such an approach will likely become highly valuable in the near future for treatment decision making, as more therapeutic targets emerge and genetic information enters the classification of brain tumors.