Preoperative classification of brain tumors is critical to developing personalized treatment plans, however existing classification methods rely on manual intervention and often have problems with efficiency and accuracy, which may lead to misdiagnosis or delayed diagnosis in clinical practice and affect the therapeutic effect. We propose a fully automated approach to brain tumor magnetic resonance imaging (MRI) classification, consisted by a feature extractor based on the improved U-Net and a classifier based on convolutional recurrent neural network (CRNN). The encoder of the feature extractor based on dense block, is used to enhance feature propagation and reduce the number of parameters. The decoder uses residual block to reduce the weight of some features for improving the effect of MRI spatial sequence reconstruction, and avoid gradient disappearance. Skip connections between the encoder and the decoder effectively merge low-level features and high-level features. The extract feature sequence is input into the CRNN-based classifier for final classification. We assessed the performance of our method for grading glioma, glioma isocitrate dehydrogenase1 (IDH1) mutation status classification and pituitary tumor texture classification on two datasets, glioma or pituitary tumors collected in a local affiliated hospital and glioma imaging data from TCIA. Compared with commonly models and new models, our model achieves higher accuracy, with an accuracy of 90.72%, classified glioma IDH1 mutation status with an accuracy of 94.35%, and classified pituitary tumor texture with an accuracy of 94.64%.

Lower grade glioma (LGG) can be clinically grouped into major subtypes based on the mutation of the IDH1 and IDH2 genes and subsequently the co-deletion of the chromosome arms 1p and 19q. Chemical modification of DNA cytosine base at the cytosine-phosphate-guanine (CpG) di-nucleotide context plays an important role in mammalian cells and is altered severely in cancer including LGG. Due to the impact of IDH genes on genome-wide CpG methylation, the LGG subtypes show characteristically distinct methylation landscapes. Therefore, it is very feasible to use CpG methylation profiles to predict glioma subtypes, which are multi-class and hierarchically natured.

Here, two cohorts of LGGs with both genome-wide methylation profiles were curated from sources including Synapse TCGA Live, cBioPortal, and Gene Expression Omnibus data repositories.

• The Cancer Genome Atlas (n=507, "American Cohort")
• German Glioma Network (n=122, "German Cohort")

In both cohorts, CpG methylation was measured using the Illumina HumanMethylation450 BeadChip platform at the individual CpG level. In this Kaggle dataset, the methylation datasets have been re-processed such that:

• Instead of individual CpGs, CpG islands (CGIs), where CpG sites are densely distributed, are used as the feature unit. The value of each CGI was calculated by unweighted averaging the methylation beta-values of all available CpGs in each dataset, and then logit-transformed (i.e. M-values). The name of a CGI is simply its chromosomal coordinate on the human hg19 genome build.
• Only CGIs on chromosomes 1-22, longer than 800 nucleotides, and have at least 6 Illumina methylation probes were included

Each cohort's clinical (including subtype, the main outcome of interest) and processed CGI methylation data are organized into a CSV file. Each row is a tumor sample. Both cohorts have a column named "dummy", which is coded from the column "Subtype" where IDH-normal (wild type) = 2, IDH-mutated only = 0, and IDH mutated with co-deletion = 1. This column can be conveniently used for predictive modeling. The methylation M-values of individual CGIs (predictive features) are the 3,017 columns after the "dummy" column.

Abstract

This dataset includes MR imaging from 203 glioma patients with 617 different post-treatment MR time points, and tumor segmentations. Clinical data includes patient demographics, genomics, and treatment details. Preprocessing of MR images followed a standardized pipeline with automatic tumor segmentation based on nnUNet deep learning approach. The automatic tumor segmentations were manually validated and refined by neuroradiologists.

The heterogeneity of glioma imaging characteristics and management strategies contributes to a lack of reliable findings when evaluating treatment outcomes with conventional MRI, and the overlapping imaging features of radiation necrosis and tumor progression post-treatment can be particularly challenging for radiologists. This robust dataset should contribute to the development of AI models to improve evaluation of treatment outcomes.

Introduction

The dataset consists of institutional review board-approved retrospective analysis of pathologically proven glioma patients at University Hospital of The University of Missouri - Anatomic Pathology CoPathPlus database was used to collect glioma cases over the last 10 years.

Sharing segmented postoperative glioma data with clinical information significantly accelerates research and improves clinical practice by providing a comprehensive, readily available dataset. This eliminates the time-consuming burden of manual segmentation, enhances the accuracy and consistency of tumor delineation, and allows researchers to focus on analysis and interpretation, ultimately driving the development of more accurate segmentation algorithms, predictive models for personalized treatment strategies, and improved patient outcome predictions. Standardized longitudinal follow-up and benchmarking capabilities further facilitate multi-center studies and objective evaluation of treatment efficacy, leading to advancements in glioma biology and personalized patient care.

Methods

The following subsections provide information about how the data were selected, acquired, and prepared for publication.

Subject Inclusion and Exclusion Criteria

The selection criteria for the CoPath Natural Language II Search included accession dates ranging from 01/01/2021 to 02/20/2024. To ensure all relevant diagnoses for this study were included; three separate keyword searches were performed using "glioma", "astrocytoma", and "glioblastoma". The search only included keyword results that were present in the Final Diagnoses. "Glioma" returned 85 cases; "Astrocytoma" returned 67 cases; and "Glioblastoma" returned 215 cases. Following the exclusion of duplicate cases, those missing any of the four requisite MR imaging sequences, and cases that failed processing through our pipeline, our final cohort comprised 203 patients.

Data Acquisition

Radiology: MRI studies on our McKesson Radiology 12.2 Picture archiving and communication system (PACS) (Change Healthcare Radiology Solutions, Nashville, Tennessee, U.S) were exported. The image exportation process involved multiple personnels of varying ranks, including medical graduates, radiology residents, neuroradiology fellows, and neuroradiologists. Our team exported the four basic conventional MR sequences including T1, T1 with IV gadolinium-based contrast agent administration, T2, and Fluid Attenuated Inversion Recovery (FLAIR) into a HIPPA compliant MU secured research server.

For each patient, the images were thoroughly checked for including up to six post-treatment images as available. The post-treatment images were captured on different dates, though not all patients had the maximum number of follow-up images; some had as few as one post-treatment follow-up MRI. For patients with more frequent follow-up MRIs, the immediate post-operative scan, at least one time point of progression and another follow-up study. The MR images were comprehensively reviewed to exclude significantly motion degraded or suboptimal studies.

The majority of the studies were conducted using Siemens MRI machines 97.47%, n=579 with a smaller proportion performed on MRI machines from other vendors: GE (2.02%, n=12) and Philips (0.51%, n=3). Table 1 shows the distribution of studies across different Siemens MR machines. Regarding the magnetic field strength, 1.5T MRIs accounted for 48.14% (n=1,126), 3T MRIs accounted for 45.08% (n=318), and 3T MRIs accounted for 45.08% (n=261). Table 2 summarizes the MRI parameters of each MR sequence.

Our team made efforts to obtain 3D sequences whenever available. Scans were performed using 3D acquisition methods in 40.28% of cases (n=975) and 2D acquisition methods in 59.82% of cases (n=1,419). In cases where 3D images were not available, 2D images were utilized instead. Table 3 summarizes the counts and percentage of studies performed with 2D vs 3D acquisition across different MR sequences.

Clinical: Basic demographic data, clinical data points, and tumor pathology were obtained through review of the electronic medical record (EMR). Clinical data points included the date of diagnosis, date of first surgery or treatment, date and characterization of first and/or subsequent disease progression and/or recurrence, and date of any follow-up resections. Survival information included the date of death and, if that was unknown, the date of last known contact while alive. Disease progression and/or recurrence was characterized as imaging only, clinical only, or both based on information obtained through review of each patient’s clinical notes, brain imaging, and clinical impression as documented by the primary care team. Brief summaries of the reasoning behind each characterization were also included. Patients with no further clinical contact beyond their primary treatment were documented as “lost to follow-up.” Pathological information was obtained through review of the initial pathology note and any subsequent addenda for each tumor sample and included final tumor diagnosis, grade, and any identified genetic mutations. This information was then compiled into a spreadsheet for analysis.

Data Analysis

The image data underwent preprocessing using the Federated Tumor Segmentation (FeTS) tool. The pipeline began with converting DICOM files to the Neuroimaging Informatics Technology Initiative (NIfTI) format, ensuring the removal of any remaining PHI not eliminated by the anonymization/de-identification tool. The converted NIfTI images were then resampled to an isotropic 1mm³ resolution and co-registered to the standard anatomical human brain atlas, SRI24. A deep learning brain extraction method was applied to strip the skull and extracranial tissues, thereby mitigating any potential facial reconstruction or recognition risks.

The preprocessed images were segmented using a deep network based on nnU-Net, resulting in four distinct labels that correspond to different components of each tumor:

• Label 1: Non-enhancing Tumor Core (NETC). This label identifies non-enhancing components within the tumor, such as cystic, necrotic, or hemorrhagic portions.
• Label 2: Surrounding Non-enhancing FLAIR Hyperintensity (SNFH). This label represents both non-enhancing infiltrative tumor components and peritumoral vasogenic edema.
• Label 3: Enhancing Tissue (ET). This label highlights the viable nodular-enhancing components of the tumor.
• Label 4: Resection Cavity (RC). This label covers post-surgical changes, including recent changes like blood products and air foci, as well as chronic changes with materials isointense to CSF signal.

A spreadsheet is also provided that includes tumor volumes and signal intensity of different tumor components across various MR sequences.

Usage Notes

Each scan was manually exported using the built-in McKesson DICOM export tool into separate folders labeled as post-treatment 1, post-treatment 2, etc. In a subsequent step, a subset of the data was selected to contribute for the development of FeTS 2 toolbox. Consequently, the naming convention was updated to replace "post-treatment" with "timepoint" (e.g., post-treatment 1 became timepoint 1) to adhere to the instructions of the FeTS development team. Each sequence was saved in its own folder within these categories to a HIPPA compliant and secured server within the University of Missouri network. Exportation was conducted in DICOM format, maintaining the original image compression settings to preserve quality. To ensure patient privacy and HIPPA compliance, all images were anonymized and all protected health information (PHI) e.g. patient name, MRN, accession number, etc. were deleted from the metadata DICOM headers.

The folders are labeled in the following structure:

• Main folder: PatientID_XXXX
• Subfolders: Timepoint_X, Timepoint_X
• Each time point folder has the NIfTI images associated with the respective timepoints.

Clinical and genetic features of WHO grade II gliomas (n = 417).

Preoperative classification of brain tumors is critical to developing personalized treatment plans, however existing classification methods rely on manual intervention and often have problems with efficiency and accuracy, which may lead to misdiagnosis or delayed diagnosis in clinical practice and affect the therapeutic effect. We propose a fully automated approach to brain tumor magnetic resonance imaging (MRI) classification, consisted by a feature extractor based on the improved U-Net and a classifier based on convolutional recurrent neural network (CRNN). The encoder of the feature extractor based on dense block, is used to enhance feature propagation and reduce the number of parameters. The decoder uses residual block to reduce the weight of some features for improving the effect of MRI spatial sequence reconstruction, and avoid gradient disappearance. Skip connections between the encoder and the decoder effectively merge low-level features and high-level features. The extract feature sequence is input into the CRNN-based classifier for final classification. We assessed the performance of our method for grading glioma, glioma isocitrate dehydrogenase1 (IDH1) mutation status classification and pituitary tumor texture classification on two datasets, glioma or pituitary tumors collected in a local affiliated hospital and glioma imaging data from TCIA. Compared with commonly models and new models, our model achieves higher accuracy, with an accuracy of 90.72%, classified glioma IDH1 mutation status with an accuracy of 94.35%, and classified pituitary tumor texture with an accuracy of 94.64%.

This collection contains 406 ROI masks in MATLAB format defining the low grade glioma (LGG) tumour region on T1-weighted (T1W), T2-weighted (T2W), T1-weighted post-contrast (T1CE) and T2-flair (T2F) MR images of 108 different patients from the TCGA-LGG collection. From this subset of 108 patients, 81 patients have ROI masks drawn for the four MRI sequences (T1W, T2W, T1CE and T2F), and 27 patients have ROI masks drawn for three or less of the four MRI sequences. The ROI masks were used to extract texture features in order to develop radiomic-based multivariable models for the prediction of isocitrate dehydrogenase 1 (IDH1) mutation, 1p/19q codeletion status, histological grade and tumour progression. Clinical data (188 patients in total from the TCGA-LGG collection, some incomplete depending on the clinical attribute), VASARI scores (188 patients in total from the TCGA-LGG collection, 178 complete) with feature keys, and source code used in this study are also available with this collection. Please contact Martin Vallières (mart.vallieres@gmail.com) of the Medical Physics Unit of McGill University for any scientific inquiries about this dataset.

ObjectivesTo assess the diagnostic accuracy of machine learning (ML)-based radiomics for predicting isocitrate dehydrogenase (IDH) mutations in patients with glioma.MethodsA systematic search of PubMed, Web of Science, Embase, and the Cochrane Library from inception to 1 September 2023, was conducted to collect all articles investigating the diagnostic performance of ML for the prediction of IDH mutations in gliomas. Two reviewers independently screened all papers for eligibility. Methodological quality and risk of bias were assessed using the METhodological RadiomICs Score and Quality Assessment of Diagnostic Accuracy Studies-2, respectively. The pooled sensitivity, specificity, and 95% confidence intervals were calculated, and the area under the receiver operating characteristic curve (AUC) was obtained.ResultsIn total, 14 original articles assessing 1740 patients with gliomas were included. The AUC of ML for predicting IDH mutation was 0.90 (0.87–0.92). The pooled sensitivity, specificity, and diagnostic odds ratio were 0.83 (0.71–0.90), 0.84 (0.74–0.90), and 25 (12,50) respectively. In subgroup analyses, modeling methods, glioma grade, and the combination of magnetic resonance imaging and clinical features affected the diagnostic performance in predicting IDH mutations in gliomas.ConclusionML-based radiomics demonstrated excellent diagnostic performance in predicting IDH mutations in gliomas. Factors influencing the diagnosis included the modeling methods employed, glioma grade, and whether the model incorporated clinical features.Systematic review registrationhttps://www.crd.york.ac.uk/PROSPERO/#myprospero, PROSPERO registry (CRD 42023395444).

Aims and methods: astroblastoma is a rare glial brain tumor with singular morphology. Recurrent MN1-BEND2 fusions have been recently identified in most of pediatric cases. Adolescent and adult cases, however, remain molecularly poorly defined. Here, we performed clinical and molecular characterization of a retrospective cohort of 14 adult and one adolescent gliomas with astroblastic features. Results: strikingly, we found MN1 fusions a rare event in this age group (1/15). Using methylation profiling and targeted sequencing, most cases were reclassified as either pleomorphic xanthoastrocytomas (PXA) or high-grade glioma (HGG). PXA-like ABM show BRAF mutation (6/7 with V600E mutation and 1/7 with G466E mutation) and CD34 expression. Conversely, HGG-like ABM harbored specific mutations of diffuse midline glioma (2/5) or glioblastoma (3/5). These latter patients showed an unfavorable clinical course with significantly shorter overall survival (p = 0.027). MAPK pathway alterations (including FGFR fusion, BRAF and NF1 mutations) were present in 10 of 15 patients and overrepresented in the HGG group (3/5) compared to previously reported prevalence of these alterations in GBM and diffuse midline glioma. Conclusion: We suggest that astroblastoma comprises a variety of molecularly sharply defined entities. Adults’ astroblastomas harboring molecular features of PXA and HGG should be reclassified. CNS high-grade neuroepithelial tumors with MN1 alterations appears to be a truly pediatric entity and is uncommon in adult cases with a histology of astroblastoma. Astroblastic morphology in adults should thus prompt thorough molecular investigation aiming at a clear histomolecular diagnosis and identifying actionable drug targets, especially in MAPK pathway.

Tumor mutation burden (TMB) is a useful biomarker for predicting the prognosis and efficacy of immune checkpoint inhibitor (ICIs). In this study, we aimed to explore the prognostic value of TMB and TMB-related PRLHR immune genes as prognostic markers in patients with gliomas. We downloaded MAF files, RNA-seq data, and clinical information from the Cancer Genome Atlas (TCGA) database. The TMB of glioma was calculated and its correlation with clinical features and prognosis was analyzed. We found that TMB was statistically correlated with the grade and age of patients with gliomas. Kaplan-Meier curve analysis showed that low TMB was associated with better clinical outcome in patients with gliomas. Additionally, a predictive model based on five HUB genes (FABP5, VEGFA, SAA1, ADM, and PRLHR) was constructed to predict OS in patients with gliomas. Receiver operating characteristic curve analysis shows that the model is reliable in predicting the risk of survival and prognosis. Immune microenvironment analysis revealed a correlation between TMB and infiltrating immune cells. The clinical-related immune gene, PRLHR, can be used as an independent prognostic factor for patients with brain glioma, and it is negatively correlated with the grade of glioma and age of patients with glioma. We found that the higher the tumor grade and the older the age, the lower the PRLHR expression, which was verified by CGGA database and independent experimental data. These results suggest that PRLHR may be a tumor suppressor for the progression of glioma and might provide a new therapeutic target for the treatment and improvement of survival rate in patients with glioma.

Forest plot data for glioma IDH1 mutation status classification.

Background: Molecular characteristics can be good indicators of tumor prognosis and have been introduced into the classification of gliomas. The prognosis of patients with newly classified lower-grade gliomas (LGGs, including grade 2 and grade 3 gliomas) is highly heterogeneous, and new molecular markers are urgently needed.Methods: Autophagy related genes (ATGs) were obtained from Human Autophagy Database (HADb). From the Cancer Genome Atlas (TCGA) and the Chinese Glioma Genome Atlas (CGGA), gene expression profiles including ATG expression information and patient clinical data were downloaded. Cox regression analysis, receiver operating characteristic (ROC) analysis, Kaplan–Meier analysis, random survival forest algorithm (RSFVH) and stratification analysis were performed.Results: Through univariate Cox regression analysis, we found a total of 127 ATGs associated with the prognosis of LGG patients from TCGA dataset and a total of 131 survival-related ATGs from CGGA dataset. Using TCGA dataset as the training group (n = 524), we constructed a five-ATG signature (including BAG1, BID, MAP1LC3C, NRG3, PTK6), which could divide LGG patients into two risk groups with significantly different overall survival (Log Rank P < 0.001). Then we confirmed in the independent CGGA dataset that the five-ATG signature had the ability to predict prognosis (n = 431, Log Rank P < 0.001). We further discovered that the predictive ability of the five-ATG signature was better than the existing clinical indicators and IDH mutation status. In addition, the five-ATG signature could further classify patients after receiving radiotherapy or chemotherapy into groups with different prognosis.Conclusions: We identified a five-ATG signature that could be a reliable prognostic marker and might be therapeutic targets for autophagy therapy for LGG patients.

Background: Diffuse lower-grade gliomas (LGGs) are infiltrative and heterogeneous neoplasms. Gene signature including multiple protein-coding genes (PCGs) is widely used as a tumor marker. This study aimed to construct a multi-PCG signature to predict survival for LGG patients.Methods: LGG data including PCG expression profiles and clinical information were downloaded from The Cancer Genome Atlas (TCGA) and the Chinese Glioma Genome Atlas (CGGA). Survival analysis, receiver operating characteristic (ROC) analysis, and random survival forest algorithm (RSFVH) were used to identify the prognostic PCG signature.Results: From the training (n = 524) and test (n = 431) datasets, a five-PCG signature which can classify LGG patients into low- or high-risk group with a significantly different overall survival (log rank P < 0.001) was screened out and validated. In terms of prognosis predictive performance, the five-PCG signature is stronger than other clinical variables and IDH mutation status. Moreover, the five-PCG signature could further divide radiotherapy patients into two different risk groups. GO and KEGG analysis found that PCGs in the prognostic five-PCG signature were mainly enriched in cell cycle, apoptosis, DNA replication pathways.Conclusions: The new five-PCG signature is a reliable prognostic marker for LGG patients and has a good prospect in clinical application.

Glioma is a highly common pathological brain tumor. Misfolded protein response, which is strongly associated with the growth of cancerous tumors, is mediated by the gene, endoplasmic reticulum degradation-enhancing alpha-mannosidase-like protein 2. However, this gene has not been linked to glioma. To assess the same, we used The Cancer Genome Atlas, Chinese Glioma Genome Atlas, and Genotype-Tissue Expression datasets. The gene was overexpressed in gliomas. This overexpression was linked to unfavorable clinical characteristics, such as the World Health Organization grade, isocitrate dehydrogenase mutation, and the combined loss of the short arm chromosome 1 and the long arm of chromosome 19. Quantitative polymerase chain reaction experiments and immunohistochemistry on clinical samples from our institution verified the gene’s expression and clinical importance. The Human Protein Atlas website verified the messenger ribonucleic acid expression of the gene in glioma cell lines, and immunohistochemistry verified the presence of its protein. A previous survival study indicated that its high expression is substantially related to a bad prognosis. It was identified as an independent predictor of primary glioma prognosis using multivariate Cox regression analysis. To forecast individual survival, we created a nomogram based on this (concordance-index = 0.847). Additionally, functional annotation demonstrated its major role in the control of the extracellular matrix and immune system. The scratch assay and transwell migration assay confirmed the decreased invasive ability of U251 glioma cells with the gene knockdown. Its increased expression was found to be related to the extent of macrophage infiltration using the CIBERSORT, ESTIMATE, Single-sample Gene Set Enrichment Analysis, and Tumor Immune Single-Cell Hub (TISCH) algorithms. The Tumor Immune Dysfunction and Exclusion algorithm revealed that the gene can accurately predict the response of immunotherapy (area under the receiver operating characteristic curve = 0.857). Further, isocitrate dehydrogenase 1 mutation is typically more frequent when the gene expression is high. Finally, five medicines targeting this gene were discovered utilizing the molecular docking program and drug sensitivity analysis of the RNAactDrug website. Low expression of the gene inhibited glioma cell invasion. Therefore, the gene is helpful for the diagnosis, prognosis, and case-specific immunotherapy of glioma.

Cuproptosis is the most recently discovered type of regulated cell death and is mediated by copper ions. Studies show that cuproptosis plays a significant role in cancer development and progression. Lower-grade gliomas (LGGs) are slow-growing brain tumors. The majority of LGGs progress to high-grade glioma, which makes it difficult to predict the prognosis. However, the prognostic value of cuproptosis-related genes (CRGs) in LGG needs to be further explored. mRNA expression profiles and clinical data of LGG patients were collected from public sources for this study. Univariate Cox regression analysis and the least absolute shrinkage and selection operator (LASSO) Cox regression model were used to build a multigene signature that could divide patients into different risk groups. The differences in clinical pathological characteristics, immune infiltration characteristics, and mutation status were evaluated in risk subgroups. In addition, drug sensitivity and immune checkpoint scores were estimated in risk subgroups to provide LGG patients with precision medication. We found that all CRGs were differentially expressed in LGG and normal tissues. Patients were divided into high- and low-risk groups based on the risk score of the CRG signature. Patients in the high-risk group had a considerably lower overall survival rate than those in the low-risk group. According to functional analysis, pathways related to the immune system were enriched, and the immune state differed across the two risk groups. Immune characteristic analysis showed that the immune cell proportion and immune scores were different in the different groups. High-risk group was characterized by low sensitivity to chemotherapy but high sensitivity to immune checkpoint inhibitors. The current study revealed that the novel CRG signature was related to the prognosis, clinicopathological features, immune characteristics, and treatment perference of LGG.

Preoperative classification of brain tumors is critical to developing personalized treatment plans, however existing classification methods rely on manual intervention and often have problems with efficiency and accuracy, which may lead to misdiagnosis or delayed diagnosis in clinical practice and affect the therapeutic effect. We propose a fully automated approach to brain tumor magnetic resonance imaging (MRI) classification, consisted by a feature extractor based on the improved U-Net and a classifier based on convolutional recurrent neural network (CRNN). The encoder of the feature extractor based on dense block, is used to enhance feature propagation and reduce the number of parameters. The decoder uses residual block to reduce the weight of some features for improving the effect of MRI spatial sequence reconstruction, and avoid gradient disappearance. Skip connections between the encoder and the decoder effectively merge low-level features and high-level features. The extract feature sequence is input into the CRNN-based classifier for final classification. We assessed the performance of our method for grading glioma, glioma isocitrate dehydrogenase1 (IDH1) mutation status classification and pituitary tumor texture classification on two datasets, glioma or pituitary tumors collected in a local affiliated hospital and glioma imaging data from TCIA. Compared with commonly models and new models, our model achieves higher accuracy, with an accuracy of 90.72%, classified glioma IDH1 mutation status with an accuracy of 94.35%, and classified pituitary tumor texture with an accuracy of 94.64%.

BackgroundThe 2021 World Health Organization (WHO) Central Nervous System (CNS) Tumor Classification has suggested that isocitrate dehydrogenase wildtype (IDH-wt) WHO grade-2/3 astrocytomas with molecular features of glioblastoma should be designated as “Glioblastoma, IDH-wildtype, WHO grade-4.” This study analyzed the metabolic correlates of progression free and overall survival in “Glioblastoma, IDH-wildtype, WHO grade-4” patients using short echo time single voxel 1H-MRS.MethodsFifty-seven adult patients with hemispheric glioma fulfilling the 2021 WHO CNS Tumor Classification criteria for “Glioblastoma, IDH-wildtype, WHO grade-4” at presurgery time point were included. All patients were IDH1/2-wt and TERTp-mut. 1H-MRS was performed on a 3 T MR scanner and post-processed using LCModel. A Mann–Whitney U test was used to assess the metabolic differences between gliomas with or without contrast enhancement and necrosis. Cox regression analysis was used to assess the effects of age, extent of resection, presence of contrast enhancement and necrosis, and metabolic intensities on progression-free survival (PFS) and overall survival (OS). Machine learning algorithms were employed to discern possible metabolic patterns attributable to higher PFS or OS.ResultsContrast enhancement (p = 0.015), necrosis (p = 0.012); and higher levels of Glu/tCr (p = 0.007), GSH/tCr (p = 0.019), tCho/tCr (p = 0.032), and Glx/tCr (p = 0.010) were significantly associated with shorter PFS. Additionally, necrosis (p = 0.049), higher Glu/tCr (p = 0.039), and Glx/tCr (p = 0.047) were significantly associated with worse OS. Machine learning models differentiated the patients having longer than 12 months OS with 81.71% accuracy and the patients having longer than 6 months PFS with 77.41% accuracy.ConclusionGlx and GSH have been identified as important metabolic correlates of patient survival among “IDH-wt, TERT-mut diffuse gliomas” using single-voxel 1H-MRS on a clinical 3 T MRI scanner.

Biomarkers for classifying and grading gliomas have been extensively explored, whereas populations in public databases were mostly Western/European. Based on public databases cannot accurately represent Chinese population. To identify molecular characteristics associated with clinical outcomes of lower-grade glioma (LGG) and glioblastoma (GBM) in the Chinese population, we performed whole-exome sequencing (WES) in 16 LGG and 35 GBM tumor tissues. TP53 (36/51), TERT (31/51), ATRX (16/51), EFGLAM (14/51), and IDH1 (13/51) were the most common genes harboring mutations. IDH1 mutation (c.G395A; p.R132H) was significantly enriched in LGG, whereas PCDHGA10 mutation (c.A265G; p.I89V) in GBM. IDH1-wildtype and PCDHGA10 mutation were significantly related to poor prognosis. IDH1 is an important biomarker in gliomas, whereas PCDHGA10 mutation has not been reported to correlate with gliomas. Different copy number variations (CNVs) and oncogenic signaling pathways were identified between LGG and GBM. Differential genomic landscapes between LGG and GBM were revealed in the Chinese population, and PCDHGA10, for the first time, was identified as the prognostic factor of gliomas. Our results might provide a basis for molecular classification and identification of diagnostic biomarkers and even potential therapeutic targets for gliomas.

Glioblastoma is the most common and lethal brain cancer globally. Clinically, this cancer has heterogenous molecular and clinical characteristics. Studies have shown that UBE2S is highly expressed in many cancers. But its expression profile in glioma, and the correlation with clinical outcomes is unknown. RNA sequencing data of glioma samples was downloaded from the Chinese Glioma Genome Atlas and The Cancer Genome Atlas. A total of 114 cases of glioma tissue samples (WHO grades II-IV) were used to conduct protein expression assays. The molecular and biological characteristics of UBE2S, and its prognostic value were analyzed. The results showed that high UBE2S expression was associated with a higher grade of glioma and PTEN mutations. In addition, UBE2S affected the degree of malignancy of glioma and the development of chemo-radiotherapy resistance. It was also found to be an independent predictor of worse survival of LGG patients. Furthermore, we identified five UBE2S ubiquitination sites and found that UBE2S was associated with Akt phosphorylation in malignant glioblastoma. The results also revealed that UBE2S expression was negatively correlated with 1p19q loss and IDH1 mutation; positively correlated with epidermal growth factor receptor amplification and PTEN mutation. This study demonstrates that UBE2S expression strongly correlates with glioma malignancy and resistance to chemo-radiotherapy. It is also a crucial biomarker of poor prognosis.

Background: As reflected in the WHO classification of glioma since 2020, genomic information has been an important criterion in addition to histology for glioma classification. There is a significant intergrade difference as well as intragrade difference of survival probability among glioma patients. Except the molecular criteria used in the WHO classification, few studies have explored other genomic factors that may be underlying these survival differences, especially in Chinese populations. Here, we used integrative genomic approaches to characterize a Chinese glioma cohort to search for potential prognostic biomarkers.Methods: We recruited 46 Chinese patients with primary malignant glioma. All the patients were analyzed with whole-exome sequencing (WES) and 27 of them were analyzed with RNA-seq. We compared the molecular features between patients in different WHO grades. We classified the glioblastoma (GBM) patients into two groups (good vs poor survival) using six-month progression-free survival (PFS6) status and compared the genomic profiles between the two groups.Results: We found grade II and grade III patients cluster together (LGG) and they are different from GBM in unsupervised clustering analysis with RNA-seq data. Gene set enrichment analysis (GSEA) comparing GBM and the LGG group suggested that GBM had upregulation of multiple pathways related to genome integrity and immune cell infiltration. Further comparison of somatic mutations between the two groups revealed TOPAZ1 as a novel mutation associated with GBM and prevalence of CNV in multiple genes in GBM. Comparison between PFS6 good and poor GBM patients revealed six genes (TRIML2, ROCK1, PKD1, OBSCN, HECTD4, and ADCY7) were significantly mutated and two genes (NTRK1 and B2M) had more CNV alterations in the poor prognosis group.Conclusion: Taken together, our molecular data revealed that GBM patient showed distinct characteristics related to individual gene, chromosome integrity, and infiltrating immune cells compared to LGG (grade II/III) patients. We also identified few novel genes with SNV or CNV, which might be the potential markers for clinical outcome of GBM.

BackgroundGliomas account for 75% of all primary malignant brain tumors in adults and result in high mortality. Accumulated evidence has declared the minichromosome maintenance protein complex (MCM) gene family plays a critical role in modulating the cell cycle and DNA replication stress. However, the biological function and clinic characterization of nine MCM members in low-grade glioma are not yet clarified.MethodsIn this study, we utilized diverse public databases, including The Cancer Genome Atlas (TCGA), Chinese Glioma Genome Atlas (CGGA), Rembrandt, Human Protein Atlas (HPA), Linkedomics, cbioportal, Tumor and Immune System Interaction Database (TISIDB), single-sample GSEA (ssGSEA), Tumor Immune Estimation Resource (TIMER), Genomics of Drug Sensitivity in Cancer (GDSC) and Cancer Therapeutics Response Portal databases to explore the mRNA and protein expression profiles, gene mutation, clinical features, diagnosis, prognosis, signaling pathway, tumor mutational burden (TMB), immune subtype, immune cell infiltration, immune modulator and drug sensitivity of nine MCMs. Afterward, qRT-PCR was utilized to detect the expression of the MCM family in glioblastoma multiforme (GBM) cell lines. The one-, three-, or five-year survival rate was predicted by utilizing a nomogram established by cox proportional hazard regression.ResultsIn this study, we found that nine MCMs were consistently up-regulated in glioma tissues and glioma cell lines. Elevated nine MCMs expressions were significantly correlated with a higher tumor stage, isocitrate dehydrogenase (IDH) mutates, 1p/19q codeletion, histological type, and primary therapy outcome. Survival analyses showed that higher expression of MCM2-MCM8 (minichromosome maintenance protein2-8) and MCM10 (minichromosome maintenance protein 10) were linked with poor overall survival (OS) and progression-free survival (PFS) in glioma patients. On the other hand, up-regulated MCM2-MCM8 and MCM10 were significantly associated with shorter disease-specific survival (DSS) in glioma patients. Univariate and multivariate analyses revealed that MCM2 (minichromosome maintenance protein2), MCM4 (minichromosome maintenance protein 4), MCM6 (minichromosome maintenance protein 6), MCM7 (minichromosome maintenance protein 7) expression and tumor grade, 1p/19q codeletion, age, and primary therapy outcome were independent factors correlated with the clinical outcome of glioma patients. More importantly, a prognostic MCMs model constructed using the above five prognostic genes could predict the overall survival of glioma patients with medium-to-high accuracy. Furthermore, functional enrichment analysis indicated that MCMs principal participated in regulating cell cycle and DNA replication. DNA copy number variation (CNV) and DNA methylation significantly affect the expression of MCMs. Finally, we uncover that MCMs expression is highly correlated with immune cell infiltration, immune modulator, TMB, and drug sensitivity.ConclusionsIn summary, this finding confirmed that MCM4 is a potential target of precision therapy for patients with glioma.