循环miRNAs检测对乳腺癌诊断价值的meta分析  

Circulating miRNAs in the Diagnosis of Breast Cancer：A Meta-Analysis

作　　者：胡道军[1] 秦兵[1] 郁淼[1] 沈宝岩 张莉[1] 

机构地区：[1]上海交通大学医学院附属新华医院崇明分院检验科,上海202150

出　　处：《现代检验医学杂志》2014年第5期93-96,共4页Journal of Modern Laboratory Medicine

摘　　要：目的：评价循环miRNAs对乳腺癌的诊断价值。方法检索 PubMed，Embase和 Cochrane外文数据库以及中国知网、维普、万方等中文数据库2014年6月之前发表的相关文献，12篇参考文献纳入研究，采用 Meta-Disc1.4软件统计，随机效应模型计算合并敏感度、特异度、阳/阴性似然比和诊断比值比，以sROC曲线法分析miRNAs总体诊断性能。采用stata12.0软件分析现有的参考文献是否存在发表偏倚。结果循环 miRNAs诊断乳腺癌的合并灵敏度是80％（95％CI 0.77-0.82），合并特异度是78％（95％CI 0.75-0.81）；合并阳性似然比是4.09（95％CI 2.80-5.99），合并阴性似然比是0.22（95％CI 0.15-0.31）；合并诊断比值比是20.64（95％CI 10.24-41.62）；sROC曲线下面积（AUC）为0.89，Q值为0.82。已有的研究存在发表偏倚。结论循环miRNAs可作为一项潜在的生物标记物用于乳腺癌的诊断。Objective To assess circulating miRNAs’diagnosis value of breast cancer.Methods Conducted PubMed,Embase, the Cochrane Library,CNKI,cqvip,and wangfang database search in any language before June 2014.A total of 12 publica-tions were included in the meta-analysis.Then,the meta-analysis was performed using Meta-Disc 1.4.Meanwhile,the diag-nostic sensitivity,specitivity,positive likelyhood ratio,negative likelyhood ratio and diagnostic odds ratio were pooled by ran-dom-effects models.And the overall diagnostic performance was estimated by summary receiver operating characteristic curves （sROC）approaches.Stata12.0 was used to evaluated the publication bias.Results The pooled sensitivity was 80%（95%CI 0.77-0.82）,specificity was 78% （95%CI 0.75-0.81）;positive likelihood ratio was 4.09 （95%CI 2.80-5.99）, negative likelihood ratio was 0.22 （95%CI 0.15-0.31）,diagnostic odds rations was 20.64 （95%CI 10.24-41.62）.The AUC for circulating miRNAs was 0.89 with Q value of 0.82.Publication bias was observed in existing literatures.Conclu-sion Circulating miRNAs is a potential biomarker in the diagnosis of breast cancer.

关 键 词：乳腺癌 META分析 

分 类 号：R737.9[医药卫生—肿瘤]