Glioblastoma multiforme (GBM) is the most frequent and deadly human brain cancer. Early diagnosis through non-invasive biomarkers may render GBM more easily treatable, improving the prognosis of this currently incurable disease. We suggest the use of serum extracellular vesicle (sEV)-derived circular RNAs (circRNAs) as highly stable minimally invasive diagnostic biomarkers for GBM diagnosis. EVs were isolated by size exclusion chromatography from sera of 23 GBM and 5 grade 3 glioma (GIII) patients, and 10 unaffected controls (UC). The expression of two candidate circRNAs (circSMARCA5 and circHIPK3) was assayed by droplet digital PCR. CircSMARCA5 and circHIPK3 were significantly less abundant in sEVs from GBM patients with respect to UC (fold-change (FC) of −2.15 and −1.92, respectively) and GIII (FC of −1.75 and −1.4, respectively). Receiver operating characteristic curve (ROC) analysis, based on the expression of sEV-derived circSMARCA5 and circHIPK3, allowed us to distinguish GBM from UC (area under the curve (AUC) 0.823 (0.667–0.979) and 0.855 (0.704 to 1.000), with a 95% confidence interval (CI), respectively). Multivariable ROC analysis, performed by combining the expression of sEV-derived circSMARCA5 and circHIPK3 with preoperative neutrophil to lymphocyte (NLR), platelet to lymphocyte (PLR) and lymphocyte to monocyte (LMR) ratios, three known diagnostic and prognostic GBM markers, allowed an improvement in the GBM diagnostic accuracy (AUC 0.901 (0.7912 to 1.000), 95% CI). Our data suggest sEV-derived circSMARCA5 and circHIPK3 as good diagnostic biomarkers for GBM, especially when associated with preoperative NLR, PLR and LMR.

Circular RNAs (circRNAs) are a large class of RNAs with regulatory functions within cells. We recently showed that circSMARCA5 is a tumor suppressor in glioblastoma multiforme (GBM) and acts as a decoy for Serine and Arginine Rich Splicing Factor 1 (SRSF1) through six predicted binding sites (BSs). Here we characterized RNA motifs functionally involved in the interaction between circSMARCA5 and SRSF1. Three different circSMARCA5 molecules (Mut1, Mut2, Mut3), each mutated in two predicted SRSF1 BSs at once, were obtained through PCR-based replacement of wild-type (WT) BS sequences and cloned in three independent pcDNA3 vectors. Mut1 significantly decreased its capability to interact with SRSF1 as compared to WT, based on the RNA immunoprecipitation assay. In silico analysis through the “Find Individual Motif Occurrences” (FIMO) algorithm showed GAUGAA as an experimentally validated SRSF1 binding motif significantly overrepresented within both predicted SRSF1 BSs mutated in Mut1 (q-value = 0.0011). U87MG and CAS-1, transfected with Mut1, significantly increased their migration with respect to controls transfected with WT, as revealed by the cell exclusion zone assay. Immortalized human brain microvascular endothelial cells (IM-HBMEC) exposed to conditioned medium (CM) harvested from U87MG and CAS-1 transfected with Mut1 significantly sprouted more than those treated with CM harvested from U87MG and CAS-1 transfected with WT, as shown by the tube formation assay. qRT-PCR showed that the intracellular pro- to anti-angiogenic Vascular Endothelial Growth Factor A (VEGFA) mRNA isoform ratio and the amount of total VEGFA mRNA secreted in CM significantly increased in Mut1-transfected CAS-1 as compared to controls transfected with WT. Our data suggest that GAUGAA is the RNA motif responsible for the interaction between circSMARCA5 and SRSF1 as well as for the circSMARCA5-mediated control of GBM cell migration and angiogenic potential.

Radiomics holds great promise in the field of cancer management. However, the clinical application of radiomics has been hampered by uncertainty about the robustness of the features extracted from the images. Previous studies have reported that radiomics features are sensitive to changes in voxel size resampling and interpolation, image perturbation, or slice thickness. This study aims to observe the variability of positron emission tomography (PET) radiomics features under the impact of co-registration with magnetic resonance imaging (MRI) using the difference percentage coefficient, and the Spearman’s correlation coefficient for three groups of images: (i) original PET, (ii) PET after co-registration with T1-weighted MRI and (iii) PET after co-registration with FLAIR MRI. Specifically, seventeen patients with brain cancers undergoing [11C]-Methionine PET were considered. Successively, PET images were co-registered with MRI sequences and 107 features were extracted for each mentioned group of images. The variability analysis revealed that shape features, first-order features and two subgroups of higher-order features possessed a good robustness, unlike the remaining groups of features, which showed large differences in the difference percentage coefficient. Furthermore, using the Spearman’s correlation coefficient, approximately 40% of the selected features differed from the three mentioned groups of images. This is an important consideration for users conducting radiomics studies with image co-registration constraints to avoid errors in cancer diagnosis, prognosis, and clinical outcome prediction.

Cellular composition and molecular signatures of the glioma core compared with infiltrative margins are different, and it is well known that the tumor edge is enriched in microglia. In this review of the literature, we summarize the role of the peritumoral area in high-grade gliomas (HGGs) from surgical and biological points of view. There is evidence on the dual role of microglia in HGGs—a scavenger-tumoricidal role when microglia are activated in an M1 phenotype and a role favoring tumor growth and infiltration/migration when microglia are activated in an M2 phenotype. Microglia polarization is mediated by complex pathways involving cross-talk with glioma cells. In this scenario, extracellular vesicles and their miRNA cargo seem to play a central role. The switch to a specific phenotype correlates with prognosis and the pathological assessment of a specific microglial setting can predict a patient’s outcome. Some authors have designed an engineered microglial cell as a biologically active vehicle for the delivery of intraoperative near-infrared fluorescent dye with the aim of helping surgeons detect peritumoral infiltrated areas during resection. Furthermore, the pharmacological modulation of microglia-glioma cross-talk paves the way to more effective therapies.

Background/Aim: Nowadays, Machine Learning (ML) algorithms have demonstrated remarkable progress in image-recognition tasks and could be useful for the new concept of precision medicine in order to help physicians in the choice of therapeutic strategies for brain tumours. Previous data suggest that, in the central nervous system (CNS) tumours, amino acid PET may more accurately demarcate the active disease than paramagnetic enhanced MRI, which is currently the standard method of evaluation in brain tumours and helps in the assessment of disease grading, as a fundamental basis for proper clinical patient management. The aim of this study is to evaluate the feasibility of ML on 11[C]-MET PET/CT scan images and to propose a radiomics workflow using a machine-learning method to create a predictive model capable of discriminating between low-grade and high-grade CNS tumours. Materials and Methods: In this retrospective study, fifty-six patients affected by a primary brain tumour who underwent 11[C]-MET PET/CT were selected from January 2016 to December 2019. Pathological examination was available in all patients to confirm the diagnosis and grading of disease. PET/CT acquisition was performed after 10 min from the administration of 11C-Methionine (401–610 MBq) for a time acquisition of 15 min. 11[C]-MET PET/CT images were acquired using two scanners (24 patients on a Siemens scan and 32 patients on a GE scan). Then, LIFEx software was used to delineate brain tumours using two different semi-automatic and user-independent segmentation approaches and to extract 44 radiomics features for each segmentation. A novel mixed descriptive-inferential sequential approach was used to identify a subset of relevant features that correlate with the grading of disease confirmed by pathological examination and clinical outcome. Finally, a machine learning model based on discriminant analysis was used in the evaluation of grading prediction (low grade CNS vs. high-grade CNS) of 11[C]-MET PET/CT. Results: The proposed machine learning model based on (i) two semi-automatic and user-independent segmentation processes, (ii) an innovative feature selection and reduction process, and (iii) the discriminant analysis, showed good performance in the prediction of tumour grade when the volumetric segmentation was used for feature extraction. In this case, the proposed model obtained an accuracy of ~85% (AUC ~79%) in the subgroup of patients who underwent Siemens tomography scans, of 80.51% (AUC 65.73%) in patients who underwent GE tomography scans, and of 70.31% (AUC 64.13%) in the whole patients’ dataset (Siemens and GE scans). Conclusions: This preliminary study on the use of an ML model demonstrated to be feasible and able to select radiomics features of 11[C]-MET PET with potential value in prediction of grading of disease. Further studies are needed to improve radiomics algorithms to personalize predictive and prognostic models and potentially support the medical decision process.