Neurosurgery 98:1221–1230, 2026
This systematic review and meta-analysis evaluates microRNA (miRNA)–based liquid biopsies for diagnosing glioblastoma (GBM), synthesizing data from 15 studies and 28 biomarker evaluations across 868 samples. Key diagnostic metrics—pooled sensitivity 0.84 and specificity 0.89—indicate strong potential, with miR-21 showing the highest accuracy among single-miRNA assays.
The report details search strategy, inclusion criteria, statistical methods, subgroup analyses (miRNA type, biofluid source, control type), and study quality assessment. Limitations include methodological heterogeneity, high risk of bias in patient selection, inconsistent reporting (CSF source, IDH status), and small cohort sizes, underscoring need for standardized clinical validation.
Aim Evaluate the diagnostic accuracy of microRNA (miRNA) liquid biopsies (blood/serum/plasma/CSF) for diagnosing glioblastoma (GBM) via systematic review and meta-analysis.
Methods PRISMA-guided searches of Ovid Medline and Embase (updated through Oct 16, 2024); included studies had histologic GBM confirmation and extractable 2×2 diagnostic data; pooled estimates generated using a random-effects bivariate model.
Evidence base 15 included articles (published 2011–2022) provided 28 miRNA evaluations, totaling 868 samples from 551 GBM patients and 811 samples from 578 controls.
Overall accuracy Pooled sensitivity 0.84 and specificity 0.89, with heterogeneity of 66% (sensitivity) and 39% (specificity); pooled AUC 0.89.
Likelihood ratios Pooled PLR 7.26, NLR 0.19, and DOR 40.17, indicating strong overall discriminatory performance.
Key biomarker miR-21 showed the highest pooled performance among assessed groupings (sensitivity 0.90, specificity 0.95).
Subgroups Single miRNAs had higher specificity than multi-miRNA panels, while diagnostic capability did not differ clearly by biofluid source (CSF vs blood) in this dataset.
Limitations All included studies had high risk of bias in patient selection, and many had bias in index test interpretation/flow-timing; limited CSF comparisons and inconsistent qPCR thresholds/normalization contributed to heterogeneity and constrain clinical translation without standardization and validation.
















You must be logged in to post a comment.