Circulating miRNAs as Diagnostic and Prognostic Biomarkers in High-Grade Gliomas

Abstract

Objectives: miR-181a/b and miR-410 downregulation and miR-155 upregulation has been shown to play important roles in the oncogenesis and progression of gliomas including high-grade gliomas. However, the potential role of plasma miR-181a/b, miR-410 and miR-155 in the diagnosis and prognosis of high-grade gliomas remains poorly known.

Methods: We retrieved published articles from the PubMed, the Cochrane Central Register of Controlled Trials, and Web of Science database and obtained different sets of data on microRNAs (miRNAs) expression profiling in glioma and highlighted the most frequently dysregulated miRNAs and their gene-targets (PDCD4, WNT5A, MET, and EGFR) in high-grade gliomas. Quantitative reverse transcription polymerase chain reaction (qRT-PCR) was carried out to measure the pre- and postoperative plasma levels of miR-181a/b, miR-410 and miR-155 in 114 Grade 3-4 glioma patients, 77 Grade 1-2 glioma patients and 85 healthy volunteers as control group. The diagnostic and prognostic value of circulating miR-181a/b, miR-410 and miR-155 as biomarker was estimated by the Receiver Operating Characteristic (ROC) curve and the area under the curve (AUC) and Kaplan-Meier analysis.

Results: We found a plasma miRNA signature including three downexpressed miRNAs and one overexpressed (miR-181a, miR-181b and miR-410; miR-155) in high-grade glioma patients in comparison with low-grade glioma patients control group. The ROC curve AUC of these four circulating miRNAs were ≥ 0.75 for high-grade glioma patients in before and after surgery. Higher circulating miR-155 and lower miR-181a/b and miR-410 expression is associated with clinical data, clinic pathological variables, worse overall survival (OS) of patients and negative correlated with potential gene-targets expression. Moreover, Kaplan-Meier analysis showed that miR-181a/b, miR-410 and miR-155 were independent predictors of OS in high-grade glioma patients.

Conclusions: Our data, for the first time, demonstrated that circulating miR-181a/b, miR-410 and miR-155 could be a useful diagnostic and prognostic non-invasive biomarkers in high-grade gliomas.

Keywords: biomarker; circulating; diagnosis; high-grade glioma; miRNA; prognosis.

Figure 1

Cases of low-grade glioma patients (A, B). (A) A large intracerebral tumor of the right temporo-occipital areas with a heterogeneous cystic-solid structure, contrasted as a group of nodules and small closed rings in the solid part. The contents of the cysts are similar to the cerebrospinal fluid in terms of signal characteristics, with the exception of a cyst in the pole of the temporal lobe, where a protein admixture and traces of hemorrhage are detected, the walls of the cysts do not increase. Diffusion restriction in the structure of education is not determined. In the posterior sections, a node with elevated median cerebral blood volume (CBV) values according to multi-slice computed tomography (MSCT) perfusion. (B) Predominantly in the middle frontal gyrus of the left frontal lobe, an intracerebral mass with diffuse distribution, clear contours and a sign of T2/fluid attenuation inversion recovery (FLAIR) mismatch. The tumor involves white and gray matter with “swelling” of the cortical plate. There is no restriction of diffusion, contrasting in the formation. Reliable areas of hyperperfusion according to dynamic susceptibility contrast (DSC) MR perfusion are not determined.

Figure 2

Cases of high-grade glioma patients (A, B). (A) Intracranial tumor of the right frontal lobe and genu corpus callosum, in addition to the white matter, involving the cortical plate with a mass effect on the anterior horn of the right lateral ventricle. The parasagittal solid component has small foci of hyperperfusion on dynamic susceptibility contrast (DSC) MR perfusion and areas of increased cellularity in apparent diffusion coefficient (ADC), is heterogeneous in signal in T2 weighted image (T2WI), and shows heterogeneous contrast enhancement. (B) A large “intra-axial” area of in homogeneously elevated signal in T2 and fluid attenuation inversion recovery (FLAIR) in the right frontal lobe and basal ganglia, extending into the gray matter, the opposite hemisphere along the genu corpus callosum, with a group of subpial (and one subependial) nodules of homogeneous intense contrast enhancement. Against the background of uneven diffusion, zones of increased relative cerebral blood volume (rCBV) are revealed along the lateral contour of the right frontal lobe.

Figure 3

Flow diagram illustrating the steps for microRNAs (miRNAs) selection, expression profiling and differential expression analysis. qRT-PCR, quantitative reverse transcription polymerase chain reaction; miRNAs, microRNAs.

Figure 4

Circulating miR-181a/b, miR-410 and miR-155 expression levels within different grade gliomas and control group (A–D). (A) The expression levels of circulating miR-155 in the plasma of high-grade glioma patients were significantly higher than in low-grade glioma patients and control group. In the same time, (B–D) the expression levels of circulating miR-181a, miR-181b and miR-410 in the plasma of high-grade glioma patients were significantly lower than in low-grade glioma patients and control group. A probability of p < 0.05 (*) or p < 0.001 (**) was considered statistically significant.

Figure 5

Box plots showing the expression levels of (A) miR-155, (B) miR-410, (C) miR-181a and (D) miR-181b before and after 10 days surgery of patients with high-grade glioma with compare control group. A probability of p < 0.05 (*) or p < 0.001 (**) was considered statistically significant.

Figure 6

The dynamic change of circulating miR-181a/b, miR-410 and miR-155 in plasma samples of high-grade glioma patients before and after surgery. The expression levels of the 4 circulating miRNAs in the patients before and after surgery (A–D). Each point represents the mean of the triplicate samples. A probability of p < 0.05 (*) or p < 0.001 (**) was considered statistically significant.

Figure 7

Receiver Operating Characteristic (ROC) curves for circulating miR-181a/b, miR-410 and miR-155 for glioma patients before surgery (A–D). In panel (A, B) of the figure the area under the curves (AUCs) for circulating miR-155 and miR-410 were 0.68, 0.92, 0.83, 0.67, 0.97, 0.93 and in panel (C, D) for circulating miR-155 and miR-410 were 0.83, 0.97, 0.79, 0.78, 0.94, 0.78 respectively.

Figure 8

Receiver Operating Characteristic (ROC) curves for circulating miR-181a/b, miR-410 and miR-155 for high-grade glioma patients after surgery (A–D). The area under the curves (AUCs) for circulating miR-155 (A), miR-410 (B), miR-181a (C), and miR-181b (D) were 0.87, 0.76, 0.84, and 0.87, respectively, suggesting that these circulating miRNAs can distinguish high-grade glioma patients after surgery from control group.

Figure 9

Kaplan–Meier curves of overall survival (OS) for high-grade glioma patients based on the circulating miR-155, miR-410, miR-181a and miR-181b signature in patients with total (n = 71) (A–D) or partial tumor resection (n = 43) (E–H).

Figure 10

Relative expression of circulating miR-181a/b, miR-410 and miR-155 and its potential gene-targets, programmed cell death 4 (PDCD4), Wnt family member 5A (WNT5A), MET, and epidermal growth factor receptor (EGFR), in plasma samples from high-grade glioma patients (A–E). (A, B) Correlation between circulating miR-155 expression and PDCD4 and WNT5A levels in plasma samples (p < 0.0001; Spearman rank correlation test: r = 0.753, r = -0.671). (C) Correlation between miR-410 expression and MET levels in in plasma samples (p < 0.0001; Spearman rank correlation test: r = -0.4778). (D, E) Correlation between circulating miR-181a/b expression and EGFR levels in plasma samples (p < 0.0001; Spearman rank correlation test: r = -0.5235, r = -0.5217).