Long Non-Coding RNAs as Diagnostic Biomarkers for Ischemic Stroke: A Systematic Review and Meta-Analysis

Abstract

Ischemic stroke is a serious cerebrovascular disease, highlighting the urgent need for reliable biomarkers for early diagnosis. Recent reports suggest that long non-coding RNAs (lncRNAs) can be potential biomarkers for ischemic stroke. Therefore, our study seeks to investigate the potential diagnostic value of lncRNAs for ischemic stroke by analyzing existing research. A comprehensive literature search was conducted across the PubMed, ScienceDirect, Wiley Online Library, and Web of Science databases for articles published up to July 10, 2024. Statistical analyses were performed using Stata 17.0 software to calculate pooled sensitivity, specificity, positive likelihood ratio (PLR), diagnostic odds ratio (DOR), negative likelihood ratio (NLR), and area under the curve (AUC). Heterogeneity was explored with the Cochran-Q test and the I² statistical test, and publication bias was assessed with Deeks' funnel plot. A total of 44 articles were included, involving 4302 ischemic stroke patients and 3725 healthy controls. Results demonstrated that lncRNAs H19, GAS5, PVT1, TUG1, and MALAT1 exhibited consistent trends across multiple studies. The pooled sensitivity of lncRNAs in the diagnosis of ischemic stroke was 79% (95% CI: 73-84%), specificity was 88% (95% CI: 77-94%), PLR was 6.63 (95% CI: 3.11-14.15), NLR was 0.23 (95% CI: 0.16-0.33), DOR was 28.5 (95% CI: 9.88-82.21), and AUC was 0.88 (95% CI: 0.85-0.90). Furthermore, the results of subgroup analysis indicated that lncRNA H19 had superior diagnostic performance. LncRNAs demonstrated strong diagnostic accuracy in distinguishing ischemic stroke patients from healthy controls, underscoring their potential as reliable biomarkers. Because most of the articles included in this study originate from China, large-scale, high-quality, multi-country prospective studies are required to further validate the reliability of lncRNAs as biomarkers for ischemic stroke.

Keywords: LncRNAs; biomarker; epigenetic; ischemic stroke; meta-analysis.

Figure 1

Flow diagram of study search and selection.

Figure 2

Risk of bias assessment of eligible studies using QUADAS-2. (A) Summary of bias risk items in the QUADAS-2 quality assessment. (B) Percentile of risk of bias in the QUADAS-2 quality assessment [18,19,21,25,28,30,32,38,41,43,51,52,53,56,57,58,59].

Figure 3

(A) Forest plot showing the pooled sensitivity and specificity of lncRNAs in diagnosing ischemic stroke. Squares represent individual studies, while line segments indicate the 95% confidence interval (CI) for each study. The center of the diamond and the red dashed line represent the pooled effect size, and the width of the diamond corresponds to the 95% CI of the pooled results. (B) Summary receiver operating characteristic (SROC) curve with the 95% confidence and prediction contours. The Y-axis represents sensitivity, and the X-axis represents specificity. Numbers represent individual studies, and the curves depict combined diagnostic performance.

Figure 4

Fagan nomogram (A) and likelihood ratio scattergram (B) are illustrated. (A) If two values are known, the nomogram can be used to calculate a third value. (B) The ordinate represents the positive likelihood ratio, indicating the likelihood of a positive result in a patient compared to a non-patient. The abscissa represents the negative likelihood ratio, indicating the likelihood of a negative result in a patient compared to a non-patient.

Figure 5

Deeks’ funnel plot for publication bias analysis. A p-value > 0.05 indicates no significant publication bias.

Figure 6

LncRNAs as diagnostic markers for ischemic stroke. Created using https://BioRender.com (accessed on 3 December 2024).