Diagnostic potential of circulating LncRNAs in human cardiovascular disease: a meta-analysis

Cardiovascular disease (CVD) is a major killer of the human population around the world. Identifying effective diagnostic biomarkers for CVDs is particularly important in order to guide optimizing treatment. Accumulating evidence on aberrantly regulated circulating long non-coding RNAs (LncRNAs) promise to serve as a diagnostic or prognostic biomarker for various types of CVDs. We summarized studies to identify the potential diagnostic values of LncRNAs in CVD patients. We included articles reporting on the association between LncRNAs and diagnosis in CVDs. We calculated sensitivities, specificities, and area under the curves of LncRNAs. The pooled overall sensitivity and specificity for LncRNAs expression profile in differentiating CVD patients from controls (non-CVDs or healthy subjects) were 0.74 (95%CI 0.68-0.80) and 0.81 (95%CI 0.76-0.85), respectively; the overall positive likelihood ratio, 3.9 (95%CI 3.1-4.9); the negative likelihood ratio, 0.32 (95%CI 0.25-0.40); corresponding to an area under curve of 0.85 (95%CI 0.82-0.88) and overall diagnostic odds ratio 12 (95%CI 9-18). Subgroup analysis showed that the detection of LncRNAs expression in plasma substantially improved the diagnostic accuracy. Likewise, meta-regression analysis indicated that the detection method and sample size were the main source of heterogeneity. All these results suggested a relatively good reference value of LncRNAs as auxiliary biomarkers for CVDs, and should be considered in cases where the diagnosis is uncertain. Population-based prospective cohort studies are warranted to confirm our findings.