EpiDiP/NanoDiP: a versatile unsupervised machine learning edge computing platform for epigenomic tumour diagnostics

DNA methylation analysis based on supervised machine learning algorithms with static reference data, allowing diagnostic tumour typing with unprecedented precision, has quickly become a new standard of care. Whereas genome-wide diagnostic methylation profiling is mostly performed on microarrays, an increasing number of institutions additionally employ nanopore sequencing as a faster alternative. In addition, methylation-specific parallel sequencing can generate methylation and genomic copy number data. Given these diverse approaches to methylation profiling, to date, there is no single tool that allows (1) classification and interpretation of microarray, nanopore and parallel sequencing data, (2) direct control of nanopore sequencers, and (3) the integration of microarray-based methylation reference data. Furthermore, no software capable of entirely running in routine diagnostic laboratory environments lacking high-performance computing and network infrastructure exists. To overcome these shortcomings, we present EpiDiP/NanoDiP as an open-source DNA methylation and copy number profiling suite, which has been benchmarked against an established supervised machine learning approach using in-house routine diagnostics data obtained between 2019 and 2021. Running locally on portable, cost- and energy-saving system-on-chip as well as gpGPU-augmented edge computing devices, NanoDiP works in offline mode, ensuring data privacy. It does not require the rigid training data annotation of supervised approaches. Furthermore, NanoDiP is the core of our public, free-of-charge EpiDiP web service which enables comparative methylation data analysis against an extensive reference data collection. We envision this versatile platform as a useful resource not only for neuropathologists and surgical pathologists but also for the tumour epigenetics research community. In daily diagnostic routine, analysis of native, unfixed biopsies by NanoDiP delivers molecular tumour classification in an intraoperative time frame.

Heterogeneity of DNA methylation profiles and copy number alterations in 10782 adult-type glioblastomas, IDH-wildtype

The morphological patterns leading to the diagnosis of glioblastoma may also commonly be observed in several other distinct tumor entities, which can result in a mixed bag of tumors subsumed under this diagnosis. The 2021 WHO Classification of CNS Tumors has separated several of these entities from the diagnosis of glioblastoma, IDH-wildtype. This study determines the DNA methylation classes most likely receiving the diagnosis glioblastoma, IDH wildtype according to the definition by the WHO 2021 Classification and provides comparative copy number analyses. We identified 10782 methylome datasets uploaded to the web page www.molecularneuropathology.org with a calibrated score of ≥0.9 by the Heidelberg Brain Tumor Classifier version v12.8. These methylation classes were characterized by the diagnosis glioblastoma being the most frequent classification encountered in each of the classes according to the WHO 2021 definition. Further, methylation classes selected for this study predominantly contained adult patients. Unsupervised clustering confirmed the presence of nine methylation classes containing tumors most likely receiving the diagnosis glioblastoma, IDH-wildtype according to the WHO 2021 definition. Copy number analysis and a focus on genes with typical numerical alterations in glioblastoma revealed clear differences between the nine methylation classes. Although great progress in diagnostic precision has been achieved over the last decade, our data clearly demonstrate that glioblastoma, IDH-wildtype still is a heterogeneous group in need of further stratification.

Concurrent gliomas in patients with multiple sclerosis

BACKGROUND

Concurrent malignant brain tumors in patients with multiple sclerosis (MS) constitute a rare but paradigmatic phenomenon for studying neuroimmunological mechanisms from both molecular and clinical perspectives.

METHODS

A multicenter cohort of 26 patients diagnosed with both primary brain tumors and multiple sclerosis was studied for disease localization, tumor treatment-related MS activity, and molecular characteristics specific for diffuse glioma in MS patients.

RESULTS

MS neither predisposes nor protects from the development of gliomas. Patients with glioblastoma WHO grade 4 without isocitratdehydrogenase (IDH) mutations have a longstanding history of MS, whereas patients diagnosed with IDH-mutant astrocytoma WHO grade 2 receive multiple sclerosis diagnosis mostly at the same time or later. Concurrent MS is associated with a lesser extent of tumor resection and a worse prognosis in IDH-mutant glioma patients (PFS 32 vs. 64 months, p = 0.0206). When assessing tumor-intrinsic differences no distinct subgroup-defining methylation pattern is identified in gliomas of MS patients compared to other glioma samples. However, differential methylation of immune-related genetic loci including human leukocyte antigen locus on 6p21 and interleukin locus on 5q31 is found in MS patients vs. matched non-MS patients. In line, inflammatory disease activity increases in 42% of multiple sclerosis patients after brain tumor radiotherapy suggesting a susceptibility of multiple sclerosis brain tissue to pro-inflammatory stimuli such as ionizing radiation.

CONCLUSIONS

Concurrent low-grade gliomas should be considered in multiple sclerosis patients with slowly progressive, expansive T2/FLAIR lesions. Our findings of typically reduced extent of resection in MS patients and increased MS activity after radiation may inform future treatment decisions.

Integrated molecular analysis reveals hypermethylation and overexpression of HOX genes to be poor prognosticators in isocitrate dehydrogenase mutant glioma

BACKGROUND

Diffuse gliomas represent over 80% of malignant brain tumors ranging from low-grade to aggressive high-grade lesions. Within isocitrate dehydrogenase (IDH)-mutant gliomas, there is a high variability in survival and a need to more accurately predict outcome.

METHODS

To identify and characterize a predictive signature of outcome in gliomas, we utilized an integrative molecular analysis (using methylation, mRNA, copy number variation (CNV), and mutation data), analyzing a total of 729 IDH-mutant samples including a test set of 99 from University Health Network (UHN) and 2 validation cohorts including the German Cancer Research Center (DKFZ) and The Cancer Genome Atlas (TCGA).

RESULTS

Cox regression analysis of methylation data from the UHN cohort identified CpG-based signatures that split the glioma cohort into 2 prognostic groups strongly predicting survival that were validated using 2 independent cohorts from TCGA and DKFZ (all P-values < .0001). The methylation signatures that predicted poor outcomes also exhibited high CNV instability and hypermethylation of HOX gene probes. Integrated multi-platform analyses using mRNA and methylation (iRM) showed that parallel HOX gene overexpression and simultaneous hypermethylation were significantly associated with increased mutational load, high aneuploidy, and worse survival (P-value < .0001). A 7-HOX gene signature was developed and validated using the most significantly associated HOX genes with patient outcome in both 1p/19q codeleted and non-codeleted IDHmut gliomas.

CONCLUSIONS

HOX gene methylation and expression provide important prognostic information in IDH-mutant gliomas that are not captured by current molecular diagnostics. A 7-HOX gene signature of outcome shows significant survival differences in both 1p/19q codeleted and non-codeleted IDH-mutant gliomas.

Prognostic Markers of DNA Methylation and Next-Generation Sequencing in Progressive Glioblastoma from the EORTC-26101 Trial

PURPOSE

The EORTC-26101 study was a randomized phase II and III clinical trial of bevacizumab in combination with lomustine versus lomustine alone in progressive glioblastoma. Other than for progression-free survival (PFS), there was no benefit from addition of bevacizumab for overall survival (OS). However, molecular data allow for the rare opportunity to assess prognostic biomarkers from primary surgery for their impact in progressive glioblastoma.

EXPERIMENTAL DESIGN

We analyzed DNA methylation array data and panel sequencing from 170 genes of 380 tumor samples of the EORTC-26101 study. These patients were comparable with the overall study cohort in regard to baseline characteristics, study treatment, and survival.

RESULTS

Of patients' samples, 295/380 (78%) were classified into one of the main glioblastoma groups, receptor tyrosine kinase (RTK)1, RTK2 and mesenchymal. There were 10 patients (2.6%) with isocitrate dehydrogenase mutant tumors in the biomarker cohort. Patients with RTK1 and RTK2 classified tumors had lower median OS compared with mesenchymal (7.6 vs. 9.2 vs. 10.5 months). O6-methylguanine DNA-methyltransferase (MGMT) promoter methylation was prognostic for PFS and OS. Neurofibromin (NF)1 mutations were predictive of response to bevacizumab treatment.

CONCLUSIONS

Thorough molecular classification is important for brain tumor clinical trial inclusion and evaluation. MGMT promoter methylation and RTK1 classifier assignment were prognostic in progressive glioblastoma. NF1 mutation may be a predictive biomarker for bevacizumab treatment.

Identification of Epigenetically Regulated Genes Distinguishing Intracranial from Extracranial Melanoma Metastases

Despite remarkable advances in treating patients with metastatic melanoma, the management of melanoma brain metastases remains challenging. Recent evidence suggests that epigenetic reprogramming is an important mechanism for the adaptation of melanoma cells to the brain environment. In this study, the methylomes and transcriptomes of a cohort of matched melanoma metastases were evaluated by integrated omics data analysis. The identified 38 candidate genes displayed distinct promoter methylation and corresponding gene expression changes in intracranial compared with extracranial metastases. The 11 most promising genes were validated on protein level in both tumor and surrounding normal tissue using immunohistochemistry. In accordance with the underlying promoter methylation and gene expression changes, a significantly different protein expression was confirmed for STK10, PDXK, WDR24, CSSP1, NMB, RASL11B, phosphorylated PRKCZ, PRKCZ, and phosphorylated GRB10 in the intracranial metastases. The observed changes imply a distinct intracranial phenotype with increased protein kinase B phosphorylation and a higher frequency of proliferating cells. Knockdown of PRKCZ or GRB10 altered the expression of phosphorylated protein kinase B and decreased the viability of a brain-specific melanoma cell line. In summary, epigenetically regulated cancer-relevant alterations were identified that provide insights into the molecular mechanisms that discriminate brain metastases from other organ metastases, which could be exploited by targeting the affected signaling pathways.

Transcriptome analysis stratifies second-generation non-WNT/non-SHH medulloblastoma subgroups into clinically tractable subtypes

Medulloblastoma (MB), one of the most common malignant pediatric brain tumor, is a heterogenous disease comprised of four distinct molecular groups (WNT, SHH, Group 3, Group 4). Each of these groups can be further subdivided into second-generation MB (SGS MB) molecular subgroups, each with distinct genetic and clinical characteristics. For instance, non-WNT/non-SHH MB (Group 3/4) can be subdivided molecularly into eight distinct and clinically relevant tumor subgroups. A further molecular stratification/summarization of these SGS MB would allow for the assignment of patients to risk-associated treatment protocols. Here, we performed DNA- and RNA-based analysis of 574 non-WNT/non-SHH MB and analyzed the clinical significance of various molecular patterns within the entire cohort and the eight SGS MB, with the aim to develop an optimal risk stratification of these tumors. Multigene analysis disclosed several survival-associated genes highly specific for each molecular subgroup within this non-WNT/non-SHH MB cohort with minimal inter-subgroup overlap. These subgroup-specific and prognostically relevant genes were associated with pathways that could underlie SGS MB clinical-molecular diversity and tumor-driving mechanisms. By combining survival-associated genes within each SGS MB, distinct metagene sets being appropriate for their optimal risk stratification were identified. Defined subgroup-specific metagene sets were independent variables in the multivariate models generated for each SGS MB and their prognostic value was confirmed in a completely non-overlapping validation cohort of non-WNT/non-SHH MB (n = 377). In summary, the current results indicate that the integration of transcriptome data in risk stratification models may improve outcome prediction for each non-WNT/non-SHH SGS MB. Identified subgroup-specific gene expression signatures could be relevant for clinical implementation and survival-associated metagene sets could be adopted for further SGS MB risk stratification. Future studies should aim at validating the prognostic role of these transcriptome-based SGS MB subtypes in prospective clinical trials.

Multiomic neuropathology improves diagnostic accuracy in pediatric neuro-oncology

The large diversity of central nervous system (CNS) tumor types in children and adolescents results in disparate patient outcomes and renders accurate diagnosis challenging. In this study, we prospectively integrated DNA methylation profiling and targeted gene panel sequencing with blinded neuropathological reference diagnostics for a population-based cohort of more than 1,200 newly diagnosed pediatric patients with CNS tumors, to assess their utility in routine neuropathology. We show that the multi-omic integration increased diagnostic accuracy in a substantial proportion of patients through annotation to a refining DNA methylation class (50%), detection of diagnostic or therapeutically relevant genetic alterations (47%) or identification of cancer predisposition syndromes (10%). Discrepant results by neuropathological WHO-based and DNA methylation-based classification (30%) were enriched in histological high-grade gliomas, implicating relevance for current clinical patient management in 5% of all patients. Follow-up (median 2.5 years) suggests improved survival for patients with histological high-grade gliomas displaying lower-grade molecular profiles. These results provide preliminary evidence of the utility of integrating multi-omics in neuropathology for pediatric neuro-oncology.

Genomic characterization of DICER1-associated neoplasms uncovers molecular classes

DICER1 syndrome is a tumor predisposition syndrome that is associated with up to 30 different neoplastic lesions, usually affecting children and adolescents. Here we identify a group of mesenchymal tumors which is highly associated with DICER1 syndrome, and molecularly distinct from other DICER1-associated tumors. This group of DICER1-associated mesenchymal tumors encompasses multiple well-established clinicopathological tumor entities and can be further divided into three clinically meaningful classes designated "low-grade mesenchymal tumor with DICER1 alteration" (LGMT DICER1), "sarcoma with DICER1 alteration" (SARC DICER1), and primary intracranial sarcoma with DICER1 alteration (PIS DICER1). Our study not only provides a combined approach to classify DICER1-associated neoplasms for improved clinical management but also suggests a role for global hypomethylation and other recurrent molecular events in sarcomatous differentiation in mesenchymal tumors with DICER1 alteration. Our results will facilitate future investigations into prognostication and therapeutic approaches for affected patients.

Pediatric-type high-grade neuroepithelial tumors with CIC gene fusion share a common DNA methylation signature

Pediatric neoplasms in the central nervous system (CNS) show extensive clinical and molecular heterogeneity and are fundamentally different from those occurring in adults. Molecular genetic testing contributes to accurate diagnosis and enables an optimal clinical management of affected children. Here, we investigated a rare, molecularly distinct type of pediatric high-grade neuroepithelial tumor (n = 18), that was identified through unsupervised visualization of genome-wide DNA methylation array data, together with copy number profiling, targeted next-generation DNA sequencing, and RNA transcriptome sequencing. DNA and/or RNA sequencing revealed recurrent fusions involving the capicua transcriptional repressor (CIC) gene in 10/10 tumor samples analyzed, with the most common fusion being CIC::LEUTX (n = 9). In addition, a CIC::NUTM1 fusion was detected in one of the tumors. Apart from the detected fusion events, no additional oncogenic alteration was identified in these tumors. The histopathological review demonstrated a morphologically heterogeneous group of high-grade neuroepithelial tumors with positive immunostaining for markers of glial differentiation in combination with weak and focal expression of synaptophysin, CD56 and CD99. All tumors were located in the supratentorial compartment, occurred during childhood (median age 8.5 years) and typically showed early relapses. In summary, we expand the spectrum of pediatric-type tumors of the CNS by reporting a previously uncharacterized group of rare high-grade neuroepithelial tumors that share a common DNA methylation signature and recurrent gene fusions involving the transcriptional repressor CIC. Downstream functional consequences of the fusion protein CIC::LEUTX and potential therapeutic implications need to be further investigated.

Glioneuronal tumor with ATRX alteration, kinase fusion and anaplastic features (GTAKA): a molecularly distinct brain tumor type with recurrent NTRK gene fusions

Glioneuronal tumors are a heterogenous group of CNS neoplasms that can be challenging to accurately diagnose. Molecular methods are highly useful in classifying these tumors-distinguishing precise classes from their histological mimics and identifying previously unrecognized types of tumors. Using an unsupervised visualization approach of DNA methylation data, we identified a novel group of tumors (n = 20) that formed a cluster separate from all established CNS tumor types. Molecular analyses revealed ATRX alterations (in 16/16 cases by DNA sequencing and/or immunohistochemistry) as well as potentially targetable gene fusions involving receptor tyrosine-kinases (RTK; mostly NTRK1-3) in all of these tumors (16/16; 100%). In addition, copy number profiling showed homozygous deletions of CDKN2A/B in 55% of cases. Histological and immunohistochemical investigations revealed glioneuronal tumors with isomorphic, round and often condensed nuclei, perinuclear clearing, high mitotic activity and microvascular proliferation. Tumors were mainly located supratentorially (84%) and occurred in patients with a median age of 19 years. Survival data were limited (n = 18) but point towards a more aggressive biology as compared to other glioneuronal tumors (median progression-free survival 12.5 months). Given their molecular characteristics in addition to anaplastic features, we suggest the term glioneuronal tumor with ATRX alteration, kinase fusion and anaplastic features (GTAKA) to describe these tumors. In summary, our findings highlight a novel type of glioneuronal tumor driven by different RTK fusions accompanied by recurrent alterations in ATRX and homozygous deletions of CDKN2A/B. Targeted approaches such as NTRK inhibition might represent a therapeutic option for patients suffering from these tumors.

Comparison of transcriptome profiles between medulloblastoma primary and recurrent tumors uncovers novel variance effects in relapses

Nowadays medulloblastoma (MB) tumors can be treated with risk-stratified approaches with up to 80% success rate. However, disease relapses occur in approximately 30% of patients and successful salvage treatment strategies at relapse remain scarce. Acquired copy number changes or TP53 mutations are known to occur frequently in relapses, while methylation profiles usually remain highly similar to those of the matching primary tumors, indicating that in general molecular subgrouping does not change during the course of the disease. In the current study, we have used RNA sequencing data to analyze the transcriptome profiles of 43 primary-relapse MB pairs in order to identify specific molecular features of relapses within various tumor groups. Gene variance analysis between primary and relapse samples demonstrated the impact of age in SHH-MB: the changes in gene expression relapse profiles were more pronounced in the younger patients (< 10 years old), which were also associated with increased DNA aberrations and somatic mutations at relapse probably driving this effect. For Group 3/4 MB transcriptome data analysis uncovered clear sets of genes either active or decreased at relapse that are significantly associated with survival, thus could be potential predictive markers. In addition, deconvolution analysis of bulk transcriptome data identified progression-associated differences in cell type enrichment. The proportion of undifferentiated progenitors increased in SHH-MB relapses with a concomitant decrease of differentiated neuron-like cells, while in Group 3/4 MB relapses cell cycle activity increases and differentiated neuron-like cells proportion decreases as well. Thus, our findings uncovered significant transcriptome changes in the molecular signatures of relapsed MB and could be potentially useful for further clinical purposes.

Reference on copy number variations in pleomorphic xanthoastrocytoma: Implications for diagnostic approach

Pleomorphic xanthoastrocytoma (PXA) poses a diagnostic challenge. The present study relies on methylation-based predictions and focuses on copy number variations (CNV) in PXA. We identified 551 tumors from patients having received the histologic diagnosis or differential diagnosis pleomorphic xanthoastrocytoma (PXA) uploaded to the web page www.molecularneuropathology.org. Of these 551 tumors, 165 received the prediction "methylation class (anaplastic) pleomorphic xanthoastrocytoma" with a calibrated score >=0.9 by the brain tumor classifier version v12.8 and, therefore, were defined the PXA reference set designated mcPXAref. In addition to these 165 mcPXAref, 767 other tumors received the prediction mcPXA with a calibrated score >=0.9 but without a histological PXA diagnosis. The total number of individual tumors predicted by histology and/or by methylome based classification as PXA, mcPXA or both was 1318, and these were designated the study cohort. The selection of a control cohort was guided by methylation-based predictions recurrently observed for the other 386/551 tumors diagnosed as histologic PXA. 131/386 received predictions for another entity besides PXA with a score >=0.9. Control tumors corresponding to the 11 most common other predictions were selected, adding up to 1100 reference cases. CNV profiles were calculated from all methylation datasets of the study and control cohorts. Special attention was given to the 7/10 signature, gene amplifications and homozygous deletion of CDKN2A/B. Comparison of CNV in the subsets of the study cohort and the control cohort were used to establish relations independent of histological diagnoses. Tumors in mcPXA were highly homogenous in regard to CNV alterations, irrespective of the histological diagnoses. The 7/10 signature commonly present in glioblastoma, IDH-wildtype, was present in 15-20% of mcPXA, whereas amplification of oncogenes (likewise common in glioblastoma) was very rare in mcPXA (<1%). In contrast, the histology-based PXA group exhibited high variance in regard to methylation classes as well as to CNVs. Our data add to the notion, that histologically defined PXA likely only represent a subset of the biological disease.

PATH-46. COMPUTATIONAL HISTOPATHOLOGY INFORMED RAPID TARGETED NANOPORE SEQUENCING ENABLES AFFORDABLE NEXT DAY REPORTING OF COMPREHENSIVE MOLECULAR MARKERS FOR CNS TUMOUR DIAGNOSTICS

BACKGROUND

Integrative brain tumour diagnostics indisputably requires comprehensive reporting of molecular markers. The 2021 WHO classification of central nervous system (CNS) tumours substantially increased the set of markers for routine evaluation, with greater significance to DNA methylation analysis in diagnostics. Limited by investment and batching, smaller labs and clinics might suffer major delays in delivering clinical decisions. To make precision diagnostics accessible, we introduce an integrated computational histopathology and adaptive nanopore sequencing workflow for next day CNS tumour diagnostics.

METHODS

We used CNS-CHiP- a multitask deep transfer learning model to predict key molecular alterations and methylation classification from H&E stained CNS tumour slides. For further characterisation and subtyping, we used the predictions to formulate a custom panel for each patient. Targeted sequencing and analyses were performed using Rapid-CNS2- a custom neurooncology nanopore sequencing pipeline for parallel copy-number, mutational and methylation analysis that is flexible in target selection with no additional library preparation and can be initiated upon receipt of frozen sections. Sequencing was performed on a portable MinION or GridION.

RESULTS

We demonstrate our workflow on diagnostic samples received by the Department of Neuropathology, University Hospital Heidelberg. CNS-CHiP predicted multiple pathognomonic alterations (eg. IDH mutation, 7 gain/10 loss) with reasonable accuracy. This provided basic information regarding the tumor type instantly. Personalised panels enabled small target sizes, resulting in low sequencing time (up to 24h) and competitive costs. The GPU-accelerated bioinformatics pipeline reduced analysis time from &gt; 24h to &lt; 3h.

CONCLUSIONS

Our workflow harnessing histology-based molecular predictions to instruct targeted nanopore sequencing can be set up with low initial investment and has the potential to facilitate reporting of molecular results on the next day of sample collection. CNS-CHiP combined with Rapid-CNS2 thus aims to make CNS tumour diagnostics affordable and accessible to smaller hospitals and labs especially in low- and middle-income countries.

Anaplastic ganglioglioma - a diagnosis comprising several distinct tumour types

Anaplastic ganglioglioma is a rare tumour and diagnosis has been based on histological criteria. The 5^th edition of the World Health Organization Classification of Tumours of the Central Nervous System (CNS WHO) does not list anaplastic ganglioglioma as a distinct diagnosis due to lack of molecular data in previous publications AIM: We retrospectively compiled a cohort of 54 histologically diagnosed anaplastic gangliogliomas to explore whether the molecular profiles of these tumours represent a separate type or resolve into other entities METHODS: Samples were subjected to histological review, DNA methylation profiling and next generation sequencing. Morphologic and molecular data were summarised to an integrated diagnosis RESULTS: The majority of histologically diagnosed anaplastic gangliogliomas resolved into CNS WHO diagnoses of glial tumours, most commonly pleomorphic xanthoastrocytoma (16/54), glioblastoma, IDH wildtype and diffuse paediatric-type high-grade glioma, H3 wildtype and IDH wildtype (11 and 2/54) followed by low-grade glial or glioneuronal tumours including pilocytic astrocytoma, dysembryoplastic neuroepithelial tumour and diffuse leptomeningeal glioneuronal tumour (5/54), IDH mutant astrocytoma (4/54) and others (6/54). A subset of tumours (10/54) was not assignable to a CNS WHO diagnosis and common molecular profiles pointing to a separate entity were not evident CONCLUSION: In summary, we show that tumours histologically diagnosed as anaplastic ganglioglioma comprise a wide spectrum of CNS WHO tumour types with different prognostic and therapeutic implications. We therefore suggest assigning this designation with caution and recommend comprehensive molecular workup.

DNA-methylome-assisted classification of patients with poor prognostic subventricular zone associated IDH-wildtype glioblastoma

Glioblastoma (GBM) derived from the “stem cell” rich subventricular zone (SVZ) may constitute a therapy-refractory subgroup of tumors associated with poor prognosis. Risk stratification for these cases is necessary but is curtailed by error prone imaging-based evaluation. Therefore, we aimed to establish a robust DNA methylome-based classification of SVZ GBM and subsequently decipher underlying molecular characteristics. MRI assessment of SVZ association was performed in a retrospective training set of IDH-wildtype GBM patients (n = 54) uniformly treated with postoperative chemoradiotherapy. DNA isolated from FFPE samples was subject to methylome and copy number variation (CNV) analysis using Illumina Platform and cnAnalysis450k package. Deep next-generation sequencing (NGS) of a panel of 130 GBM-related genes was conducted (Agilent SureSelect/Illumina). Methylome, transcriptome, CNV, MRI, and mutational profiles of SVZ GBM were further evaluated in a confirmatory cohort of 132 patients (TCGA/TCIA). A 15 CpG SVZ methylation signature (SVZM) was discovered based on clustering and random forest analysis. One third of CpG in the SVZM were associated with MAB21L2/LRBA. There was a 14.8% (n = 8) discordance between SVZM vs. MRI classification. Re-analysis of these patients favored SVZM classification with a hazard ratio (HR) for OS of 2.48 [95% CI 1.35–4.58], p = 0.004 vs. 1.83 [1.0–3.35], p = 0.049 for MRI classification. In the validation cohort, consensus MRI based assignment was achieved in 62% of patients with an intraclass correlation (ICC) of 0.51 and non-significant HR for OS (2.03 [0.81–5.09], p = 0.133). In contrast, SVZM identified two prognostically distinct subgroups (HR 3.08 [1.24–7.66], p = 0.016). CNV alterations revealed loss of chromosome 10 in SVZM– and gains on chromosome 19 in SVZM– tumors. SVZM– tumors were also enriched for differentially mutated genes (p < 0.001). In summary, SVZM classification provides a novel means for stratifying GBM patients with poor prognosis and deciphering molecular mechanisms governing aggressive tumor phenotypes.

HIP1R and vimentin immunohistochemistry predict 1p/19q status in IDH-mutant glioma

BACKGROUND

IDH-mutant gliomas are separate based on the codeletion of the chromosomal arms 1p and 19q into oligodendrogliomas IDH-mutant 1p/19q-codeleted and astrocytomas IDH-mutant. While nuclear loss of ATRX expression excludes 1p/19q codeletion, its limited sensitivity prohibits to conclude on 1p/19q status in tumors with retained nuclear ATRX expression.

METHODS

Employing mass spectrometry based proteomic analysis in a discovery series containing 35 fresh frozen and 72 formalin fixed and paraffin embedded tumors with established IDH and 1p/19q status, potential biomarkers were discovered. Subsequent validation immunohistochemistry was conducted on two independent series (together 77 oligodendrogliomas IDH-mutant 1p/19q-codeleted and 92 astrocytomas IDH-mutant).

RESULTS

We detected highly specific protein patterns distinguishing oligodendroglioma and astrocytoma. In these patterns, high HIP1R and low vimentin levels were observed in oligodendroglioma while low HIP1R and high vimentin levels occurred in astrocytoma. Immunohistochemistry for HIP1R and vimentin expression in 35 cases from the FFPE discovery series confirmed these findings. Blinded evaluation of the validation cohorts predicted the 1p/19q status with a positive and negative predictive value as well as an accuracy of 100% in the first cohort and with a positive predictive value of 83%; negative predictive value of 100% and an accuracy of 92% in the second cohort. Nuclear ATRX loss as marker for astrocytoma increased the sensitivity to 96% and the specificity to 100%.

CONCLUSIONS

We demonstrate that immunohistochemistry for HIP1R, vimentin, and ATRX predict 1p/19q status with 100% specificity and 95% sensitivity and therefore, constitutes a simple and inexpensive approach to the classification of IDH-mutant glioma.

Osteosarcoma: Novel prognostic biomarkers using circulating and cell-free tumour DNA

AIM

Osteosarcoma (OS) is the most common primary bone tumour in children and adolescents. Circulating free (cfDNA) and circulating tumour DNA (ctDNA) are promising biomarkers for disease surveillance and prognostication in several cancer types; however, few such studies are reported for OS. The purpose of this study was to discover and validate methylation-based biomarkers to detect plasma ctDNA in patients with OS and explore their utility as prognostic markers.

METHODS

Candidate CpG markers were selected through analysis of methylation array data for OS, non-OS tumours and germline samples. Candidates were validated in two independent OS datasets (n = 162, n = 107) and the four top-performing markers were selected. Methylation-specific digital droplet PCR (ddPCR) assays were designed and experimentally validated in OS tumour samples (n = 20) and control plasma samples. Finally, ddPCR assays were applied to pre-operative plasma and where available post-operative plasma from 72 patients with OS, and findings correlated with outcome.

RESULTS

Custom ddPCR assays detected ctDNA in 69% and 40% of pre-operative plasma samples (n = 72), based on thresholds of one or two positive markers respectively. ctDNA was detected in 5/17 (29%) post-operative plasma samples from patients, which in four cases were associated with or preceded disease relapse. Both pre-operative cfDNA levels and ctDNA detection independently correlated with overall survival (p = 0.0015 and p = 0.0096, respectively).

CONCLUSION

Our findings illustrate the potential of mutation-independent methylation-based ctDNA assays for OS. This study lays the foundation for multi-institutional collaborative studies to explore the utility of plasma-derived biomarkers in the management of OS.

Prognostic impact of genetic alterations and methylation classes in meningioma

Meningiomas are classified based on histological features, but genetic and epigenetic features are emerging as relevant biomarkers for outcome prediction and may supplement histomorphological evaluation. We investigated meningioma‐relevant mutations and their correlation with DNA methylation clusters and patient survival times. Formalin‐fixed and paraffin‐embedded samples of 126 meningioma patients (WHO grade I 52/126; 41.3%; WHO grade II: 48/126; 38.1%; WHO grade III: 26/126; 20.6%) were investigated. We analyzed NF2, TRAF7, KLF4, ARID, SMO, AKT,TERT promotor, PIK3CA, and SUFU mutations using panel sequencing and correlated them to DNA methylation classes (MC) determined using 850k EPIC arrays. The TRAKL mutation genotype was characterized by the presence of any of the following mutations: TRAF7, AKT1, and KLF4. Survival data including progression‐free survival (PFS) and overall survival (OS) was retrieved from chart review. Mutations were evident in 90/126 (71.4%) specimens with mutations in NF2 (39/126; 31.0%), TRAF7 (39/126; 31.0%) and KLF4 (25/126; 19.8%) being the most frequent ones. Two or more mutations were observed in 35/126 (27.8%) specimens. While TRAKL was predominantly found in benign MC, NF2 was associated with malign MC (p < 0.05). TRAF7, KLF4, and TRAKL mutation genotype were associated with improved PFS and OS (p < 0.05). TERT promotor methylation, intermediate, and malign MC were associated with impaired PFS and OS (p < 0.05). Methylation cluster showed better prognostic discrimination for PFS and OS (c‐index 0.77/0.75) than each of the individual mutations (c‐index 0.63/0.68). In multivariate analysis correcting for age, gender, MC, and WHO grade, none of the individual mutations except TERT remained an independent significant prognostic factor for PFS. Molecular profiling including mutational analysis and DNA methylation classification may facilitate more precise prognostic assessment and identification of potential targets for personalized therapy in meningioma patients.

Oligosarcomas, IDH-mutant are distinct and aggressive

Oligodendrogliomas are defined at the molecular level by the presence of an IDH mutation and codeletion of chromosomal arms 1p and 19q. In the past, case reports and small studies described gliomas with sarcomatous features arising from oligodendrogliomas, so called oligosarcomas. Here, we report a series of 24 IDH-mutant oligosarcomas from 23 patients forming a distinct methylation class. The tumors were recurrences from prior oligodendrogliomas or developed de novo. Precursor tumors of 12 oligosarcomas were histologically and molecularly indistinguishable from conventional oligodendrogliomas. Oligosarcoma tumor cells were embedded in a dense network of reticulin fibers, frequently showing p53 accumulation, positivity for SMA and CALD1, loss of OLIG2 and gain of H3K27 trimethylation (H3K27me3) as compared to primary lesions. In 5 oligosarcomas no 1p/19q codeletion was detectable, although it was present in the primary lesions. Copy number neutral LOH was determined as underlying mechanism. Oligosarcomas harbored an increased chromosomal copy number variation load with frequent CDKN2A/B deletions. Proteomic profiling demonstrated oligosarcomas to be highly distinct from conventional CNS WHO grade 3 oligodendrogliomas with consistent evidence for a smooth muscle differentiation. Expression of several tumor suppressors was reduced with NF1 being lost frequently. In contrast, oncogenic YAP1 was aberrantly overexpressed in oligosarcomas. Panel sequencing revealed mutations in NF1 and TP53 along with IDH1/2 and TERT promoter mutations. Survival of patients was significantly poorer for oligosarcomas as first recurrence than for grade 3 oligodendrogliomas as first recurrence. These results establish oligosarcomas as a distinct group of IDH-mutant gliomas differing from conventional oligodendrogliomas on the histologic, epigenetic, proteomic, molecular and clinical level. The diagnosis can be based on the combined presence of (a) sarcomatous histology, (b) IDH-mutation and (c) TERT promoter mutation and/or 1p/19q codeletion, or, in unresolved cases, on its characteristic DNA methylation profile.

Pleomorphic xanthoastrocytoma is a heterogeneous entity with pTERT mutations prognosticating shorter survival

Pleomorphic xanthoastrocytoma (PXA) in its classic manifestation exhibits distinct morphological features and is assigned to CNS WHO grade 2 or grade 3. Distinction from glioblastoma variants and lower grade glial and glioneuronal tumors is a common diagnostic challenge. We compared a morphologically defined set of PXA (histPXA) with an independent set, defined by DNA methylation analysis (mcPXA). HistPXA encompassed 144 tumors all subjected to DNA methylation array analysis. Sixty-two histPXA matched to the methylation class mcPXA. These were combined with the cases that showed the mcPXA signature but had received a histopathological diagnosis other than PXA. This cohort constituted a set of 220 mcPXA. Molecular and clinical parameters were analyzed in these groups. Morphological parameters were analyzed in a subset of tumors with FFPE tissue available. HistPXA revealed considerable heterogeneity in regard to methylation classes, with methylation classes glioblastoma and ganglioglioma being the most frequent mismatches. Similarly, the mcPXA cohort contained tumors of diverse histological diagnoses, with glioblastoma constituting the most frequent mismatch. Subsequent analyses demonstrated the presence of canonical pTERT mutations to be associated with unfavorable prognosis among mcPXA. Based on these data, we consider the tumor type PXA to be histologically more varied than previously assumed. Histological approach to diagnosis will predominantly identify cases with the established archetypical morphology. DNA methylation analysis includes additional tumors in the tumor class PXA that share similar DNA methylation profile but lack the typical morphology of a PXA. DNA methylation analysis also assist in separating other tumor types with morphologic overlap to PXA. Our data suggest the presence of canonical pTERT mutations as a robust indicator for poor prognosis in methylation class PXA.

Gene expression profiling of Group 3 medulloblastomas defines a clinically tractable stratification based on KIRREL2 expression

Medulloblastomas (MB) molecularly designated as Group 3 (Grp 3) MB represent a more clinically aggressive tumor variant which, as a group, displays heterogeneous molecular characteristics and disease outcomes. Reliable risk stratification of Grp 3 MB would allow for appropriate assignment of patients to aggressive treatment protocols and, vice versa, for sparing adverse effects of high-dose radio-chemotherapy in patients with standard or low-risk tumors. Here we performed RNA-based analysis on an international cohort of 179 molecularly designated Grp 3 MB treated with HIT protocols. We analyzed the clinical significance of differentially expressed genes, thereby developing optimal prognostic subdivision of this MB molecular group. We compared the transcriptome profiles of two Grp 3 MB subsets with various outcomes (76 died within the first 60 months vs. 103 survived this period) and identified 224 differentially expressed genes (DEG) between these two clinical groups (Limma R algorithm, adjusted p-value < 0.05). We selected the top six DEG overexpressed in the unfavorable cohort for further survival analysis and found that expression of all six genes strongly correlated with poor outcomes. However, only high expression of KIRREL2 was identified as an independent molecular prognostic indicator of poor patients' survival. Based on clinical and molecular patterns, four risk categories were outlined for Grp 3 MB patients: i. low-risk: M0-1/MYC non-amplified/KIRREL2 low (n = 48; 5-year OS-95%); ii. standard-risk: M0-1/MYC non-amplified/KIRREL2 high or M2-3/MYC non-amplified/KIRREL2 low (n = 65; 5-year OS-70%); iii. high-risk: M2-3/MYC non-amplified/KIRREL2 high (n = 36; 5-year OS-30%); iv. very high risk-all MYC amplified tumors (n = 30; 5-year OS-0%). Cross-validated survival models incorporating KIRREL2 expression with clinical features allowed for the reclassification of up to 50% of Grp 3 MB patients into a more appropriate risk category. Finally, KIRREL2 immunopositivity was also identified as a predictive indicator of Grp 3 MB poor survival, thus suggesting its application as a possible prognostic marker in routine clinical settings. Our results indicate that integration of KIRREL2 expression in risk stratification models may improve Grp 3 MB outcome prediction. Therefore, simple gene and/or protein expression analyses for this molecular marker could be easily adopted for Grp 3 MB prognostication and may help in assigning patients to optimal therapeutic approaches in prospective clinical trials.

Integrated Molecular-Morphologic Meningioma Classification: A Multicenter Retrospective Analysis, Retrospectively and Prospectively Validated

PURPOSE

Meningiomas are the most frequent primary intracranial tumors. Patient outcome varies widely from benign to highly aggressive, ultimately fatal courses. Reliable identification of risk of progression for individual patients is of pivotal importance. However, only biomarkers for highly aggressive tumors are established (CDKN2A/B and TERT), whereas no molecularly based stratification exists for the broad spectrum of patients with low- and intermediate-risk meningioma.

METHODS

DNA methylation data and copy-number information were generated for 3,031 meningiomas (2,868 patients), and mutation data for 858 samples. DNA methylation subgroups, copy-number variations (CNVs), mutations, and WHO grading were analyzed. Prediction power for outcome was assessed in a retrospective cohort of 514 patients, validated on a retrospective cohort of 184, and on a prospective cohort of 287 multicenter cases.

RESULTS

Both CNV- and methylation family-based subgrouping independently resulted in increased prediction accuracy of risk of recurrence compared with the WHO classification (c-indexes WHO 2016, CNV, and methylation family 0.699, 0.706, and 0.721, respectively). Merging all risk stratification approaches into an integrated molecular-morphologic score resulted in further substantial increase in accuracy (c-index 0.744). This integrated score consistently provided superior accuracy in all three cohorts, significantly outperforming WHO grading (c-index difference P = .005). Besides the overall stratification advantage, the integrated score separates more precisely for risk of progression at the diagnostically challenging interface of WHO grade 1 and grade 2 tumors (hazard ratio 4.34 [2.48-7.57] and 3.34 [1.28-8.72] retrospective and prospective validation cohorts, respectively).

CONCLUSION

Merging these layers of histologic and molecular data into an integrated, three-tiered score significantly improves the precision in meningioma stratification. Implementation into diagnostic routine informs clinical decision making for patients with meningioma on the basis of robust outcome prediction.

PATH-48. RAPID-CNS2: RAPID COMPREHENSIVE ADAPTIVE NANOPORE-SEQUENCING OF CNS TUMORS, A PROOF OF CONCEPT STUDY

BACKGROUND

The WHO classification 2021 includes multiple molecular markers for routine diagnostics in addition to histology. Sequencing setup for complete molecular profiling requires considerable investment, while batching samples for sequencing and methylation profiling can delay turnaround time. We introduce RAPID-CNS2, a nanopore adaptive sequencing pipeline that enables comprehensive mutational, methylation and copy number profiling of CNS tumours with a single third generation sequencing assay. It can be run for single samples and offers highly flexible target selection requiring no additional library preparation.

METHODS

Utilising ReadFish, a toolkit enabling targeted nanopore sequencing, we sequenced DNA from 22 diffuse glioma patient samples on a MinION device. Target regions comprised our Heidelberg brain tumour NGS panel and pre-selected CpG sites for methylation classification by an adapted random forest classifier. Pathognomonic alterations, copy number profiles, and methylation classes were called using a custom bioinformatics pipeline. Results were compared to their corresponding NGS panel-seq and EPIC array outputs.

RESULTS

Complete concordance with the EPIC array was found for copy number profiles from RAPID-CNS2. 94% pathognomonic mutations were congruent with NGS panel-seq. MGMT promoter status was correctly identified in all samples. Methylation families were detected with 96% congruence. Among the alterations decisive for rendering a classification-compatible integrated diagnosis, 97% of the alterations were consistent over the entire cohort (completely congruent in 19/22 cases and sufficient for unequivocal diagnosis in all).

CONCLUSIONS

RAPID-CNS2 provides a swift and highly flexible alternative to conventional NGS and array- based methods for SNV/Indel analysis, detection of copy number alterations and methylation classification. The turnaround time of ~4 days can be further shortened to &lt; 12h by altering target sizes. It offers a low-capital approach that would be cost-efficient for low throughput settings and invaluable in cases requiring immediate diagnoses.

PATH-39. INTEGRATED MOLECULAR-MORPHOLOGICAL MENINGIOMA CLASSIFICATION: A MULTICENTER RETROSPECTIVE ANALYSIS, RETRO- AND PROSPECTIVELY VALIDATED

PURPOSE

Meningiomas are the most frequent primary intracranial tumors. Patient outcome varies widely from cases with benign to highly aggressive, ultimately fatal courses. Reliable identification of risk of progression for the individual patient is of pivotal importance in clinical management. However, only biomarkers for highly aggressive tumors are established at present (CDKN2A/B and TERT), while no molecularly-based stratification exists for the broad spectrum of low- and intermediate-risk meningioma patients.

PATIENTS AND METHODS

DNA methylation data and copy-number information were generated for 3,031 meningiomas of 2,868 individual patients, with mutation data for 858 samples. DNA methylation subgroups, copy-number variations (CNV), mutations and WHO grading were comparatively analyzed. Prediction power for outcome of these parameters was assessed in an initial retrospective cohort of 514 patients, and validated on a retrospective cohort of 184, and on a prospective cohort of 287 multi-center cases, respectively.

RESULTS

Both CNV and methylation family- (MF)-based subgrouping independently resulted in an increase in prediction accuracy of risk of recurrence compared to the WHO classification (c-indexes WHO 2016, CNV, and MF 0.699, 0.706 and 0.721, respectively). Merging all independently powerful risk stratification approaches into an integrated molecular-morphological score resulted in a further, substantial increase in accuracy (c-index 0.744). This integrated score consistently provided superior accuracy in all three cohorts, significantly outperforming WHO grading (c-index difference p=0.005). Besides the overall stratification advantage, the integrated score separates more precisely for risk of progression at the diagnostically challenging interface of WHO grade 1 and grade 2 tumors (HR 4.56 [2.97;7.00], 4.34 [2.48;7.57] and 3.34 [1.28; 8.72] for discovery, retrospective, and prospective validation cohort, respectively).

CONCLUSIONS

Merging these layers of histological and molecular data into an integrated, three-tiered score significantly improves the precision in meningioma stratification. Implementation into diagnostic routine informs clinical decision-making for meningioma patients on the basis of robust outcome prediction.

PATH-23. OLIGOSARCOMA, IDH-MUTANT IS A DISTINCT AGGRESSIVE TYPE

Oligodendrogliomas are defined by IDH-mutations and the co-deletion of chromosomal arms 1p and 19q. In the past, case reports and small series described gliomas with sarcomatous features arising from oligodendrogliomas, so called oligosarcomas. Here, we report a group of 23 IDH-mutant oligosarcomas that form a distinct methylation group. The tumors were recurrences from prior oligodendrogliomas or developed de novo. Precursor tumors of 11 oligosarcomas were histologically and molecularly indistinguishable from conventional oligodendrogliomas. Tumor cells were embedded in a reticulin fibers network in all oligosarcomas, frequently showing p53 accumulation, positivity for SMA, and regain of H3K27me3 as compared to primary lesions. In 5 oligosarcomas no 1p/19q co-deletion was detectable, even though it was present in the primary lesions. Oligosarcomas harbored an increased chromosomal copy number variation load with frequent CDKN2A/B deletions. Proteomic profiling demonstrated oligosarcomas to be highly distinct from conventional grade 3 oligodendrogliomas with consistent evidence for smooth muscle differentiation. Expression of several tumor suppressors including NF1 and STAG2 was lost whereas oncogenic YAP1 was aberrantly overexpressed in oligosarcomas. Panel sequencing revealed mutations in TP53 and NF1 along with IDH1/2 and TERT promoter mutations. Survival of patients was significantly poorer for oligosarcomas as for grade 3 oligodendrogliomas but comparable to grade 4 IDH-mutant astrocytomas. These results establish oligosarcoma as a distinct type of IDH-mutant glioma differing from conventional oligodendrogliomas on the histologic, epigenetic, proteomic, molecular and clinical level. Diagnosis can be based on the characteristic DNA methylation profile or the combined evidence of sarcomatous histology, IDH-mutation and TERT promoter mutation with or without 1p/19q co-deletion.

Recurrent fusions in PLAGL1 define a distinct subset of pediatric-type supratentorial neuroepithelial tumors

Ependymomas encompass a heterogeneous group of central nervous system (CNS) neoplasms that occur along the entire neuroaxis. In recent years, extensive (epi-)genomic profiling efforts have identified several molecular groups of ependymoma that are characterized by distinct molecular alterations and/or patterns. Based on unsupervised visualization of a large cohort of genome-wide DNA methylation data, we identified a highly distinct group of pediatric-type tumors (n = 40) forming a cluster separate from all established CNS tumor types, of which a high proportion were histopathologically diagnosed as ependymoma. RNA sequencing revealed recurrent fusions involving the pleomorphic adenoma gene-like 1 (PLAGL1) gene in 19 of 20 of the samples analyzed, with the most common fusion being EWSR1:PLAGL1 (n = 13). Five tumors showed a PLAGL1:FOXO1 fusion and one a PLAGL1:EP300 fusion. High transcript levels of PLAGL1 were noted in these tumors, with concurrent overexpression of the imprinted genes H19 and IGF2, which are regulated by PLAGL1. Histopathological review of cases with sufficient material (n = 16) demonstrated a broad morphological spectrum of tumors with predominant ependymoma-like features. Immunohistochemically, tumors were GFAP positive and OLIG2- and SOX10 negative. In 3/16 of the cases, a dot-like positivity for EMA was detected. All tumors in our series were located in the supratentorial compartment. Median age of the patients at the time of diagnosis was 6.2 years. Median progression-free survival was 35 months (for 11 patients with data available). In summary, our findings suggest the existence of a novel group of supratentorial neuroepithelial tumors that are characterized by recurrent PLAGL1 fusions and enriched for pediatric patients.

Clinically Tractable Outcome Prediction of non-WNT/non-SHH Medulloblastoma Based on TPD52 Immunohistochemistry in a Multicohort Study

PURPOSE

International consensus and the 2021 WHO classification recognize eight molecular subgroups among non-WNT/non-SHH (Group 3/4) medulloblastoma, representing ~60% of tumors. However, very few clinical centers worldwide possess the technical capabilities to determine DNA-methylation profiles or other molecular parameters of high-risk for Group 3/4 tumors. As a result, biomarker-driven risk stratification and therapy assignment constitutes a major challenge in medulloblastoma research. Here, we identify an immunohistochemistry (IHC) marker as a clinically tractable method for improved medulloblastoma risk stratification.

EXPERIMENTAL DESIGN

We bioinformatically analyzed published medulloblastoma transcriptomes and proteomes identifying as a potential biomarker TPD52, whose IHC prognostic value was validated across three Group 3/4 medulloblastoma clinical cohorts (n = 387) treated with conventional therapies.

RESULTS

TPD52 IHC positivity represented a significant independent predictor of early relapse and death for Group 3/4 medulloblastoma (HRs between 3.67-26.7 [95% CIs between 1.00-706.23], p = 0.05, 0.017 and 0.0058). Cross-validated survival models incorporating TPD52 IHC with clinical features outperformed existing state-of-the-art risk stratification schemes, and reclassified ~50% of patients into more appropriate risk categories. Finally, TPD52 immunopositivity was a predictive indicator of poor response to chemotherapy (HR 12.66 [95% CI 3.53-45.40], p < 0.0001), suggesting important implication for therapeutic choices.

CONCLUSIONS

The current study redefines the approach to risk stratification in Group 3/4 medulloblastoma in global practice. Since integration of TPD52 IHC in classification algorithms significantly improved outcome prediction, this test could be rapidly adopted for risk stratification on a global scale, independently of advanced but technically challenging molecular profiling techniques.

Integrated molecular analysis of adult sonic hedgehog (SHH)-activated medulloblastomas reveals two clinically relevant tumor subsets with VEGFA as potent prognostic indicator

BACKGROUND

Up to now, adult medulloblastoma (MB) patients are treated according to the protocols elaborated for pediatric MB although these tumors are different in terms of clinical outcomes and biology. Approximately 70% of adult MB disclose a sonic hedgehog (SHH) molecular signature in contrast to about 30% in pediatric cohorts. In addition, adult SHH-MB (aSHH-MB) are clinically heterogeneous but there is consensus neither on their optimal treatment nor on risk stratification. Thus, the identification of clinically relevant molecular subsets of aSHH-MB and identification of potential treatment targets remains inconclusive.

METHODS

We analyzed 96 samples of institutionally diagnosed aSHH-MB through genome-wide DNA methylation profiling, targeted DNA sequencing, and RNA sequencing to identify molecular subcategories of these tumors and assess their prognostic significance.

RESULTS

We defined two aSHH-MB numerically comparable epigenetic subsets with clinical and molecular variability. The subset "aSHH-MBI" (46%/48%) was associated with PTCH1/SMO (54%/46%) mutations, "neuronal" transcriptional signatures, and favorable outcomes after combined radio-chemotherapy (5-year PFS = 80% and OS = 92%). The clinically unfavorable "aSHH-MBII" subset (50%/52%; 5-year PFS = 24% and OS = 45%) disclosed GLI2 amplifications (8%), loss of 10q (22%), and gene expression signatures associated with angiogenesis and embryonal development. aSHH-MBII tumors revealed strong and ubiquitous expression of VEGFA both at transcript and protein levels that was correlated with unfavorable outcome.

CONCLUSIONS

(1) The histologically uniform aSHH-MB cohort exhibits clear molecular heterogeneity separating these tumors into two molecular subsets (aSHH-MBI and aSHH-MBII), which are associated with different cytogenetics, mutational landscapes, gene expression signatures, and clinical course. (2) VEGFA appears to be a promising biomarker to predict clinical course, which needs further prospective validation as its potential role in the pathogenesis of this subset.

Glioblastomas with primitive neuronal component harbor a distinct methylation and copy-number profile with inactivation of TP53, PTEN, and RB1

Glioblastoma IDH-wildtype presents with a wide histological spectrum. Some features are so distinctive that they are considered as separate histological variants or patterns for the purpose of classification. However, these usually lack defined (epi-)genetic alterations or profiles correlating with this histology. Here, we describe a molecular subtype with overlap to the unique histological pattern of glioblastoma with primitive neuronal component. Our cohort consists of 63 IDH-wildtype glioblastomas that harbor a characteristic DNA methylation profile. Median age at diagnosis was 59.5 years. Copy-number variations and genetic sequencing revealed frequent alterations in TP53, RB1 and PTEN, with fewer gains of chromosome 7 and homozygous CDKN2A/B deletions than usually described for IDH-wildtype glioblastoma. Gains of chromosome 1 were detected in more than half of the cases. A poorly differentiated phenotype with frequent absence of GFAP expression, high proliferation index and strong staining for p53 and TTF1 often caused misleading histological classification as carcinoma metastasis or primitive neuroectodermal tumor. Clinically, many patients presented with leptomeningeal dissemination and spinal metastasis. Outcome was poor with a median overall survival of only 12 months. Overall, we describe a new molecular subtype of IDH-wildtype glioblastoma with a distinct histological appearance and genetic signature.

DNA methylation-based profiling of bone and soft tissue tumours: a validation study of the 'DKFZ Sarcoma Classifier'

Diagnosing bone and soft tissue neoplasms remains challenging because of the large number of subtypes, many of which lack diagnostic biomarkers. DNA methylation profiles have proven to be a reliable basis for the classification of brain tumours and, following this success, a DNA methylation-based sarcoma classification tool from the Deutsches Krebsforschungszentrum (DKFZ) in Heidelberg has been developed. In this study, we assessed the performance of their classifier on DNA methylation profiles of an independent data set of 986 bone and soft tissue tumours and controls. We found that the 'DKFZ Sarcoma Classifier' was able to produce a diagnostic prediction for 55% of the 986 samples, with 83% of these predictions concordant with the histological diagnosis. On limiting the validation to the 820 cases with histological diagnoses for which the DKFZ Classifier was trained, 61% of cases received a prediction, and the histological diagnosis was concordant with the predicted methylation class in 88% of these cases, findings comparable to those reported in the DKFZ Classifier paper. The classifier performed best when diagnosing mesenchymal chondrosarcomas (CHSs, 88% sensitivity), chordomas (85% sensitivity), and fibrous dysplasia (83% sensitivity). Amongst the subtypes least often classified correctly were clear cell CHSs (14% sensitivity), malignant peripheral nerve sheath tumours (27% sensitivity), and pleomorphic liposarcomas (29% sensitivity). The classifier predictions resulted in revision of the histological diagnosis in six of our cases. We observed that, although a higher tumour purity resulted in a greater likelihood of a prediction being made, it did not correlate with classifier accuracy. Our results show that the DKFZ Classifier represents a powerful research tool for exploring the pathogenesis of sarcoma; with refinement, it has the potential to be a valuable diagnostic tool.

EMBR-21. CLINICALLY TRACTABLE OUTCOME PREDICTION OF GROUP 3/4 MEDULLOBLASTOMA BASED ON TPD52 IMMUNOHISTOCHEMISTRY: A MULTICOHORT STUDY

Background

International consensus and the 2021 WHO classification recognize eight molecular subtypes among Group 3/4 medulloblastoma (representing ~60% of tumors). However, very few clinical centers worldwide possess the technical capabilities to determine DNA-methylation patterns or other molecular parameters of high-risk for Group 3/4 tumors. As a result, biomarker-driven risk stratification and therapy assignment constitutes a major challenge in medulloblastoma research. Here, we identify an immunohistochemistry (IHC) marker as a clinically tractable method for improved medulloblastoma risk-stratification.

Patients and Methods

We bioinformatically analyzed published medulloblastoma transcriptomes and proteomes identifying as a potential biomarker TPD52, whose IHC prognostic value was validated across three Group 3/4 medulloblastoma clinical cohorts (n = 387) treated with conventional therapies. Risk stratification and prediction capability were computed utilizing uni- and multivariate survival analysis. Newly developed risk classifiers including TPD52 IHC were compared to state-of-the-art risk stratification schemes in terms of prediction error, area under the time-dependent receiver operating characteristic (ROC) curves and C-statistic. Biomarker-driven prognostic stratification models identified were cross validated in different cohorts.

Results

TPD52 IHC positivity represents a significant independent predictor of early relapse and death for Group 3/4 medulloblastoma (HRs between 3.67–26.7 [95% CIs between 1.00–706.23], p = 0.05, 0.017 and 0.0058). Cross-validated survival models incorporating TPD52 IHC with clinical features outperformed existing disease risk-stratification schemes, and reclassified ~50% of patients into more appropriate risk categories. Finally, TPD52 immunopositivity is a predictive indicator of poor response to chemotherapy (HR 12.66 [95% CI 3.53–45.40], p < 0.0001), suggesting important implication for therapeutic choices.

Conclusion

The current study redefines the approach to risk-stratification in Group 3/4 medulloblastoma. Integration of TPD52 IHC in classification algorithms significantly improves outcome prediction and can be rapidly adopted for risk stratification on a global scale, independently of advanced but technically challenging molecular profiling techniques.

Molecular analysis of pediatric CNS-PNET revealed nosologic heterogeneity and potent diagnostic markers for CNS neuroblastoma with FOXR2-activation

Primitive neuroectodermal tumors of the central nervous system (CNS-PNETs) are highly malignant neoplasms posing diagnostic challenge due to a lack of defining molecular markers. CNS neuroblastoma with forkhead box R2 (FOXR2) activation (CNS_NBL) emerged as a distinct pediatric brain tumor entity from a pool previously diagnosed as primitive neuroectodermal tumors of the central nervous system (CNS-PNETs). Current standard of identifying CNS_NBL relies on molecular analysis. We set out to establish immunohistochemical markers allowing safely distinguishing CNS_NBL from morphological mimics. To this aim we analyzed a series of 84 brain tumors institutionally diagnosed as CNS-PNET. As expected, epigenetic analysis revealed different methylation groups corresponding to the (1) CNS-NBL (24%), (2) glioblastoma IDH wild-type subclass H3.3 G34 (26%), (3) glioblastoma IDH wild-type subclass MYCN (21%) and (4) ependymoma with RELA_C11orf95 fusion (29%) entities. Transcriptome analysis of this series revealed a set of differentially expressed genes distinguishing CNS_NBL from its mimics. Based on RNA-sequencing data we established SOX10 and ANKRD55 expression as genes discriminating CNS_NBL from other tumors exhibiting CNS-PNET. Immunohistochemical detection of combined expression of SOX10 and ANKRD55 clearly identifies CNS_NBL discriminating them to other hemispheric CNS neoplasms harboring “PNET-like” microscopic appearance. Owing the rarity of CNS_NBL, a confirmation of the elaborated diagnostic IHC algorithm will be necessary in prospective patient series.

A subset of pediatric-type thalamic gliomas share a distinct DNA methylation profile, H3K27me3 loss and frequent alteration of EGFR

BACKGROUND

Malignant astrocytic gliomas in children show a remarkable biological and clinical diversity. Small in-frame insertions or missense mutations in the epidermal growth factor receptor gene (EGFR) have recently been identified in a distinct subset of pediatric-type bithalamic gliomas with a unique DNA methylation pattern.

METHODS

Here, we investigated an epigenetically homogeneous cohort of malignant gliomas (n = 58) distinct from other subtypes and enriched for pediatric cases and thalamic location, in comparison with this recently identified subtype of pediatric bithalamic gliomas.

RESULTS

EGFR gene amplification was detected in 16/58 (27%) tumors, and missense mutations or small in-frame insertions in EGFR were found in 20/30 tumors with available sequencing data (67%; 5 of them co-occurring with EGFR amplification). Additionally, 8 of the 30 tumors (27%) harbored an H3.1 or H3.3 K27M mutation (6 of them with a concomitant EGFR alteration). All tumors tested showed loss of H3K27me3 staining, with evidence of overexpression of the EZH inhibitory protein (EZHIP) in the H3 wildtype cases. Although some tumors indeed showed a bithalamic growth pattern, a significant proportion of tumors occurred in the unilateral thalamus or in other (predominantly midline) locations.

CONCLUSIONS

Our findings present a distinct molecular class of pediatric-type malignant gliomas largely overlapping with the recently reported bithalamic gliomas characterized by EGFR alteration, but additionally showing a broader spectrum of EGFR alterations and tumor localization. Global H3K27me3 loss in this group appears to be mediated by either H3 K27 mutation or EZHIP overexpression. EGFR inhibition may represent a potential therapeutic strategy in these highly aggressive gliomas.

Accurate calling of KIAA1549-BRAF fusions from DNA of human brain tumours using methylation array-based copy number and gene panel sequencing data

AIMS

KIAA1549-BRAF fusions occur in certain brain tumours and provide druggable targets due to a constitutive activation of the MAP-kinase pathway. We introduce workflows for calling the KIAA1549-BRAF fusion from DNA methylation array-derived copy number as well as DNA panel sequencing data.

METHODS

Copy number profiles were analysed by automated screening and visual verification of a tandem duplication on chromosome 7q34, indicative of the KIAA1549-BRAF fusion. Pilocytic astrocytomas of the ICGC cohort with known fusion status were used for validation. KIAA1549-BRAF fusions were called from DNA panel sequencing data using the fusion callers Manta, Arriba with modified filtering criteria and deFuse. We screened DNA methylation and panel sequencing data of 7790 specimens from brain tumour and sarcoma entities.

RESULTS

We identified the fusion in 337 brain tumours with both DNA methylation and panel sequencing data. Among these, we detected the fusion from copy number data in 84% and from DNA panel sequencing data in more than 90% using Arriba with modified filters. While in 74% the KIAA1549-BRAF fusion was detected from both methylation array-derived copy number and panel sequencing data, in 9% it was detected from copy number data only and in 16% from panel data only. The fusion was almost exclusively found in pilocytic astrocytomas, diffuse leptomeningeal glioneuronal tumours and high-grade astrocytomas with piloid features.

CONCLUSIONS

The KIAA1549-BRAF fusion can be reliably detected from either DNA methylation array or DNA panel data. The use of both methods is recommended for the most sensitive detection of this diagnostically and therapeutically important marker.

Primary mismatch repair deficient IDH-mutant astrocytoma (PMMRDIA) is a distinct type with a poor prognosis

Diffuse IDH-mutant astrocytoma mostly occurs in adults and carries a favorable prognosis compared to IDH-wildtype malignant gliomas. Acquired mismatch repair deficiency is known to occur in recurrent IDH-mutant gliomas as resistance mechanism towards alkylating chemotherapy. In this multi-institutional study, we report a novel epigenetic group of 32 IDH-mutant gliomas with proven or suspected hereditary mismatch repair deficiency. None of the tumors exhibited a combined 1p/19q deletion. These primary mismatch repair-deficient IDH-mutant astrocytomas (PMMRDIA) were histologically high-grade and were mainly found in children, adolescents and young adults (median age 14 years). Mismatch repair deficiency syndromes (Lynch or Constitutional Mismatch Repair Deficiency Syndrom (CMMRD)) were clinically diagnosed and/or germline mutations in DNA mismatch repair genes (MLH1, MSH6, MSH2) were found in all cases, except one case with a family and personal history of colon cancer and another case with MSH6-deficiency available only as recurrent tumor. Loss of at least one of the mismatch repair proteins was detected via immunohistochemistry in all, but one case analyzed. Tumors displayed a hypermutant genotype and microsatellite instability was present in more than half of the sequenced cases. Integrated somatic mutational and chromosomal copy number analyses showed frequent inactivation of TP53, RB1 and activation of RTK/PI3K/AKT pathways. In contrast to the majority of IDH-mutant gliomas, more than 60% of the samples in our cohort presented with an unmethylated MGMT promoter. While the rate of immuno-histochemical ATRX loss was reduced, variants of unknown significance were more frequently detected possibly indicating a higher frequency of ATRX inactivation by protein malfunction. Compared to reference cohorts of other IDH-mutant gliomas, primary mismatch repair-deficient IDH-mutant astrocytomas have by far the worst clinical outcome with a median survival of only 15 months irrespective of histological or molecular features. The findings reveal a so far unknown entity of IDH-mutant astrocytoma with high prognostic relevance. Diagnosis can be established by aligning with the characteristic DNA methylation profile, by DNA-sequencing-based proof of mismatch repair deficiency or immunohistochemically demonstrating loss-of-mismatch repair proteins.

PATH-04. AN ENHANCED AI-DRIVEN PLATFORM FOR PRECISION MOLECULAR BRAIN TUMOR DIANOSTICS

Tumors of the CNS represent one of the most complex groups of human cancer, with a vast number of different entities occurring across a spectrum of ages and anatomic locations. This heterogeneity makes accurate diagnosis challenging, with the current gold standard relying on multiple subjective elements. We recently proposed a classification algorithm based on tumor DNA methylation profiling as an objective way to assign samples to over 80 distinct molecular classes. Here we present a substantial update to our machine learning-based algorithm, with more than 170 molecular classes now being represented amongst the 5,915 samples in our reference cohort. These new classes include further subclassification of known groups such as medulloblastoma and ependymoma, as well as multiple new molecular entities described here for the first time. A further improvement is the introduction of a more rationally layered output, making use of ‘families’ of closely-related molecular classes to improve the compatibility with the current WHO classification of CNS tumors. This approach is designed to increase the clinical relevance of the primary output, while also retaining the full information content from the random forest-driven classification. Benchmarking our new algorithm by cross-validation and on an independent validation cohort indicates a retention of the excellent accuracy of diagnosis (error-rate &lt; 4%), with a significant improvement in the proportion of confidently classifiable tumors compared with our previous tool. We believe that this approach, freely accessible through an online web portal, has the potential to enhance diagnostic precision and thereby support clinical care for brain tumor patients.

PATH-26. RNA SEQUENCING OF FORMALIN-FIXED PARAFFIN-EMBEDDED SPECIMENS IN DIAGNOSTIC ROUTINE IDENTIFIES CLINICALLY RELEVANT GENE FUSIONS

Pediatric brain tumor entities harbor a variety of gene fusions. Whilst other molecular parameters like somatic mutations and copy number alterations have become pivotal for brain tumor diagnostics, gene fusions are only less well covered by routinely applied methylation arrays or targeted next-generation sequencing of DNA. In a routine diagnostic setting we established and optimized a workflow for investigation of gene fusions in formalin-fixed paraffin-embedded (FFPE) tumor tissues by using RNA sequencing. Assessing different tools for calling fusions from raw data, we found relevant fusions in 66 out of 101 (65%) analyzed cases in a prospective cohort collected over 26 months. In 43 (43%) cases the fusions were of decisive diagnostic relevance and in 40 (40%) cases the fusion genes rendered a druggable target. Besides the relevance of pathognomonic fusions for diagnostics, especially the detection of druggable gene fusions yields direct benefit to the patients. This approach allows for an unbiased search for fusion events in the tested samples. Besides rare variants of established fusions which were not detected by prior targeted analyses, we identified previously unreported fusion events. Exemplified on KIAA1549:BRAF fusion, we in addition provide an overview of the detection accuracy of different methods, including breakpoint detection in DNA methylation array data and fusion gene detection in DNA panel sequencing data. Our data show that RNA sequencing has great diagnostic as well as therapeutic value by clinically detecting relevant alterations.

LGG-33. ISOMORPHIC DIFFUSE GLIOMA HAS RECURRENT GENE FUSIONS OF MYBL1 OR MYB AND CAN BE DISTINGUISHED FROM OTHER MYB/MYBL1 ALTERED GLIOMAS BASED ON A DISTINCT MORPHOLOGY AND DNA METHYLATION PROFILE

Isomorphic diffuse glioma (IDG) was first described in 2004 as an epilepsy-associated supratentorial diffuse glioma with low cellularity, low proliferation and very monomorphic tumour cells. Most patients had seizures since childhood but were operated on as adults. To study the position of these lesions among brain tumours we histologically, molecularly and clinically analysed 26 histologically typical IDGs. Tumour cells were GFAP-positive, MAP2-, OLIG2- and CD34-negative and the nuclear ATRX-expression was retained. Proliferation was very low. Sequencing of 24 cases revealed an IDH-wildtype status. Cluster analyses of DNA methylation data showed that IDG has a DNA methylation profile distinct from those of different glial/glio-neuronal brain tumours and normal hemispheric tissue. About half of IDGs had copy number alterations of MYBL1 or MYB (13/25) and half of the cases analysed by RNA-sequencing had gene fusions of MYBL1 or MYB with various gene partners (11/22), often associated with an increased RNA-expression of the respective MYB-family gene. Integrating all data available, 77% of IDGs had either MYBL1 (54%) or MYB (23%) alterations. All patients had a good outcome and most were seizure-free after surgery. In summary, we show that isomorphic diffuse glioma is a distinct benign tumour in the family of MYB/MYBL1-altered gliomas. DNA methylation analysis is very helpful for their identification. More recent analyses of a large cohort of MYB/MYBL1-altered brain tumours suggest the presence of a third methylation group that primarily contains paediatric cases and seems to be distinct from IDG and angiocentric gliomas. Further histological, molecular and clinical analyses are ongoing.

Infratentorial IDH-mutant astrocytoma is a distinct subtype

Diffuse IDH-mutant astrocytic tumors are rarely diagnosed in the cerebellum or brainstem. In this multi-institutional study, we characterized a series of primary infratentorial IDH-mutant astrocytic tumors with respect to clinical and molecular parameters. We report that about 80% of IDH mutations in these tumors are of non-IDH1-R132H variants which are rare in supratentorial astrocytomas. Most frequently, IDH1-R132C/G and IDH2-R172S/G mutations were present. Moreover, the frequencies of ATRX-loss and MGMT promoter methylation, which are typically associated with IDH mutations in supratentorial astrocytic tumors, were significantly lower in the infratentorial compartment. Gene panel sequencing revealed two samples with IDH1-R132C/H3F3A-K27M co-mutations. Genome-wide DNA methylation as well as chromosomal copy number profiling provided further evidence for a molecular distinctiveness of infratentorial IDH-mutant astrocytomas. Clinical outcome of patients with infratentorial IDH-mutant astrocytomas is significantly better than that of patients with diffuse midline gliomas, H3K27M-mutant (p < 0.005) and significantly worse than that of patients with supratentorial IDH-mutant astrocytomas (p = 0.028). The presented data highlight the very existence and distinctiveness of infratentorial IDH-mutant astrocytomas that have important implications for diagnostics and prognostication. They imply that molecular testing is critical for detection of these tumors, since many of these tumors cannot be identified by immunohistochemistry applied for the mutated IDH1-R132H protein or loss of ATRX.

DNA methylation profiling is a method of choice for molecular verification of pediatric WNT-activated medulloblastomas

BACKGROUND

Wingless-activated medulloblastoma (WNT MB) represents a well-characterized molecular variant accounting for 10-15% of all MB and is associated with a favorable clinical outcome. Patients with localized WNT MBs could benefit from de-intensification of combined treatment, which would require an accurate diagnosis of these tumors. However, despite the presence of molecular features related with a WNT MB signature (nuclear ß-catenin immunoexpression, CTNNB1 mutation, and monosomy 6), a prompt and reliable diagnostic verification of these tumors is not yet feasible.

METHODS

In the current study, we analyzed 78 samples of WNT MB treated in a single institute through genome-wide DNA methylation and targeted next generation sequencing to elaborate an optimal method for WNT MB molecular verification.

RESULTS

We found that DNA methylation profiling discloses significant advantages for molecular diagnostic of WNT MB. All other "routine" methods applied, such as ß-catenin immunohistochemistry, CTNNB1 mutation analysis, and detection of monosomy 6, failed to identify all WNT MB cases. Survival analysis revealed that application of a reduced radiotherapy protocol for WNT MB treatment had no influence on patients' survival. Only one patient died due to local relapse but recurrent tumor was pathologically and molecularly diagnosed as a secondary glioblastoma.

CONCLUSIONS

DNA methylation analysis should be considered as a method of choice for further clinically relevant stratification of WNT MB and for correct diagnosis of the recurrent tumors. WNT MB patients with localized disease could benefit from treatment de-intensification.

Isomorphic diffuse glioma is a morphologically and molecularly distinct tumour entity with recurrent gene fusions of MYBL1 or MYB and a benign disease course

The "isomorphic subtype of diffuse astrocytoma" was identified histologically in 2004 as a supratentorial, highly differentiated glioma with low cellularity, low proliferation and focal diffuse brain infiltration. Patients typically had seizures since childhood and all were operated on as adults. To define the position of these lesions among brain tumours, we histologically, molecularly and clinically analysed 26 histologically prototypical isomorphic diffuse gliomas. Immunohistochemically, they were GFAP-positive, MAP2-, OLIG2- and CD34-negative, nuclear ATRX-expression was retained and proliferation was low. All 24 cases sequenced were IDH-wildtype. In cluster analyses of DNA methylation data, isomorphic diffuse gliomas formed a group clearly distinct from other glial/glio-neuronal brain tumours and normal hemispheric tissue, most closely related to paediatric MYB/MYBL1-altered diffuse astrocytomas and angiocentric gliomas. Half of the isomorphic diffuse gliomas had copy number alterations of MYBL1 or MYB (13/25, 52%). Gene fusions of MYBL1 or MYB with various gene partners were identified in 11/22 (50%) and were associated with an increased RNA-expression of the respective MYB-family gene. Integrating copy number alterations and available RNA sequencing data, 20/26 (77%) of isomorphic diffuse gliomas demonstrated MYBL1 (54%) or MYB (23%) alterations. Clinically, 89% of patients were seizure-free after surgery and all had a good outcome. In summary, we here define a distinct benign tumour class belonging to the family of MYB/MYBL1-altered gliomas. Isomorphic diffuse glioma occurs both in children and adults, has a concise morphology, frequent MYBL1 and MYB alterations and a specific DNA methylation profile. As an exclusively histological diagnosis may be very challenging and as paediatric MYB/MYBL1-altered diffuse astrocytomas may have the same gene fusions, we consider DNA methylation profiling very helpful for their identification.

DNA methylation-based profiling of uterine neoplasms: a novel tool to improve gynecologic cancer diagnostics

PURPOSE

Uterine neoplasms comprise a broad spectrum of lesions, some of which may pose a diagnostic challenge even to experienced pathologists. Recently, genome-wide DNA methylation-based classification of central nervous system tumors has been shown to increase diagnostic precision in clinical practice when combined with standard histopathology. In this study, we describe DNA methylation patterns of a diverse set of uterine neoplasms and test the applicability of array-based DNA methylation profiling.

METHODS

A multicenter cohort including prototypical epithelial and mesenchymal uterine neoplasms was collected. Tumors were subject to pathology review and array-based DNA methylation profiling (Illumina Infinium HumanMethylation450 or EPIC [850k] BeadChip). Methylation data were analyzed by unsupervised hierarchical clustering and t-SNE analysis.

RESULTS

After sample retrieval and pathology review the study cohort consisted of 49 endometrial carcinomas (EC), 5 carcinosarcomas (MMMT), 8 uterine leiomyomas (ULMO), 7 uterine leiomyosarcomas (ULMS), 15 uterine tumor resembling ovarian sex cord tumors (UTROSCT), 17 low-grade endometrial stromal sarcomas (LGESS) and 9 high-grade endometrial stromal sarcomas (HGESS). Analysis of methylation data identified distinct methylation clusters, which correlated with established diagnostic categories of uterine neoplasms. MMMT clustered together with EC, while ULMO, ULMS and UTROSCT each formed distinct clusters. The LGESS cluster differed from that of HGESS, and within the branch of HGESS, we observed a notable subgrouping of YWHAE- and BCOR-rearranged tumors.

CONCLUSION

Herein, we describe distinct DNA methylation signatures in uterine neoplasms and show that array-based DNA methylation analysis holds promise as an ancillary tool to further characterize uterine neoplasms, especially in cases which are diagnostically challenging by conventional techniques.

Bevacizumab plus hypofractionated radiotherapy versus radiotherapy alone in elderly patients with glioblastoma: the randomized, open-label, phase II ARTE trial

Background

The addition of bevacizumab to temozolomide-based chemoradiotherapy (TMZ/RT → TMZ) did not prolong overall survival (OS) in patients with newly diagnosed glioblastoma in phase III trials. Elderly and frail patients are underrepresented in clinical trials, but early reports suggested preferential benefit in this population.

Patients and methods

ARTE was a 2 : 1 randomized, multi-center, open-label, non-comparative phase II trial of hypofractionated RT (40 Gy in 15 fractions) with bevacizumab (10 mg/kg×14 days) (arm A, N = 50) or without bevacizumab (arm B, N = 25) in patients with newly diagnosed glioblastoma aged ≥65 years. The primary objective was to obtain evidence for prolongation of median OS by the addition of bevacizumab to RT. Response was assessed by RANO criteria. Quality of life (QoL) was monitored by the EORTC QLQ-C30/BN20 modules. Exploratory studies included molecular subtyping by 450k whole methylome and gene expression analyses.

Results

Median PFS was longer in arm A than in arm B (7.6 and 4.8 months, P = 0.003), but OS was similar (12.1 and 12.2 months, P = 0.77). Clinical deterioration was delayed and more patients came off steroids in arm A. Prolonged PFS in arm A was confined to tumors with the receptor tyrosine kinase (RTK) I methylation subtype (HR 0.25, P = 0.014) and proneural gene expression (HR 0.29, P = 0.025). In a Cox model of OS controlling for established prognostic factors, associations with more favorable outcome were identified for age <70 years (HR 0.52, P = 0.018) and Karnofsky performance score 90%-100% (HR 0.51, P = 0.026). Including molecular subtypes into that model identified an association of the RTK II gene methylation subtype with inferior OS (HR 1.73, P = 0.076).

Conclusion

Efficacy outcomes and exploratory analyses of ARTE do not support the hypothesis that the addition of bevacizumab to RT generally prolongs survival in elderly glioblastoma patients. Molecular biomarkers may identify patients with preferential benefit from bevacizumab.

Clinical trial registration number

NCT01443676.