Identification of circulating microRNAs as biomarkers in diagnosis of hematologic cancers: a meta-analysis

Recent studies have provided new insights into the diagnostic value of circulating microRNAs (miRNAs) for hematologic cancers. However, inconsistent results have been reported on the diagnostic performance of various kinds of miRNAs. To systematically assess the potential diagnostic value of miRNAs in hematologic cancers, we conducted the present meta-analysis. Multiple databases (PubMed, Cochrane Library, EMBASE, CNKI, and Wan Fang) were carefully searched for available studies up to April 4, 2014. Sensitivity and specificity were pooled using a random-effects model. Likelihood ratio (LR), diagnostic odds ratio (DOR), and the area under the curve (AUC) were used to measure the diagnostic values. Subgroup and meta-regression analyses were used to find potential sources of heterogeneity. Thirty-four studies from 14 publications, which involved 1,159 hematologic cancer patients and 826 healthy controls, were included in this meta-analysis. The pooled estimates indicated a moderately high diagnostic accuracy for circulating miRNAs, with a sensitivity of 0.83, a specificity of 0.85, a PLR of 5.7, a NLR of 0.20, a DOR of 29, and an AUC of 0.91. The subgroup analyses showed that diagnostic accuracy was better for acute myeloid leukemia (AML) patients and Asians compared with other subgroups. In addition, multiple miRNA assays displayed a better performance than single ones. Furthermore, we found that plasma might be a more promising matrix for detecting the expression of miRNAs than serum. Our results identified the potential use of circulating miRNAs in second-line diagnosis for hematologic cancers, especially the value of miRNA panels. However, further large cohort studies are still required to confirm our findings.