Mapping p53 mutations in low-grade glioma: a voxel-based neuroimaging analysis

Glioma lateralization: Focus on the anatomical localization and the distribution of molecular alterations (Review)

It is well known how the precise localization of glioblastoma multiforme (GBM) predicts the direction of tumor spread in the surrounding neuronal structures. The aim of the present review is to reveal the lateralization of GBM by evaluating the anatomical regions where it is frequently located as well as the main molecular alterations observed in different brain regions. According to the literature, the precise or most frequent lateralization of GBM has yet to be determined. However, it can be said that GBM is more frequently observed in the frontal lobe. Tractus and fascicles involved in GBM appear to be focused on the corticospinal tract, superior longitudinal I, II and III fascicles, arcuate fascicle long segment, frontal strait tract, and inferior fronto‑occipital fasciculus. Considering the anatomical features of GBM and its brain involvement, it is logical that the main brain regions involved are the frontal‑temporal‑parietal‑occipital lobes, respectively. Although tumor volumes are higher in the right hemisphere, it has been determined that the prognosis of patients diagnosed with cancer in the left hemisphere is worse, probably reflecting the anatomical distribution of some detrimental alterations such as TP53 mutations, PTEN loss, EGFR amplification, and MGMT promoter methylation. There are theories stating that the right hemisphere is less exposed to external influences in its development as it is responsible for the functions necessary for survival while tumors in the left hemisphere may be more aggressive. To shed light on specific anatomical and molecular features of GBM in different brain regions, the present review article is aimed at describing the main lateralization pathways as well as gene mutations or epigenetic modifications associated with the development of brain tumors.

The Relationship Between the Molecular Phenotypes of Brain Gliomas and the Imaging Features and Sensitivity of Radiotherapy and Chemotherapy

Gliomas are the most common primary malignant tumors of the brain, accounting for about 80% of all central nervous system malignancies. With the development of molecular biology, the molecular phenotypes of gliomas have been shown to be closely related to the process of diagnosis and treatment. The molecular phenotype of glioma also plays an important role in guiding treatment plans and evaluating treatment effects and prognosis. However, due to the heterogeneity of the tumors and the trauma associated with the surgical removal of tumor tissue, the application of molecular phenotyping in glioma is limited. With the development of imaging technology, functional magnetic resonance imaging (MRI) can provide structural and function information about tumors in a noninvasive and radiation-free manner. MRI is very important for the diagnosis of intracranial lesions. In recent years, with the development of the technology for tumor molecular diagnosis and imaging, the use of molecular phenotype information and imaging procedures to evaluate the treatment outcome of tumors has become a hot topic. By reviewing the related literature on glioma treatment and molecular typing that has been published in the past 20 years, and referring to the latest 2020 NCCN treatment guidelines, summarizing the imaging characteristic and sensitivity of radiotherapy and chemotherapy of different molecular phenotypes of glioma. In this article, we briefly review the imaging characteristics of different molecular phenotypes in gliomas and their relationship with radiosensitivity and chemosensitivity of gliomas.

Pan-cancer multi-omics analysis of PTBP1 reveals it as an inflammatory, progressive and prognostic marker in glioma

PTBP1 is an oncogene that regulates the splicing of precursor mRNA. However, the relationship between PTBP1 expression and gene methylation, cancer prognosis, and tumor microenvironment remains unclear. The expression profiles of PTBP1 across various cancers were derived from the TCGA, as well as the GTEx and CGGA databases. The CGGA mRNA_325, CGGA mRNA_301, and CGGA mRNA_693 datasets were utilized as validation cohorts. Immune cell infiltration scores were approximated using the TIMER 2.0 tool. Functional enrichment analysis for groups with high and low PTBP1 expression was conducted using Gene Set Enrichment Analysis (GSEA). Methylation data were predominantly sourced from the SMART and Mexpress databases. Linked-omics analysis was employed to perform functional enrichment analysis of genes related to PTBP1 methylation, as well as to conduct protein functional enrichment analysis. Single-cell transcriptome analysis and spatial transcriptome analysis were carried out using Seurat version 4.10. Compared to normal tissues, PTBP1 is significantly overexpressed and hypomethylated in various cancers. It is implicated in prognosis, immune cell infiltration, immune checkpoint expression, genomic variation, tumor neoantigen load, and tumor mutational burden across a spectrum of cancers, with particularly notable effects in low-grade gliomas. In the context of gliomas, PTBP1 expression correlates with WHO grade and IDH1 mutation status. PTBP1 expression and methylation play an important role in a variety of cancers. PTBP1 can be used as a marker of inflammation, progression and prognosis in gliomas.

The involvement of brain regions associated with lower KPS and shorter survival time predicts a poor prognosis in glioma

Background

Isocitrate dehydrogenase-wildtype glioblastoma (IDH-wildtype GBM) and IDH-mutant astrocytoma have distinct biological behaviors and clinical outcomes. The location of brain tumors is closely associated not only with clinical symptoms and prognosis but also with key molecular alterations such as IDH. Therefore, we hypothesize that the key brain regions influencing the prognosis of glioblastoma and astrocytoma are likely to differ. This study aims to (1) identify specific regions that are associated with the Karnofsky Performance Scale (KPS) or overall survival (OS) in IDH-wildtype GBM and IDH-mutant astrocytoma and (2) test whether the involvement of these regions could act as a prognostic indicator.

Methods

A total of 111 patients with IDH-wildtype GBM and 78 patients with IDH-mutant astrocytoma from the Cancer Imaging Archive database were included in the study. Voxel-based lesion-symptom mapping (VLSM) was used to identify key brain areas for lower KPS and shorter OS. Next, we analyzed the structural and cognitive dysfunction associated with these regions. The survival analysis was carried out using Kaplan-Meier survival curves. Another 72 GBM patients and 48 astrocytoma patients from Harbin Medical University Cancer Hospital were used as a validation cohort.

Results

Tumors located in the insular cortex, parahippocampal gyrus, and middle and superior temporal gyrus of the left hemisphere tended to lead to lower KPS and shorter OS in IDH-wildtype GBM. The regions that were significantly correlated with lower KPS in IDH-mutant astrocytoma included the subcallosal cortex and cingulate gyrus. These regions were associated with diverse structural and cognitive impairments. The involvement of these regions was an independent predictor for shorter survival in both GBM and astrocytoma.

Conclusion

This study identified the specific regions that were significantly associated with OS or KPS in glioma. The results may help neurosurgeons evaluate patient survival before surgery and understand the pathogenic mechanisms of glioma in depth.

Development and validation of intraoral periapical radiography-based machine learning model for periodontal defect diagnosis

Radiographic determination of the bone level is useful in the diagnosis and determination of the severity of the periodontal disease. Various two- and three-dimensional imaging modalities offer choices for imaging pathologic processes that affect the periodontium. In recent years, innovative computer techniques, especially artificial intelligence (AI), have begun to be used in many areas of dentistry and are helping increase treatment and diagnostic performance. This study was aimed at developing a machine-learning (ML) model and assessing the extent to which it was capable of classifying periodontal defects on 2D periapical images. Eighty-seven periapical images were examined as part of this research. The existence or absence of periodontal defects in the aforementioned images were evaluated by a human observer. The evaluations were subsequently repeated using a radiomics platform. A comparison was made of all data acquired through human observation and ML techniques by SVM analysis. According to the study findings the ability of human observers and the ML model to detect periodontal defects was significantly different in comparison to the gold standard. However, ML and human observers performed similarly for the detection of periodontal defects without a significant difference. This study reveals that the prediction of periodontal defects can be achieved by combining particular radiomic features with image variables. The proposed machine leaning model can be utilized for supporting clinical practitioners and eventually substitute evaluations conducted by human observers while enhancing future levels of performance.

Advanced Diagnosis of Glioma by Using Emerging Magnetic Resonance Sequences

Glioma, the most common primary brain tumor in adults, can be difficult to discern radiologically from other brain lesions, which affects surgical planning and follow-up treatment. Recent advances in MRI demonstrate that preoperative diagnosis of glioma has stepped into molecular and algorithm-assisted levels. Specifically, the histology-based glioma classification is composed of multiple different molecular subtypes with distinct behavior, prognosis, and response to therapy, and now each aspect can be assessed by corresponding emerging MR sequences like amide proton transfer-weighted MRI, inflow-based vascular-space-occupancy MRI, and radiomics algorithm. As a result of this novel progress, the clinical practice of glioma has been updated. Accurate diagnosis of glioma at the molecular level can be achieved ahead of the operation to formulate a thorough plan including surgery radical level, shortened length of stay, flexible follow-up plan, timely therapy response feedback, and eventually benefit patients individually.

Gliomas - An experience based on molecular markers

Background

Gliomas account for 45% of all intracranial tumors. Newer technologies have allowed deeper genetic and epigenetic analysis leading to the discovery of IDH (Isocitrate dehydrogenase) mutations and their association with ATRX (alpha-thalassemia/mental retardation syndrome X-linked) and p53, for better diagnosis and prognosis. In this study, we analysed their expression and correlated with various clinicopathological parameters. A follow up to prognosticate gliomas based on the molecular findings is also attempted.

Materials and Method

During last 5 years both retrospective and prospective cases were included in the study. Immunohistochemistry for IDH1, ATRX, and p53 was done and reported based on intensity and percentage of tumor cells expressing the markers.

Results

A total of 53 cases of gliomas were included, excluding primary glioblastomas and ependymomas. The patient's age ranged from 10 to 53 years. The male to female ratio was 1.3:1. IDH1 positivity was seen in 88% of diffuse astrocytoma, 80% of anaplastic astrocytoma, 90% of oligodendroglioma, 60% of anaplastic oligodendroglioma, and 54% of glioblastoma. A significant association was seen between positive IDH1 expression and low-grade gliomas (p = 0.028). A combined analysis of expression of IDH1 and ATRX versus IDH1, ATRX, and p53 with WHO grade showed a statistically significant association. A follow-up of 32 patients was available. Out of 24 IDH1+ (positive) cases, 22 patients had a median survival of 21.5 months (92%). Out of 8 IDH1- (negative) cases, 5 had a median survival of 15.8 months (62%).

Conclusion

Gliomas expressing IDH1 mutation show improved survival of patients. Combined analysis of IDH1, ATRX, and p53 has diagnostic and prognostic significance. For routine cases of gliomas, a combination of IDH1 and ATRX are sufficient; however, the use of p53 is recommended for further prognostication and for possible targeted therapy in the future.

Imaging Findings of New Entities and Patterns in Brain Tumor: Isocitrate Dehydrogenase Mutant, Isocitrate Dehydrogenase Wild-Type, Codeletion, and MGMT Methylation

Molecular features are now essential in distinguishing between glioma histologic subtypes. Currently, isocitrate dehydrogenase mutation, 1p19q codeletion, and MGMT methylation status play significant roles in optimizing medical and surgical treatment. Noninvasive pretreatment and post-treatment determination of glioma subtype is of great interest. Although imaging cannot replace the genetic panel at present, image findings have shown promising signs to identify and diagnose the types and subtypes of gliomas. This article details key imaging findings in the most common molecular glioma subtypes and highlights recent advances in imaging technologies to differentiate these lesions noninvasively.

Analysis of morphological characteristics of IDH-mutant/wildtype brain tumors using whole-lesion phenotype analysis

BACKGROUND

Although IDH-mutant tumors aggregate to the frontotemporal regions, the clustering pattern of IDH-wildtype tumors is less clear. As voxel-based lesion-symptom mapping (VLSM) has several limitations for solid lesion mapping, a new technique, whole-lesion phenotype analysis (WLPA), is developed. We utilize WLPA to assess spatial clustering of tumors with IDH mutation from The Cancer Genome Atlas and The Cancer Imaging Archive.

METHODS

The degree of tumor clustering segmented from T1 weighted images is measured to every other tumor by a function of lesion similarity to each other via the Hausdorff distance. Each tumor is ranked according to the degree to which its neighboring tumors show identical phenotypes, and through a permutation technique, significant tumors are determined. VLSM was applied through a previously described method.

RESULTS

A total of 244 patients of mixed-grade gliomas (WHO II-IV) are analyzed, of which 150 were IDH-wildtype and 139 were glioblastomas. VLSM identifies frontal lobe regions that are more likely associated with the presence of IDH mutation but no regions where IDH-wildtype was more likely to be present. WLPA identifies both IDH-mutant and -wildtype tumors exhibit statistically significant spatial clustering.

CONCLUSION

WLPA may provide additional statistical power when compared with VLSM without making several potentially erroneous assumptions. WLPA identifies tumors most likely to exhibit particular phenotypes, rather than producing anatomical maps, and may be used in conjunction with VLSM to understand the relationship between tumor morphology and biologically relevant tumor phenotypes.

Dysregulated Expression and Methylation Analysis Identified TLX1NB as a Novel Recurrence Marker in Low-Grade Gliomas

Low-grade gliomas (LGGs) are the most common CNS tumors, and the main therapy for LGGs is complete surgical resection, due to its curative effect. However, LGG recurrence occurs frequently. Biomarkers play a crucial role in evaluating the recurrence and prognosis of LGGs. Numerous studies have focused on LGG prognosis. However, the multiomics research investigating the roles played by gene methylation and expression in LGG recurrence remains limited. In this study, we integrated the TCGA and GEO datasets, analyzing RNA and methylation data for recurrence (R) and nonrecurrence (NR) groups. We found a low expression of TLX1NB and high methylation in recurrence patients. Low expression of TLX1NB is associated with poor survival (OS: p = 0.04). The expression of TLX1NB is likely to play a role in the prognosis of LGG. Therefore, TLX1NB may represent an alternative early biomarker for the recurrence of low-grade gliomas.

Predicting the Type of Tumor-Related Epilepsy in Patients With Low-Grade Gliomas: A Radiomics Study

Purpose: The majority of patients with low-grade gliomas (LGGs) experience tumor-related epilepsy during the disease course. Our study aimed to build a radiomic prediction model for LGG-related epilepsy type based on magnetic resonance imaging (MRI) data.

Methods: A total of 205 cases with LGG-related epilepsy were enrolled in the retrospective study and divided into training and validation cohorts (1:1) according to their surgery time. Seven hundred thirty-four radiomic features were extracted from T2-weighted imaging, including six location features. Pearson correlation coefficient, univariate area under curve (AUC) analysis, and least absolute shrinkage and selection operator regression were adopted to select the most relevant features for the epilepsy type to build a radiomic signature. Furthermore, a novel radiomic nomogram was developed for clinical application using the radiomic signature and clinical variables from all patients.

Results: Four MRI-based features were selected from the 734 radiomic features, including one location feature. Good discriminative performances were achieved in both training (AUC = 0.859, 95% CI = 0.787–0.932) and validation cohorts (AUC = 0.839, 95% CI = 0.761–0.917) for the type of epilepsy. The accuracies were 80.4 and 80.6%, respectively. The radiomic nomogram also allowed for a high degree of discrimination. All models presented favorable calibration curves and decision curve analyses.

Conclusion: Our results suggested that the MRI-based radiomic analysis may predict the type of LGG-related epilepsy to enable individualized therapy for patients with LGG-related epilepsy.

Detection and Correlation of Single and Concomitant TP53, PTEN, and CDKN2A Alterations in Gliomas

Gliomas are the most frequent primary tumors of central nervous system and represent a heterogeneous group of tumors that originates from the glial cells. TP53, PTEN, and CDKN2A are important tumor suppressor genes that encode proteins involved in sustaining cellular homeostasis by different signaling pathways. Though genetic alterations in these genes play a significant role in tumorigenesis, few studies are available regarding the incidence and relation of concomitant TP53, PTEN, and CDKN2A alterations in gliomas. The purpose of this study was to evaluate the occurrence of mutation and deletion in these genes, through single-strand conformational polymorphism, array-comparative genomic hybridization, and fluorescence in situ hybridization techniques, in 69 gliomas samples. Molecular results demonstrated a significant higher prevalence of TP53, PTEN, and CDKN2A alterations in astrocytoma than other tumor subtypes, and heterozygous deletion was the most frequent event. In addition, a significant association was observed between TP53 and CDKN2A alterations (p = 0.0424), which tend to coexist in low grade astrocytomas (5/46 cases (10.9%)), suggesting that they are early events in development of these tumors, and PTEN and CDKN2A deletions (p = 0.0022), which occurred concomitantly in 9/50 (18%) patients, with CDKN2A changes preceding PTEN deletions, present preferably in high-grade gliomas.

Clinical Targeted Next-Generation sequencing Panels for Detection of Somatic Variants in Gliomas

PURPOSE

Targeted next-generation sequencing (NGS) panels for solid tumors have been useful in clinical framework for accurate tumor diagnosis and identifying essential molecular aberrations. However, most cancer panels have been designed to address a wide spectrum of pan-cancer models, lacking integral prognostic markers that are highly specific to gliomas.

Materials and Methods

To address such challenges, we have developed a glioma-specific NGS panel, termed "GliomaSCAN," that is capable of capturing single nucleotide variations and insertion/deletion, copy number variation, and selected promoter mutations and structural variations that cover a subset of intron regions in 232 essential glioma-associated genes. We confirmed clinical concordance rate using pairwise comparison of the identified variants from whole exome sequencing (WES), immunohistochemical analysis, and fluorescence in situ hybridization.

RESULTS

Our panel demonstrated high sensitivity in detecting potential genomic variants that were present in the standard materials. To ensure the accuracy of our targeted sequencing panel, we compared our targeted panel to WES. The comparison results demonstrated a high correlation. Furthermore, we evaluated clinical utility of our panel in 46 glioma patients to assess the detection capacity of potential actionable mutations. Thirty-two patients harbored at least one recurrent somatic mutation in clinically actionable gene.

CONCLUSION

We have established a glioma-specific cancer panel. GliomaSCAN highly excelled in capturing somatic variations in terms of both sensitivity and specificity and provided potential clinical implication in facilitating genome-based clinical trials. Our results could provide conceptual advance towards improving the response of genomically guided molecularly targeted therapy in glioma patients.

A study on the correlation between STAT‑1 and mutant p53 expression in glioma

Glioma is the most common primary brain tumor in adults and the second most common malignant tumor in children. Aberrant expression of signal transducer and activator of transcription 1 (STAT‑1) and p53 are known to affect the occurrence and progression of malignant tumors. The aim of the present study was to investigate the expression of STAT‑1 and mutant p53 gene, as well as their correlation, in patients with glioma. The present study included 50 patients who underwent glioma resection at the First Affiliated Hospital of Inner Mongolia Medical University between December 2007 and December 2011, and 10 patients with acute cerebral contusion who underwent intracerebral hematoma removal at the same hospital between January 2013 and January 2014. The expression of STAT‑1 and mutant p53 protein in patients with different grades of glioma was assessed by immunohistochemistry. Spearman's correlation coefficient was employed to examine the correlation between STAT‑1 and the grade of glioma, and mutant p53 expression. The results demonstrated that the mean expression of STAT‑1 in glioma was significantly lower compared with normal brain tissue (P<0.05). However, there was no significant difference in the STAT‑1 positive expression rate between the two groups (χ2=1.38, P>0.05). The expression score (P<0.05) and positive expression rate (χ2=31.27, P<0.05) of mutant p53 in glioma was significantly higher compared with those in normal brain tissue. Statistical analysis revealed a negative correlation between STAT‑1 expression and the grade of glioma (r=‑0.767, P<0.05). In addition, mutant p53 expression was negatively correlated with STAT‑1 expression in glioma (r=‑0.876, P<0.05). The observed negative correlation between STAT‑1 and the pathological grade of glioma suggested an association between STAT‑1 and the occurrence and development of glioma, thus revealing the potential of STAT‑1 as a diagnostic biomarker and therapeutic target for glioma.

Regional specificity of 1p/19q co-deletion combined with radiological features for predicting the survival outcomes of anaplastic oligodendroglial tumor patients

In this study we aimed to identify the anatomic features of 1p/19q co-deletion and investigate the predictive values of tumor location and radiological characteristics for the survival of anaplastic oligodendroglial (AO) glioma patients. Voxel-based lesion-symptom mapping (VLSM) analysis was applied to define the brain regions associated with occurrence of 1p/19q co-deletion in a cohort of 206 AO tumor patients (discovery set) treated between May 2009 and September 2013. Retrospectively, the acquired clusters and radiological features were subjected to Kaplan-Meier survival analysis using data from the Chinese Glioma Genome Atlas (validation set) to evaluate their prognostic role in AO patients. The institutional review board approved this study. The right frontal lobe and right anterior insular lobe were specifically associated with high occurrence of 1p/19q co-deletion. For AO tumors not involving these areas, the absence of contrast enhancement predicted longer progression-free (p = 0.018) and overall survival (p = 0.020); moreover, in patients with contrast enhancement, edema could stratify the survival outcome (p = 0.013 for progression-free survival, p = 0.016 for overall survival). For AO tumors located in the VLSM-identified regions, edema was also able to stratify the survival outcome of patients without contrast enhancement (p = 0.025 for progression-free survival, p = 0.028 for overall survival). The 1p/19q co-deletion showed predilection for specific brain regions. According to the tumor involvement of VLSM-identified regions associated with 1p/19q co-deletion, radiological features were predictive for AO patient survival outcomes.

MRI features predict p53 status in lower-grade gliomas via a machine-learning approach

Background

P53 mutation status is a pivotal biomarker for gliomas. Here, we developed a machine-learning model to predict p53 status in lower-grade gliomas based on radiomic features extracted from conventional magnetic resonance (MR) images.

Methods

Preoperative MR images were retrospectively obtained from 272 patients with primary grade II/III gliomas. The patients were randomly allocated in a 2:1 ratio to a training (n = 180) or validation (n = 92) set. A total of 431 radiomic features were extracted from each patient. The lest absolute shrinkage and selection operator (LASSO) method was used for feature selection and radiomic signature construction. Subsequently, a machine-learning model to predict p53 status was established using the selected features and a Support Vector Machine classifier. The predictive performance of all individual features and the model was calculated using receiver operating characteristic curves in both the training and validation sets.

Results

The p53-related radiomic signature was built using the LASSO algorithm; this procedure consisted of four first-order statistics or related wavelet features (including Maximum, Median, Minimum, and Uniformity), a shape and size-based feature (Spherical Disproportion), and ten textural features or related wavelet features (including Correlation, Run Percentage, and Sum Entropy). The prediction accuracies based on the area under the curve were 89.6% in the training set and 76.3% in the validation set, which were better than individual features.

Conclusions

These results demonstrate that MR image texture features are predictive of p53 mutation status in lower-grade gliomas. Thus, our procedure can be conveniently used to facilitate presurgical molecular pathological diagnosis.

•

We established a p53-related radiomic signature in lower-grade gliomas based on LASSO algorithm.

•

We developed a machine-learning model using the radiomic signature and a support vector machine.

•

P53 mutation status of lower-grade gliomas was predicted effectively based on our machine-learning model.

Anatomic mapping of molecular subtypes in diffuse glioma

Background

Tumor location served as an important prognostic factor in glioma patients was considered to postulate molecular features according to cell origin theory. However, anatomic distribution of unique molecular subtypes was not widely investigated. The relationship between molecular phenotype and histological subgroup were also vague based on tumor location. Our group focuses on the study of glioma anatomic location of distinctive molecular subgroups and histology subtypes, and explores the possibility of their consistency based on clinical background.

Methods

We retrospectively reviewed 143 cases with both molecular information (IDH1/TERT/1p19q) and MRI images diagnosed as cerebral diffuse gliomas. The anatomic distribution was analyzed between distinctive molecular subgroups and its relationship with histological subtypes. The influence of tumor location, molecular stratification and histology diagnosis on survival outcome was investigated as well.

Results

Anatomic locations of cerebral diffuse glioma indicate varied clinical outcome. Based on that, it can be stratified into five principal molecular subgroups according to IDH1/TERT/1p19q status. Triple-positive (IDH1 and TERT mutation with 1p19q codeletion) glioma tended to be oligodendroglioma present with much better clinical outcome compared to TERT mutation only group who is glioblastoma inclined (median overall survival 39 months VS 18 months). Five molecular subgroups were demonstrated with distinctive locational distribution. This kind of anatomic feature is consistent with its corresponding histological subtypes.

Discussion

Each molecular subgroup in glioma has unique anatomic location which indicates distinctive clinical outcome. Molecular diagnosis can be served as perfect complementary tool for the precise diagnosis. Integration of histomolecular diagnosis will be much more helpful in routine clinical practice in the future.

MRI features can predict EGFR expression in lower grade gliomas: A voxel-based radiomic analysis

OBJECTIVE

To identify the magnetic resonance imaging (MRI) features associated with epidermal growth factor (EGFR) expression level in lower grade gliomas using radiomic analysis.

METHODS

270 lower grade glioma patients with known EGFR expression status were randomly assigned into training (n=200) and validation (n=70) sets, and were subjected to feature extraction. Using a logistic regression model, a signature of MRI features was identified to be predictive of the EGFR expression level in lower grade gliomas in the training set, and the accuracy of prediction was assessed in the validation set.

RESULTS

A signature of 41 MRI features achieved accuracies of 82.5% (area under the curve [AUC] = 0.90) in the training set and 90.0% (AUC = 0.95) in the validation set. This radiomic signature consisted of 25 first-order statistics or related wavelet features (including range, standard deviation, uniformity, variance), one shape and size-based feature (spherical disproportion), and 15 textural features or related wavelet features (including sum variance, sum entropy, run percentage).

CONCLUSIONS

A radiomic signature allowing for the prediction of the EGFR expression level in patients with lower grade glioma was identified, suggesting that using tumour-derived radiological features for predicting genomic information is feasible.

KEY POINTS

• EGFR expression status is an important biomarker for gliomas. • EGFR in lower grade gliomas could be predicted using radiogenomic analysis. • A logistic regression model is an efficient approach for analysing radiomic features.

Genomic profiling of long non-coding RNA and mRNA expression associated with acquired temozolomide resistance in glioblastoma cells

Temozolomide (TMZ) is an alkylating chemotherapeutic agent widely used in anti-glioma treatment. However, acquired TMZ resistance represents a major clinical challenge that leads to tumor relapse or progress. This study investigated the genomic profiles including long non-coding RNA (lncRNA) and mRNA expression associated with acquired TMZ resistance in glioblastoma (GBM) cells in vitro. The TMZ-resistant (TR) of GBM sub-cell lines were established through repetitive exposure to increasing TMZ concentrations in vitro. The differentially expressed lncRNAs and mRNAs between the parental U87 and U87TR cells were detected by human lncRNA microarray method. In this study, we identified 2,692 distinct lncRNAs demonstrating >2-fold differential expression with 1,383 lncRNAs upregulated and 1,309 lncRNAs downregulated. Moreover, 4,886 differential mRNAs displayed 2,933 mRNAs upregulated and 1,953 mRNAs downregulated. Further lncRNA classification and subgroup analysis revealed the potential functions of the lncRNA-mRNA relationship associated with the acquired TMZ resistance. Gene ontology and pathway analysis on mRNAs showed significant biological regulatory genes and pathways involved in acquired TMZ resistance. Moreover, we found the ECM‑receptor interaction pathway was significantly downregulated and ECM related collagen Ι, fibronectin, laminin and CD44 were closely associated with the TR phenotype in vitro. Our findings indicate that the dysregulated lncRNAs and mRNAs identified in this work may provide novel targets for overcoming acquired TMZ resistance in GBM chemotherapy.

IDH mutation and 1p19q codeletion distinguish two radiological patterns of diffuse low-grade gliomas

Diffuse low-grade gliomas (DLGG) prognosis is variable, depending on several factors, including the isocitrate dehydrogenase (IDH) mutation and the 1p19q codeletion. A few studies suggested associations between these parameters and tumor radiological characteristics including topography. Our aim was analyzing the correlations between the IDH and 1p19q statuses and the tumor intracerebral distribution (at the lobar and voxel levels), volume, and borders. We conducted a retrospective, monocentric study on a consecutive series of 198 DLGG patients. The IDH and 1p19q statuses were recorded. The pre-treatment magnetic resonance FLAIR imagings were reviewed for determination of lobar topography, tumor volume, and characterisation of tumor borders (sharp or indistinct). We conducted a voxel-based lesion-symptom mapping analysis to investigate the correlations between the IDH and 1p19q statuses and topography at the voxel level. The IDH mutation and 1p19q statuses were correlated with the tumor topography defined using lobar anatomy (p < 0.001 and p = 0.004, respectively). Frontal tumors were more frequently IDH-mutant (87.1 vs. 57.4%) and 1p19q codeleted (45.2 vs. 17.0%) than temporo-insular lesions. At the voxel level, these associations were not found. Tumors with sharp borders were more frequently IDH-mutant (p = 0.001) while tumors with indistinct borders were more frequently IDH wild-type and 1p19q non-codeleted (p < 0.001). Larger tumors at diagnosis (possibly linked to a slower growth rate) were more frequently IDH-mutant (p < 0.001). IDH wild-type, 1p19q non-codeleted temporo-insular tumors are distinct from IDH-mutant, 1p19q codeleted frontal tumors. Further studies are needed to determine whether the therapeutic strategy should be adapted to each pattern.

Noninvasive IDH1 mutation estimation based on a quantitative radiomics approach for grade II glioma

OBJECTIVE

The status of isocitrate dehydrogenase 1 (IDH1) is highly correlated with the development, treatment and prognosis of glioma. We explored a noninvasive method to reveal IDH1 status by using a quantitative radiomics approach for grade II glioma.

METHODS

A primary cohort consisting of 110 patients pathologically diagnosed with grade II glioma was retrospectively studied. The radiomics method developed in this paper includes image segmentation, high-throughput feature extraction, radiomics sequencing, feature selection and classification. Using the leave-one-out cross-validation (LOOCV) method, the classification result was compared with the real IDH1 situation from Sanger sequencing. Another independent validation cohort containing 30 patients was utilised to further test the method.

RESULTS

A total of 671 high-throughput features were extracted and quantized. 110 features were selected by improved genetic algorithm. In LOOCV, the noninvasive IDH1 status estimation based on the proposed approach presented an estimation accuracy of 0.80, sensitivity of 0.83 and specificity of 0.74. Area under the receiver operating characteristic curve reached 0.86. Further validation on the independent cohort of 30 patients produced similar results.

CONCLUSIONS

Radiomics is a potentially useful approach for estimating IDH1 mutation status noninvasively using conventional T2-FLAIR MRI images. The estimation accuracy could potentially be improved by using multiple imaging modalities.

KEY POINTS

• Noninvasive IDH1 status estimation can be obtained with a radiomics approach. • Automatic and quantitative processes were established for noninvasive biomarker estimation. • High-throughput MRI features are highly correlated to IDH1 states. • Area under the ROC curve of the proposed estimation method reached 0.86.

Mapping genetic factors in high-grade glioma patients

BACKGROUND

Tumor location, which serves as a prognostic factor for high-grade gliomas, may reflect the molecular and genetic phenotype of tumor initiate cells and thus predict tumor origin. Therefore, the purpose of this study was to combine radiographic atlases and tumor biomarkers through a voxel-based neuroimaging approach.

METHODS

Preoperative MRIs were collected from 65 newly diagnosed patients with histologically confirmed high-grades gliomas. These samples were analyzed for TP53 mutations and MMP-9.PTEN, MGMT, EGFR and IDH1 statuses using a statistical voxel-based lesion-symptom mapping (VLSM) method, which correlates the anatomical location of HGGs with their molecular profile.

RESULTS

VLSM analysis identified P53, Wild-type IDH and EGFR overexpression mutations in the white matter of the periventricular region in the left hemisphere, which can be predicted by a short overall survival from the time of diagnosis. The lack of MGMT promoter methylation deep in the right frontal lobe region indicates a poor prognosis.

CONCLUSIONS

Our study demonstrates that different molecular phenotypes are related to specific brain regions. In addition, the structural MRI and genetic profile-based analysis of brain regions associated with survival-associated factors could be used in planning glioma operations and clinical survival predictions.

Anatomical specificity of vascular endothelial growth factor expression in glioblastomas: a voxel-based mapping analysis

INTRODUCTION

The expression of vascular endothelial growth factor (VEGF) is a common genetic alteration in malignant gliomas and contributes to the angiogenesis of tumors. This study aimed to investigate the anatomical specificity of VEGF expression levels in glioblastomas using voxel-based neuroimaging analysis.

METHODS

Clinical information, MR scans, and immunohistochemistry stains of 209 patients with glioblastomas were reviewed. All tumor lesions were segmented manually and subsequently registered to standard brain space. Voxel-based regression analysis was performed to correlate the brain regions of tumor involvement with the level of VEGF expression. Brain regions identified as significantly associated with high or low VEGF expression were preserved following permutation correction.

RESULTS

High VEGF expression was detected in 123 (58.9 %) of the 209 patients. Voxel-based statistical analysis demonstrated that high VEGF expression was more likely in tumors located in the left frontal lobe and the right caudate and low VEGF expression was more likely in tumors that occurred in the posterior region of the right lateral ventricle.

CONCLUSION

Voxel-based neuroimaging analysis revealed the anatomic specificity of VEGF expression in glioblastoma, which may further our understanding of genetic heterogeneity during tumor origination. This finding provides primary theoretical support for potential future application of customized antiangiogenic therapy.