Subclassification Of Glioblastoma Research Articles

An Overview of Mathematical Models for RNA Sequence-based Glioblastoma Subclassification

Molecular marker-based glioblastoma (GBM) subclassification is emerging as a key factor in personalized GBM treatment planning. Multiple genetic alterations, including methylation status and mutations, have been proposed in GBM subclassification. RNA-Sequence (RNA-Seq)-based molecular profiling of GBM is widely implemented and readily quantifiable. Machine learning (ML) algorithms have been reported as an applicable method that can consistently subgroup GBM. In this study, we systematically studied the applicability of the commonly used ML algorithms based on The Cancer Genome Atlas Glioblastoma Multiforme (TCGA-GBM) dataset and cross-validated in the Chinese Glioma Genome Atlas (CGGA) dataset. ML algorithms studied include Binomial and multinomial Logistic Regression, Linear discriminant analysis, Decision trees, K-Nearest Neighbors, Gaussian Naive Bayes, Support Vector Machines, Gradient Boosting, Voting Ensemble, Multi-Layer Perceptron. RNA-Seq data of 44 biomarkers were passed through the algorithms for performance evaluation. We found ML algorithms Support Vector Machines, Multi-Layer Perceptron s, and Voting Ensemble are best equipped in assigning GBM to correct molecular subgroups of GBM without histological studies.

Rapid Processing and Drug Evaluation in Glioblastoma Patient-Derived Organoid Models with 4D Bioprinted Arrays

SummaryGlioblastoma is the most common and deadly primary brain malignancy. Despite advances in precision medicine oncology (PMO) allowing the identification of molecular vulnerabilities in glioblastoma, treatment options remain limited, and molecular assays guided by genomic and expression profiling to inform patient enrollment in life-saving trials are lacking. Here, we generate four-dimensional (4D) cell-culture arrays for rapid assessment of drug responses in glioblastoma patient-derived models. The arrays are 3D printed with thermo-responsive shape memory polymer (SMP). Upon heating, the SMP arrays self-transform in time from 3D cell-culture inserts into histological cassettes. We assess the utility of these arrays with glioblastoma cells, gliospheres, and patient derived organoid-like (PDO) models and demonstrate their use with glioblastoma PDOs for assessing drug sensitivity, on-target activity, and synergy in drug combinations. When including genomic and drug testing assays, this platform is poised to offer rapid functional drug assessments for future selection of therapies in PMO.

A Potential Prognostic Gene Signature for Predicting Survival for Glioblastoma Patients.

Objective This study aimed to screen prognostic gene signature of glioblastoma (GBM) to construct prognostic model. Methods Based on the GBM information in the Cancer Genome Atlas (TCGA, training set), prognostic genes (Set X) were screened by Cox regression. Then, the optimized prognostic gene signature (Set Y) was further screened by the Cox-Proportional Hazards (Cox-PH). Next, two prognostic models were constructed: model A was based on the Set Y; model B was based on part of the Set X. The samples were divided into low- and high-risk groups according to the median prognosis index (PI). GBM datasets in Gene Expression Ominous (GEO, GSE13041) and Chinese Glioma Genome Atlas (CGGA) were used as the testing datasets to confirm the prognostic models constructed based on TCGA. Results We identified that the prognostic 14-gene signature was significantly associated with the overall survival (OS) in the TCGA. In model A, patients in high- and low-risk groups showed the significantly different OS (P = 7.47 × 10−9, area under curve (AUC) 0.995) and the prognostic ability were also confirmed in testing sets (P=0.0098 and 0.037). The model B in training set was significant but failed in testing sets. Conclusion The prognostic model which was constructed based on the prognostic 14-gene signature presented a high predictive ability for GBM. The 14-gene signature may have clinical implications in the subclassification of GBM.

Abstract 4148: Different immune cell responses are associated with glioblastoma subclassification and typical genetic alterations

Abstract Interactions between various components in the tumor microenvironment and dysregulated immune responses are thought to play important roles in cancer development. To better understand the role of immune cells in tumor pathogenesis and destruction, we computationally model the microenvironment of an aggressive brain tumor glioblastoma multiforme (GBM). We downloaded GBM patient RNA-seq data from the Cancer Genome Atlas (TCGA). Using cluster analysis, we identified 16 clusters, each containing 10 to 933 genes that show a statistical enrichment of immune response related gene ontology terms. Utilizing a panel of RNA-seq data from normal cell types, we constructed global and cluster specific regression models to characterize the expression profiles of GBM samples in the clusters of interest as linear combinations of normal cell and reference GBM expression profiles. Simulated data was used to validate that the regression model coefficients accurately reflect the contributions of normal cell types to the expression profiles of tumor samples. Based on the regression analysis, we were able to uncover high variability in the composition of microenvironment across the TCGA cohort, suggesting diverse immune responses in tumors. The results from global regression analysis were then associated with common genetic alterations in GBM and with GBM subclassification. Especially the estimated macrophage, granulocyte, and CD8+ T lymphocyte proportions show significant differences between different GBM subgroups. Accumulation of immune cells was increased in the mesenchymal and neural subtype compared to other subtypes. Furthermore, estimated immune cell proportions were associated with alterations in EGFR and NF1. Taken together, our analysis provided a characterization of the immunomicroenvironment in GBM and linked immune cell responses to typical GBM alterations and GBM subtypes. More detailed characterization of diverse immune responses will facilitate patient stratification and might provide tools for personalized immunotherapy in the future. Citation Format: Suvi Luoto, Juha Kesseli, Matti Nykter, Kirsi J. Granberg. Different immune cell responses are associated with glioblastoma subclassification and typical genetic alterations. [abstract]. In: Proceedings of the 107th Annual Meeting of the American Association for Cancer Research; 2016 Apr 16-20; New Orleans, LA. Philadelphia (PA): AACR; Cancer Res 2016;76(14 Suppl):Abstract nr 4148.

ATPS-16IDENTIFICATION OF CLINICALLY RELEVANT DRUGS FOR GLIOBLASTOMA WITH SPECIFIC MOLECULAR SIGNATURE USING CHEMICAL BIOLOGY FINGERPRINTING

Translating molecular subtyping of glioblastoma (GBM) into therapeutically actionable guidance remains an unfulfilled opportunity. A central challenge is to discover therapeutic targets matched to agents (drugs or tool compounds) that are predictably present in subclasses of GBM. Transcriptional patterns in GBM cases in TCGA can identify 12 distinct molecular contexts (mC), where, across a subset of genes and tumors, conditional dependencies of gene expression (signal & echo relationships) consistently emerge. We deployed a panel of 64 patient-derived xenografts from GBM patients whose gene expression profiles allow mapping against the same 12 molecular contexts identified from TCGA cases. These clinically-relevant, preclinical models afford bioinformatics mining for actionable targets as well as chemical library screening for activity against short-term cultures derived from the PDX models. Utilizing a technique we term Chemical Biology Fingerprinting (CBF), we interrogated a series of GBM PDX models using a small chemical library (650 compounds) of clinically relevant anti-cancer agents to uncover context-specific chemovulnerabilities. Preliminary data demonstrated that, mC14 GBM (GBM59, SF7300, GBM116), characterized by mt-P53 and transcriptional similarity to GBM proneural subtype, show distinct vulnerability to Arsenic Trioxide (ATO) as compared to mC4 GBM (GBM91, GBM102), characterized by CHEK2 and NF1 mutations and transcriptional patterns similar to GBM mesenchymal subtype. To clinically-validate an ATO vulnerability signature, we acquired 22 treatment naive archival patient samples, who were part of Phase I/II clinical trial to study efficacy of ATO and Temozolomide (TMZ) in combination with radiation in treatment of high grade gliomas (NCT00275067) and which exhibited varied survival with ATO treatment (91 days to >1000 days), and determined their molecular context classification using whole transcriptome sequencing (RNAseq). In summary, we demonstrate a subclassification of GBM into novel contexts and we also show that these contexts are differentially sensitive to clinically relevant drugs. Supported by NIH U01CA168397.

NIMG-11RADIOMIC SUBCLASSIFICATION OF GLIOBLASTOMA

BACKGROUND: To identify clinically relevant glioblastoma (GBM) sub-classification/subtypes based on radiomic analysis. Integrated genomic analysis has already identified genomic subtypes of GBM as characterized by specific genomic aberrations and mutations. In our study, we seek to demonstrate a GBM sub-classification based on radiomic texture analysis that is both predictive and prognostic. METHODS: We retrospectively identified 80 GBM treatment naive patients from the Cancer Genome Atlas with corresponding MRI data in the Cancer Imaging Archive. We applied texture analysis (528 features) to each tumor. We extracted texture features and performed feature selection. Univariate and multivariate analysis were also performed. Clustering identified radiomic subclasses. RESULTS: Multiple texture features were associated with GBM and we were able to classify GBM into 4 subtypes. However, subclassification into 3 subclasses were associated with differences in patient progression free and overall survival. CONCLUSIONS: GBM subtypes based on radiomic texture signatures were identified that were both predictive and prognostic. These findings are important as this can help determine patient outcome based on imaging (obtained at the time of diagnosis before treatment); this can also help determine response to therapy at the pre-treatment stage to stratify patients and serve as imaging endpoints for clinical trials.

CS-25 * MOLECULAR SUBCLASSIFICATION OF GLIOBLASTOMA BASED ON THE ABSOLUTE QUANTITATIVE PROTEOMICS

Glioblastoma (GBM) has been classified into molecular subtypes with the advent of large-scale gene expression studies. The purpose of this study is to attempt to classify GBM subtypes based on the molecular profile determined by absolute quantitative proteomics and to investigate its clinical significance. The absolute protein expression levels of 13 species of receptor tyrosine kinases (RTKs) in 33 adult GBM surgical specimens were measured by the mass spectrometer which was set up to run a multi-channel reaction monitoring. For each tumor, we ascertained information on patient age, gender, IDH1 mutation, methylation status of MGMT promoter, progression-free survival (PFS) and overall survival (OS). Eight RTK proteins (EGFR, ERBB2, PDGFRα, PDGFRβ, VEGFR1, CD33, CD37, c-kit) were successfully quantified, whereas 5 protein levels were below detection limit. Large individual differences were observed in protein expression levels of each RTK. The highest mean expression level was detected for EGFR. GBMs in our study harbored mutually exclusive dominant expression of three RTKs; EGFR, ERBB2 and PDGFRα. Therefore, we propose a subclassification of GBM with respect to the dominant expression of these proteins; EGFR (n = 20), ERBB2 (n = 7) and PDGFRα (n = 6). The clinical characteristics of each group (EGFR, ERBB2, PDGFRα) are as follows: average age, (61.9, 58.4, 56.7 years old), gender (male/female, 15/5, 4/3, 6/0), IDH1 (wild/mutant, 18/2, 6/1, 6/0), MGMT promoter (methylated/unmethylated, 8/12, 3/4, 4/2), the median progression-free survival (PFS) (7.6, 10.0, 8.6 months), the median overall survival (OS) (13.7, 21.1, 13.8 months). PFS and OS tended to be longer in ERBB2 group in comparison with EGFR and PDGFRα group. Molecular subclassification of GBM based on the absolute quantitative proteomics reflects the clinical behavior and might have clinical significance.

Associations between microRNA expression and mesenchymal marker gene expression in glioblastoma

The subclassification of glioblastoma (GBM) into clinically relevant subtypes using microRNA (miRNA)- and messenger RNA (mRNA)-based integrated analysis has been attempted. Because miRNAs regulate multiple gene-signaling pathways, understanding miRNA-mRNA interactions is a prerequisite for understanding glioma biology. However, such associations have not been thoroughly examined using high-throughput integrated analysis. To identify significant miRNA-mRNA correlations, we selected and quantified signature miRNAs and mRNAs in 82 gliomas (grade II: 14, III: 16, IV: 52) using real-time reverse-transcriptase polymerase chain reaction. Quantitative expression data were integrated into a single analysis platform that evaluated the expression relationship between miRNAs and mRNAs. The 21 miRNAs include miR-15b, -21, -34a, -105, -124a, -128a, -135b, -184, -196a-b, -200a-c, -203, -302a-d, -363, -367, and -504. In addition, we examined 23 genes, including proneural markers (DLL3, BCAN, and OLIG2), mesenchymal markers (YKL-40, CD44, and Vimentin), cancer stem cell-related markers, and receptor tyrosine kinase genes. Primary GBM was characterized exclusively by upregulation of mesenchymal markers, whereas secondary GBM was characterized by significant downregulation of mesenchymal markers, miR-21, and -34a, and by upregulation of proneural markers and miR-504. Statistical analysis showed that expression of miR-128a, -504, -124a, and -184 each negatively correlated with the expression of mesenchymal markers in GBM. Our functional analysis of miR-128a and -504 as inhibitors demonstrated that suppression of miR-128a and -504 increased the expression of mesenchymal markers in glioblastoma cell lines. Mesenchymal signaling in GBM may be negatively regulated by miR-128a and -504.