Glioblastoma Dataset Research Articles

Multivariate Analysis of Preoperative Magnetic Resonance Imaging Reveals Transcriptomic Classification of de novo Glioblastoma Patients.

Glioblastoma, the most frequent primary malignant brain neoplasm, is genetically diverse and classified into four transcriptomic subtypes, i. e., classical, mesenchymal, proneural, and neural. Currently, detection of transcriptomic subtype is based on ex vivo analysis of tissue that does not capture the spatial tumor heterogeneity. In view of accumulative evidence of in vivo imaging signatures summarizing molecular features of cancer, this study seeks robust non-invasive radiographic markers of transcriptomic classification of glioblastoma, based solely on routine clinically-acquired imaging sequences. A pre-operative retrospective cohort of 112 pathology-proven de novo glioblastoma patients, having multi-parametric MRI (T1, T1-Gd, T2, T2-FLAIR), collected from the Hospital of the University of Pennsylvania were included. Following tumor segmentation into distinct radiographic sub-regions, diverse imaging features were extracted and support vector machines were employed to multivariately integrate these features and derive an imaging signature of transcriptomic subtype. Extracted features included intensity distributions, volume, morphology, statistics, tumors' anatomical location, and texture descriptors for each tumor sub-region. The derived signature was evaluated against the transcriptomic subtype of surgically-resected tissue specimens, using a 5-fold cross-validation method and a receiver-operating-characteristics analysis. The proposed model was 71% accurate in distinguishing among the four transcriptomic subtypes. The accuracy (sensitivity/specificity) for distinguishing each subtype (classical, mesenchymal, proneural, neural) from the rest was equal to 88.4% (71.4/92.3), 75.9% (83.9/72.8), 82.1% (73.1/84.9), and 75.9% (79.4/74.4), respectively. The findings were also replicated in The Cancer Genomic Atlas glioblastoma dataset. The obtained imaging signature for the classical subtype was dominated by associations with features related to edge sharpness, whereas for the mesenchymal subtype had more pronounced presence of higher T2 and T2-FLAIR signal in edema, and higher volume of enhancing tumor and edema. The proneural and neural subtypes were characterized by the lower T1-Gd signal in enhancing tumor and higher T2-FLAIR signal in edema, respectively. Our results indicate that quantitative multivariate analysis of features extracted from clinically-acquired MRI may provide a radiographic biomarker of the transcriptomic profile of glioblastoma. Importantly our findings can be influential in surgical decision-making, treatment planning, and assessment of inoperable tumors.

Cell Lineage-Based Glioblastoma Stratification in Mice Reveals Functional Alignment with Patient Derived Models

Glioblastoma, the predominant adult malignant brain tumor, is subject to a limited number of driver mutations and has been computationally classified into subtypes whose functional relevance remains to be established. We have previously reported genetically engineered glioblastoma mouse models initiated by identical driver mutations in two distinct cells of origin: adult subventricular zone neural stem cells and oligodendrocyte lineage progenitors. These biologically and functionally separable tumors portray unique transcriptional profiles reflective of their respective lineage. Here, we identify corresponding transcriptional profiles among human glioblastoma data sets and describe patient derived xenografts with parallel features including species-conserved subtype-discriminating functional properties. Both mouse and human oligodendrocyte lineage associated glioblastoma subtypes exhibit functional ErbB3 receptor requirement and unique shared therapeutic sensitivities. We thus provide a biological lineage-based classification for glioblastoma. Significance: Glioblastoma has no effective treatment. Elucidation of biological and functional stratification may, as demonstrated here, reveal therapeutic sensibilities that are masked in the absence of subtype stratification.

Classification of glioma based on prognostic alternative splicing

BackgroundPreviously developed classifications of glioma have provided enormous advantages for the diagnosis and treatment of glioma. Although the role of alternative splicing (AS) in cancer, especially in glioma, has been validated, a comprehensive analysis of AS in glioma has not yet been conducted. In this study, we aimed at classifying glioma based on prognostic AS.MethodsUsing the TCGA glioblastoma (GBM) and low-grade glioma (LGG) datasets, we analyzed prognostic splicing events. Consensus clustering analysis was conducted to classified glioma samples and correlation analysis was conducted to characterize regulatory network of splicing factors and splicing events.ResultsWe analyzed prognostic splicing events and proposed novel splicing classifications across pan-glioma samples (labeled pST1–7) and across GBM samples (labeled ST1–3). Distinct splicing profiles between GBM and LGG were observed, and the primary discriminator for the pan-glioma splicing classification was tumor grade. Subtype-specific splicing events were identified; one example is AS of zinc finger proteins, which is involved in glioma prognosis. Furthermore, correlation analysis of splicing factors and splicing events identified SNRPB and CELF2 as hub splicing factors that upregulated and downregulated oncogenic AS, respectively.ConclusionA comprehensive analysis of AS in glioma was conducted in this study, shedding new light on glioma heterogeneity and providing new insights into glioma diagnosis and treatment.

Astrocytes enhance glioblastoma growth.

Glioblastoma (GBM) is a deadly disease with a need for deeper understanding and new therapeutic approaches. The microenvironment of glioblastoma has previously been shown to guide glioblastoma progression. In this study, astrocytes were investigated with regard to their effect on glioblastoma proliferation through correlative analyses of clinical samples and experimental in vitro and in vivo studies. Co-culture techniques were used to investigate the GBM growth enhancing potential of astrocytes. Cell sorting and RNA sequencing were used to generate a GBM-associated astrocyte signature and to investigate astrocyte-induced GBM genes. A NOD scid GBM mouse model was used for in vivo studies. A gene signature reflecting GBM-activated astrocytes was associated with poor prognosis in the TCGA GBM dataset. Two genes, periostin and serglycin, induced in GBM cells upon exposure to astrocytes were expressed at higher levels in cases with high "astrocyte signature score". Astrocytes were shown to enhance glioblastoma cell growth in cell lines and in a patient-derived culture, in a manner dependent on cell-cell contact and involving increased cell proliferation. Furthermore, co-injection of astrocytes with glioblastoma cells reduced survival in an orthotopic GBM model in NOD scid mice. In conclusion, this study suggests that astrocytes contribute to glioblastoma growth and implies this crosstalk as a candidate target for novel therapies.

BSCI-14. SYNTHETIC METASTATIC BRAIN DISEASE MRI IMAGES CREATED USING A GENERATIVE ADVERSARY NETWORK TO OVERCOME DEEP MACHINE LEARNING CHALLENGES IN HEALTHCARE

Deep Machine Learning (DML) in commercial applications such as recognizing animal species in photographs occurred through analyzing large volumes of public data. To achieve similar success in brain tumor imaging, additional factors must be addressed such as the need to follow strict regulatory protocols, work with limited datasets, and protect patient privacy. Generative adversary network (GAN) restricted to intracranial disease is one possibility to overcome these challenges and enable training on small annotated datasets to synthesize new samples. Large fabricated brain metastases (BM) training datasets derived from patient MRI using GAN models may enable DML of BM MRI studies. METHOD: We randomly selected 82 glioma patient imaging studies from the MICCAI BraTS 2018 Challenge1. All patients underwent contouring of GD-enhancing tumor (C+), peritumoral T2 (pT2), necrotic and non-enhancing tumor core (NCR/NET). Images are co-registered to the anatomical template and skull-stripped. Our network consists of a GAN and a discriminative network. The generative model works to synthesize images from labels. Labels comprise the normal brain mask as well as the contoured C+, pT2 and NCR/NET. Normal brain mask is extracted from threshold segmentation on T2-weighted image (T2WI). A discriminative network compares the difference between synthetic and real patient image in both pixel and perceptual difference. The generative model is trained to minimize the difference from the discriminative network. This method was refined in the glioblastoma dataset and applied to BM MRI. RESULTS: Figure 1. Synthetic BM MRI images derived from human brain MRI studies using the GAN model with four modalities (T2, T2 FLAIR, T1 contrasted image, and T1 non-contrasted Image). CONCLUSION: Training DML in BM disease using GAN MRI models may overcome limitations in applying DML to healthcare, namely volume of high-quality data and patient privacy. GAN based modeling for BM needs to be further refined and validated.

Impact of image preprocessing on the scanner dependence of multi-parametric MRI radiomic features and covariate shift in multi-institutional glioblastoma datasets

Recent advances in radiomics have enhanced the value of medical imaging in various aspects of clinical practice, but a crucial component that remains to be investigated further is the robustness of quantitative features to imaging variations and across multiple institutions. In the case of MRI, signal intensity values vary according to the acquisition parameters used, yet no consensus exists on which preprocessing techniques are favorable in reducing scanner-dependent variability of image-based features. Hence, the purpose of this study was to assess the impact of common image preprocessing methods on the scanner dependence of MRI radiomic features in multi-institutional glioblastoma multiforme (GBM) datasets. Two independent GBM cohorts were analyzed: 50 cases from the TCGA-GBM dataset and 111 cases acquired in our institution, and each case consisted of 3 MRI sequences viz. FLAIR, T1-weighted, and T1-weighted post-contrast. Five image preprocessing techniques were examined: 8-bit global rescaling, 8-bit local rescaling, bias field correction, histogram standardization, and isotropic resampling. A total of 420 features divided into eight categories representing texture, shape, edge, and intensity histogram were extracted. Two distinct imaging parameters were considered: scanner manufacturer and scanner magnetic field strength. Wilcoxon tests identified features robust to the considered acquisition parameters under the selected image preprocessing techniques. A machine learning-based strategy was implemented to measure the covariate shift between the analyzed datasets using features computed using the aforementioned preprocessing methods. Finally, radiomic scores (rad-scores) were constructed by identifying features relevant to patients’ overall survival after eliminating those impacted by scanner variability. These were then evaluated for their prognostic significance through Kaplan–Meier and Cox hazards regression analyses. Our results demonstrate that overall, histogram standardization contributes the most in reducing radiomic feature variability as it is the technique to reduce the covariate shift for three feature categories and successfully discriminate patients into groups of different survival risks.

Classification of Glioma Based on Prognostic Alternative Splicing

Background: Previously developed classifications of glioma have provided enormous advantages for the diagnosis and treatment of glioma. Although the role of alternative splicing (AS) in cancer, especially in glioma, has been validated, a comprehensive analysis of prognostic AS in glioma has not yet been conducted. Methods: TCGA glioblastoma (GBM) and low-grade glioma (LGG) datasets were downloaded from TCGA data portal and used to analyzed the prognostic splicing events. Consensus clustering analysis were used to classified glioma samples and correlation analysis was used to characterize regulatory network of splicing factors and splicing events. Results: We analyzed prognostic splicing events and proposed novel splicing classifications across pan-glioma samples (labeled pST1-7) and across GBM samples (labeled ST1-3). Distinct splicing profiles between GBM and LGG were observed, and the primary discriminator for the pan-glioma splicing classification was tumor grade. Subtype-specific splicing events were identified; one example is AS of zinc finger proteins, which is involved in glioma prognosis. Furthermore, correlation analysis of splicing factors and splicing events identified SNRPB and CELF2 as hub splicing factors that upregulated and downregulated oncogenic AS, respectively. Conclusion: A comprehensive analysis of prognostic AS in glioma was conducted in this study, shedding new light on glioma heterogeneity and providing new insights into glioma diagnosis and treatment. Funding: This study was supported by the National Natural Science Foundation of China (Grant nos. 81773290, 81472315, and 81872442), Guangdong Science and Technology Department (2017A030313497, 2016A040403053). Declaration of Interest: The authors report no competing interests. Ethical Approval: MIssing

Modelling cancer progression using Mutual Hazard Networks.

MotivationCancer progresses by accumulating genomic events, such as mutations and copy number alterations, whose chronological order is key to understanding the disease but difficult to observe. Instead, cancer progression models use co-occurrence patterns in cross-sectional data to infer epistatic interactions between events and thereby uncover their most likely order of occurrence. State-of-the-art progression models, however, are limited by mathematical tractability and only allow events to interact in directed acyclic graphs, to promote but not inhibit subsequent events, or to be mutually exclusive in distinct groups that cannot overlap.ResultsHere we propose Mutual Hazard Networks (MHN), a new Machine Learning algorithm to infer cyclic progression models from cross-sectional data. MHN model events by their spontaneous rate of fixation and by multiplicative effects they exert on the rates of successive events. MHN compared favourably to acyclic models in cross-validated model fit on four datasets tested. In application to the glioblastoma dataset from The Cancer Genome Atlas, MHN proposed a novel interaction in line with consecutive biopsies: IDH1 mutations are early events that promote subsequent fixation of TP53 mutations.Availability and implementationImplementation and data are available at https://github.com/RudiSchill/MHN.Supplementary information Supplementary data are available at Bioinformatics online.

Simultaneous Integration of Multi-omics Data Improves the Identification of Cancer Driver Modules.

The identification of molecular pathways driving cancer progression is a fundamental challenge in cancer research. Most approaches to address it are limited in the number of data types they employ and perform data integration in a sequential manner. Here, we describe ModulOmics, a method to de novo identify cancer driver pathways, or modules, by integrating protein-protein interactions, mutual exclusivity of mutations and copy number alterations, transcriptional coregulation, and RNA coexpression into a single probabilistic model. To efficiently search and score the large space of candidate modules, ModulOmics employs a two-step optimization procedure that combines integer linear programming with stochastic search. Applied across several cancer types, ModulOmics identifies highly functionally connected modules enriched with cancer driver genes, outperforming state-of-the-art methods and demonstrating the power of using multiple omics data types simultaneously. On breast cancer subtypes, ModulOmics proposes unexplored connections supported by an independent patient cohort and independent proteomic and phosphoproteomic datasets.

Cell-level somatic mutation detection from single-cell RNA sequencing.

MotivationBoth single-cell RNA sequencing (scRNA-seq) and DNA sequencing (scDNA-seq) have been applied for cell-level genomic profiling. For mutation profiling, the latter seems more natural. However, the task is highly challenging due to the limited input materials from only two copies of DNA molecules, while whole-genome amplification generates biases and other technical noises. ScRNA-seq starts with a higher input amount, so generally has better data quality. There exists various methods for mutation detection from DNA sequencing, it is not clear whether these methods work for scRNA-seq data.ResultsMutation detection methods developed for either bulk-cell sequencing data or scDNA-seq data do not work well for the scRNA-seq data, as they produce substantial numbers of false positives. We develop a novel and robust statistical method—called SCmut—to identify specific cells that harbor mutations discovered in bulk-cell data. Statistically SCmut controls the false positives using the 2D local false discovery rate method. We apply SCmut to several scRNA-seq datasets. In scRNA-seq breast cancer datasets SCmut identifies a number of highly confident cell-level mutations that are recurrent in many cells and consistent in different samples. In a scRNA-seq glioblastoma dataset, we discover a recurrent cell-level mutation in the PDGFRA gene that is highly correlated with a well-known in-frame deletion in the gene. To conclude, this study contributes a novel method to discover cell-level mutation information from scRNA-seq that can facilitate investigation of cell-to-cell heterogeneity.Availability and implementationThe source codes and bioinformatics pipeline of are available at https://github.com/nghiavtr/SCmut.Supplementary information Supplementary data are available at Bioinformatics online.

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.

Leveraging Spatial Variation in Tumor Purity for Improved Somatic Variant Calling of Archival Tumor Only Samples.

Archival tumor samples represent a rich resource of annotated specimens for translational genomics research. However, standard variant calling approaches require a matched normal sample from the same individual, which is often not available in the retrospective setting, making it difficult to distinguish between true somatic variants and individual-specific germline variants. Archival sections often contain adjacent normal tissue, but this tissue can include infiltrating tumor cells. As existing comparative somatic variant callers are designed to exclude variants present in the normal sample, a novel approach is required to leverage adjacent normal tissue with infiltrating tumor cells for somatic variant calling. Here we present lumosVar 2.0, a software package designed to jointly analyze multiple samples from the same patient, built upon our previous single sample tumor only variant caller lumosVar 1.0. The approach assumes that the allelic fraction of somatic variants and germline variants follow different patterns as tumor content and copy number state change. lumosVar 2.0 estimates allele specific copy number and tumor sample fractions from the data, and uses a to model to determine expected allelic fractions for somatic and germline variants and to classify variants accordingly. To evaluate the utility of lumosVar 2.0 to jointly call somatic variants with tumor and adjacent normal samples, we used a glioblastoma dataset with matched high and low tumor content and germline whole exome sequencing data (for true somatic variants) available for each patient. Both sensitivity and positive predictive value were improved when analyzing the high tumor and low tumor samples jointly compared to analyzing the samples individually or in-silico pooling of the two samples. Finally, we applied this approach to a set of breast and prostate archival tumor samples for which tumor blocks containing adjacent normal tissue were available for sequencing. Joint analysis using lumosVar 2.0 detected several variants, including known cancer hotspot mutations that were not detected by standard somatic variant calling tools using the adjacent tissue as presumed normal reference. Together, these results demonstrate the utility of leveraging paired tissue samples to improve somatic variant calling when a constitutional sample is not available.

Disruption of peroxisome proliferator–activated receptor γ coactivator (PGC)-1α reverts key features of the neoplastic phenotype of glioma cells

The peroxisome proliferator-activated receptor γ coactivator (PGC)-1α is a master regulator of mitochondrial biogenesis and controls metabolism by coordinating transcriptional events. Here, we interrogated whether PGC-1α is involved in tumor growth and the metabolic flexibility of glioblastoma cells. PGC-1α was expressed in a subset of established glioma cell lines and primary glioblastoma cell cultures. Furthermore, a higher PGC-1α expression was associated with an adverse outcome in the TCGA glioblastoma dataset. Suppression of PGC-1α expression by shRNA in the PGC-1α-positive U343MG glioblastoma line suppressed mitochondrial gene expression, reduced mitochondrial membrane potential, and diminished oxygen as well as glucose consumption, and lactate production. Compatible with the known PGC-1α functions in reactive oxygen species (ROS) metabolism, glioblastoma cells deficient in PGC-1α displayed ROS accumulation, had reduced RNA levels of proteins involved in ROS detoxification, and were more susceptible to death induction by H2O2 compared with control cells. PGC-1αsh cells also had impaired proliferation and migration rates in vitro and displayed less stem cell characteristics. Complementary effects were observed in PGC-1α-low LNT-229 cells engineered to overexpress PGC-1α. In an in vivo xenograft experiment, tumors formed by U343MG PGC-1αsh glioblastoma cells grew much slower than control tumors and were less invasive. Interestingly, the PGC-1α knockdown conferred protection against hypoxia-induced cell death, probably as a result of less active anabolic pathways, and this effect was associated with reduced epidermal growth factor expression and mammalian target of rapamycin signaling. In summary, PGC-1α modifies the neoplastic phenotype of glioblastoma cells toward more aggressive behavior and therefore makes PGC-1α a potential target for anti-glioblastoma therapies.

Rare but Recurrent ROS1 Fusions Resulting From Chromosome 6q22 Microdeletions are Targetable Oncogenes in Glioma.

Gliomas, a genetically heterogeneous group of primary central nervous system tumors, continue to pose a significant clinical challenge. Discovery of chromosomal rearrangements involving kinase genes has enabled precision therapy, and improved outcomes in several malignancies. Positing that similar benefit could be accomplished for patients with brain cancer, we evaluated The Cancer Genome Atlas (TCGA) glioblastoma dataset. Functional validation of the oncogenic potential and inhibitory sensitivity of discovered ROS1 fusions was performed using three independent cell-based model systems, and an in vivo murine xenograft study. In silico analysis revealed previously unreported intrachromosomal 6q22 microdeletions that generate ROS1-fusions from TCGA glioblastoma dataset. ROS1 fusions in primary glioma and ependymoma were independently corroborated from MSK-IMPACT and Foundation Medicine clinical datasets. GOPC-ROS1 is a recurrent ROS1 fusion in primary central nervous system (CNS) tumors. CEP85L-ROS1 and GOPC-ROS1 are transforming oncogenes in cells of astrocytic lineage, and amenable to pharmacologic inhibition with several ROS1 inhibitors even when occurring concurrently with other cancer hotspot aberrations frequently associated with glioblastoma. Oral monotherapy with a brain-permeable ROS1 inhibitor, lorlatinib, significantly prolonged survival in an intracranially xenografted tumor model generated from a ROS1 fusion-positive glioblastoma cell line. Our findings highlight that CNS tumors should be specifically interrogated for these rare intrachromosomal 6q22 microdeletion events that generate actionable ROS1 fusions. ROS1 fusions in primary brain cancer may be amenable for clinical intervention with kinase inhibitors, and this holds the potential of novel treatment paradigms in these treatment-refractory cancer types, particularly in glioblastoma.

Identification of hub genes and pathways in glioblastoma by bioinformatics analysis.

Glioblastoma (GBM) is the most common type of malignant brain tumor, and is associated with poor patient prognosis. A comprehensive understanding of the molecular mechanism underlying GBM may help to guide the identification of novel diagnoses and treatment targets. The gene expression profile of the GSE4290 GBM dataset was analyzed in order to identify differentially expressed genes (DEGs). Enriched pathways were identified through Gene Ontology and the Kyoto Encyclopedia of Genes and Genomes analyses. A protein-protein interaction network was constructed in order to identify hub genes and for module analysis. Expression and survival analyses were conducted in order to screen and validate critical genes. A total of 1,801 DEGs were recorded, including 620 upregulated and 1,181 downregulated genes. Upregulated DEGs were enriched in the terms ‘mitotic cell cycle process’, ‘mitotic cell cycle’ and ‘cell cycle process’. Downregulated genes were enriched in ‘transsynaptic signaling’, ‘anterograde transsynaptic signaling’ and ‘synaptic signaling’. A total of 15 hub genes, which displayed a high degree of connectivity, were selected. These genes included vascular endothelial growth factor A, cyclin-dependent kinase 1 (CDK1), cell-division cycle protein 20 (CDC20), aurora kinase A (AURKA), and budding uninhibited by benzimidazoles 1 (BUB1). The identified DEGs and hub genes may help guide investigations on the mechanisms underlying the development and progression of GBM. CDK1, CDC20, AURKA and BUB1, which are involved in cell cycle pathways, may be potential targets in the diagnosis and therapy of GBM.

A Radiosensitivity Gene Signature and PD-L1 Status Predict Clinical Outcome of Patients with Glioblastoma: Integrative Analyses of Transcriptome and Methylome

Combination of radiation therapy and immune checkpoint blockade such as programmed death-1 (PD-1) or programmed death-ligand 1 (PD-L1) blockade is being actively tested in clinical trial settings. Here, we tried to identify a subset of patients that could potentially benefit from this strategy using The Cancer Genome Atlas (TCGA) dataset for glioblastoma (GBM). Total 399 patients from TCGA GBM data set were clustered into radiosensitive (RS) versus radioresistant (RR) groups using 30 genes alleged to radiosensitivity. According to median value of CD274 mRNA expression, patients were stratified as PD-L1-High versus PD-L1-Low groups. Differentially expressed genes were identified, and enrichment analysis was performed as well. Differentially methylated regions were analyzed, and integrated into differentially expressed genes. A computational cell composition characterization tool was used for enumeration of various immune repertoire from transcriptome profile. We identified a subset of GBM, “PD-L1 high-RR group” which showed worse clinical outcome compared to the other groups. In this group, differentially expressed genes were highly enriched for immune response and mapped into activation of PI3K-AKT and MAPK signaling pathways. Integration of differentially expressed genes and differentially methylated region identified that MAP3K8 involved in T cell receptor signaling was upregulated and BA1, a factor which inhibits angiogenesis, was silenced. CIBERSORT showed that higher infiltration of immune repertoire contributed to immunosuppressive tumor microenvironment such as M2 macrophage and regulatory T cells. We validated the predictive value of a radiosensitivity gene signature and identified a subset, “PD-L1-High-RR group” which showed worse outcome following current standard therapy. The results of integrative analyses suggest that this group could potentially benefit from radiation therapy combined with PD-L1 blockade and angiogenesis inhibitor.

DNA Damage and Immune Response Biomarkers May Predict Toxicity and Response to Radiation Therapy Using a Blood Based Test

Radiation therapy is a critical part of cancer patient treatment regimens. While many patients respond to radiation therapy, others will not respond or will have toxicities that significantly reduce quality of life or outcome. Our objective was to create a gene expression assay from peripheral blood to predict radiation toxicity and/or response, thus allowing doctors to modify therapy based on the patient’s genetic background. A 302 gene Ampliseq RNA sequencing panel was developed to study radiation toxicity and response by performing Whole Genome Correlation Network Analysis (WGCNA) of publicly available radiation toxicity gene expression data sets in tissue, and radiation response (total body irradiation radiation biodosimetry data sets) in blood. To identify biological pathways of significance in therapy responders, KEGG and DAVID Analysis of the gene panel was performed on microarray data from the tissue of 71 prostate cancer and 318 glioblastoma patients in The Cancer Genome Atlas (TCGA). A targeted 31 gene assay was developed on a proprietary genomics platform. and a prospective study of 81 pelvic cancer patients representing 17 different cancer types was performed. This cohort included 34 patients reporting Grade 2 or Grade 3 acute radiation toxicity (3-month follow-up), and 12 rectal cancer patients with known tumor response. KEGG and DAVID analysis of the 302-gene radiation panel revealed a predominance of genes related to DNA repair, P53 signaling, Cell Cycle, Protein Ubiquitination, and Immune Response. In-silico analysis of the TCGA data revealed clear separation of responders and non-responders in the 71-patient prostate cancer data set, and clustering by survival after treatment in the 318-patient glioblastoma data set. Testing of pre-irradiation blood samples in the 81-patient prospective pelvic cancer study with the 31-gene assay showed 80.6% sensitivity and 90.7% specificity for predicting stage 2 or higher acute radiation toxicity, and 83.3% sensitivity and 85.7% specificity for predicting rectal cancer tumor response in the 12 rectal cancer patients with known Tumor Response Grading (TRG 0-4). A second IRB approved study is currently accruing additional patients to further validate these results. We were able to develop a gene expression signature from peripheral blood samples of cancer patients to predict radiation toxicity and response prior to the initiation of therapy. The blood based gene signature, comprised primarily of DNA damage and immune response genes, identified several genes also identified as biomarkers of radiation response/radiosensitivity from rectal cancer tumor samples. These results indicate that a gene expression based blood test may be developed to assist in radiation treatment decisions.

LIM and SH3 protein 1 regulates cell growth and chemosensitivity of human glioblastoma via the PI3K/AKT pathway

BackgroundLIM and SH3 protein 1 (LASP1) is upregulated in several types of human cancer and implicated in cancer progression. However, the expression and intrinsic function of LASP1 in glioblastoma (GBM) remains unclear.MethodOncomine and The Cancer Genome Atlas (TCGA) database was analyzed for the expression and clinical significance of LASP1 in GBM. LASP1 mRNA and protein level were measured by qRT-PCR and western blotting. The effect of LASP1 on GBM proliferation was examined by MTT assay and colony formation assay, the effect of LASP1 on sensitivity of Temozolomide was measured by flow cytometry and subcutaneous tumor model. The association between LASP1 and PI3K/AKT signaling was assessed by western blotting.ResultsOncomine GBM dataset analysis indicated LASP1 is significantly upregulated in GBM tissues compared to normal tissues. GBM dataset from The Cancer Genome Atlas (TCGA) revealed that high LASP1 expression is related to poor overall survival. LASP1 mRNA and protein in clinical specimens and tumor cell lines are frequently overexpressed. LASP1 knockdown dramatically suppressed U87 and U251 cell proliferation. Silencing LASP1 potentiated cell chemosensitivity to temozolomide in vitro, LASP1 knockdown inhibited tumor growth and enhanced the therapeutic effect of temozolomide in vivo. TCGA dataset analysis indicated LASP1 was correlated with PI3K/AKT signaling pathway, and LASP1 deletion inhibited this pathway. Combination treatment with PI3K/AKT pathway inhibitor LY294002 dramatically accelerated the suppression effect of temozolomide.ConclusionLASP1 may function as an oncogene in GBM and regulate cell proliferation and chemosensitivity in a PI3K/AKT-dependent mechanism. Thus, the LASP1/PI3K/AKT axis is a promising target and therapeutic strategy for GBM treatment.

Genome-wide methylomic and transcriptomic analyses identify subtype-specific epigenetic signatures commonly dysregulated in glioma stem cells and glioblastoma

ABSTRACTGlioma stem cells (GSCs), a subpopulation of tumor cells, contribute to tumor heterogeneity and therapy resistance. Gene expression profiling classified glioblastoma (GBM) and GSCs into four transcriptomically-defined subtypes. Here, we determined the DNA methylation signatures in transcriptomically pre-classified GSC and GBM bulk tumors subtypes. We hypothesized that these DNA methylation signatures correlate with gene expression and are uniquely associated either with only GSCs or only GBM bulk tumors. Additional methylation signatures may be commonly associated with both GSCs and GBM bulk tumors, i.e., common to non-stem-like and stem-like tumor cell populations and correlating with the clinical prognosis of glioma patients. We analyzed Illumina 450K methylation array and expression data from a panel of 23 patient-derived GSCs. We referenced these results with The Cancer Genome Atlas (TCGA) GBM datasets to generate methylomic and transcriptomic signatures for GSCs and GBM bulk tumors of each transcriptomically pre-defined tumor subtype. Survival analyses were carried out for these signature genes using publicly available datasets, including from TCGA. We report that DNA methylation signatures in proneural and mesenchymal tumor subtypes are either unique to GSCs, unique to GBM bulk tumors, or common to both. Further, dysregulated DNA methylation correlates with gene expression and clinical prognoses. Additionally, many previously identified transcriptionally-regulated markers are also dysregulated due to DNA methylation. The subtype-specific DNA methylation signatures described in this study could be useful for refining GBM sub-classification, improving prognostic accuracy, and making therapeutic decisions.

Prognostic role of mitochondrial pyruvate carrier in isocitrate dehydrogenase-mutant glioma.

OBJECTIVE Gliomas are one of the most common types of primary brain tumors. Recent studies have supported the importance of key genetic alterations, including isocitrate dehydrogenase (IDH) mutations and 1p19q codeletion, in glioma prognosis. Mutant IDH produces 2-hydroxyglutarate from α-ketoglutarate, a key metabolite of the Krebs cycle. The mitochondrial pyruvate carrier (MPC) is composed of MPC1 and MPC2 subunits and is functionally essential for the Krebs cycle. The authors sought to explore the impact of MPC1 and MPC2 expression on patient prognosis. METHODS Genomic and clinical data in patients with lower-grade glioma (WHO grades II and III) from The Cancer Genome Atlas (TCGA) were evaluated using Kaplan-Meier analysis and hazards modeling. Validation was conducted with additional data sets, including glioblastoma. RESULTS A total of 286 patients with lower-grade glioma (mean age 42.7 ± 13.5 years, 55.6% males) included 54 cases of IDH-wild type (18.9%); 140 cases of IDH-mutant, 1p19q-intact (49.0%); and 85 cases of IDH-mutant, 1p19q-codeleted (29.7%) tumors. Kaplan-Meier analysis showed that an MPC1 z-score > 0 distinguished better survival, particularly in IDH-mutant (p < 0.01) but not IDH-wild type tumors. Conversely, an MPC2 z-score > 0 identified worsened survival, particularly in IDH-mutant (p < 0.01) but not IDH-wild type tumors. Consistently, neither MPC1 nor MPC2 was predictive in a glioblastoma data set containing 5% IDH-mutant cases. Within the IDH-stratified lower-grade glioma data set, MPC1 status distinguished improved survival in 1p19q-codeleted tumors (p < 0.05), whereas MPC2 expression delineated worsened survival in 1p19q-intact tumors (p < 0.01). A hazards model identified IDH and 1p19q status, age (p = 0.01, HR = 1.03), Karnofsky Performance Scale (KPS) score (p = 0.03, HR = 0.97), and MPC1 (p = 0.003, HR = 0.52) but not MPC2 (p = 0.38) as key variables affecting overall survival. Further validation confirmed MPC1 as an independent predictor of lower-grade glioma. A clinical risk score using IDH and 1p19q status, age, KPS score, and MPC1 and MPC2 z-scores defined 4 risk categories for lower-grade glioma; this score was validated using a secondary glioma data set. CONCLUSIONS These results support the importance of MPC, especially MPC1, in improving prognostication of IDH-mutant tumors. The generation of a risk score system directly translates this finding to clinical application; however, further research to improve the molecular understanding of the role of MPC in the metabologenomic regulation of gliomas is warranted.