A restricted signature of serum miRNAs distinguishes glioblastoma from lower grade gliomas

Giulia Regazzo, Irene Terrenato, Manuela Spagnuolo, Mariantonia Carosi, Gaetana Cognetti, Lucia Cicchillitti, Francesca Sperati, Veronica Villani, Carmine Carapella, Giulia Piaggio, Andrea Pelosi, Maria Giulia Rizzo

Malignant gliomas are the most common primary brain tumors in adults and challenging cancers for diagnosis and treatment. They remain a disease for which non-invasive, diagnostic and/or prognostic novel biomarkers are highly desirable. Altered microRNA (miRNA) profiles have been observed in tumor tissues and biological fluids. To date only a small set of circulating/serum miRNA is found to be differentially expressed in brain tumors compared to normal controls. Here a restricted signature of circulating/serum miRNA including miR-15b*,-23a, -99a, -125b, -133a, -150*, -197, -340, -497, -548b-5p and let-7c were investigated as potential non-invasive biomarkers in the diagnosis of glioma patients. Serum and tissues miRNAs expression in patients with brain cancers (n = 30) and healthy controls (n = 15) were detected by quantitative real-time polymerase chain reaction (qRT-PCR). Relative expression was calculated using the comparative Ct method. Statistical significance (p ≤ 0,05) was determined using the Mann-Whitney rank sum and Fisher's exact test. Diagnostic accuracy of miRNAs in distinguishing glioblastoma multiforme (GBM) from lower grade cancer was assessed by the Receiver Operating Characteristic (ROC) curve analysis. To validate the role of the identified miRNAs in cancer a comprehensive literature search was conducted using PubMed, Web of Science (Core Collection) and Scopus databases. We observed a decrease of miR-497 and miR-125b serum levels depending on tumor stages with reduced level in GBM than lower grade tumors. The ROC curve analysis distinguishing GBM from lower grade cases yielded an area under the curve (AUC) of 0.87 (95 % confidence interval (CI) = 0.712-1) and of 0.75 (95 % CI = 0.533-0.967) for miR-497 and -125b, respectively. GBM patients are more likely to show a miR-497 and -125b down-regulation than the lower grade group (p = 0.002 and p = 0.024, respectively). These results were subsequently compared with evidence from 19 studies included in the final systematic review. Although multiple biomarkers are currently leveraged in the clinic to detect specific cancer types, no such standard blood biomolecules are used as yet in gliomas. Our data suggest that serum miR-497 and -125b could be a novel diagnostic markers with good perspectives for future clinical applications in patients with glioma.