Circulating long non-coding RNAs as novel diagnostic biomarkers for Alzheimer’s disease (AD): A systematic review and meta-analysis

BACKGROUND: Long non-coding RNAs (lncRNAs) have been reported to be involved in the pathogenesis of neurodegenerative diseases. It has also been hypothesized that plasma exosomal lncRNAs may be used as Alzheimer’s disease (AD) biomarkers. In this systematic review, we compiled all studies on the subject to evaluate the accuracy of lncRNAs in identifying AD cases through meta-analysis. METHODS: A PRISMA-compliant systematic search was conducted in PubMed/MEDLINE, EMBASE, and Web of Science databases for English publications till September 2022. We included all observational studies published which investigated the sensitivity and specificity of various lncRNAs in plasma samples of AD diagnosis. Our search strategy included lncRNA and all the related spelling and abbreviation variations combined with the keyword Alzheimer’s disease. Methodological quality was assessed using the Strengthening the Reporting of Observational Studies in Epidemiology (STROBE) guidelines and the Quality Assessment of Diagnostic Accuracy Studies (QUADAS-II) tool. The meta-analysis was carried out using the area under the Receiver Operator Characteristic (ROC) curves (AUC) and sensitivity and specificity values to assess the accuracy of the identified lncRNAs in AD diagnosis. To account for the predicted heterogeneity of the study, a random-effects model was used. All the statistical analyses and visualizations were conducted using Stata 17.0 software. RESULTS: A total of seven studies (AD patients = 553, healthy controls = 513) were included in the meta-analysis. Three lncRNAs were upregulated (RNA BACE-AS1, RNA NEAT1, RNA GAS5), and one lncRNA (MALAT1) was downregulated in plasma samples of AD patients. RNA 51A and RNA BC200 were reported to have variable expression patterns. A lncRNA (RNA 17A) was not significantly different between AD and control groups. The pooled sensitivity, specificity, and AUC values of lncRNAs in identifying AD were (0.74; 95% CI [0.63, 0.82], I2 = 79.2%), (0.88; 95% CI [0.75, 0.94], I2 = 88.9%), and 0.86; 95% CI [0.82, 0.88], respectively. In addition, the pooled diagnostic odds ratio (DOR) of the five individual lncRNAs in AD diagnosis was 20. CONCLUSION: lncRNAs had high accuracy in identifying AD and must be seen as a promising diagnostic biomarker of the disease.