Diagnostic efficacy of long non-coding RNA MALAT-1 in human cancers: a meta-analysis study

Abstract

Metastasis-associated lung adenocarcinoma transcript 1 (MALAT-1) is one kind of long non-coding RNAs (lncRNAs) that has been recognized as a hallmark of the onset and development of several carcinomas. This study seek to meta-analyze the overall diagnostic efficacy of elevated MALAT-1 expression profile for human cancers. Studies on the diagnostic performance of MALAT-1 in cancers were retrieved by searching the online databases. The combined effect sizes were summarized using a bivariate meta-analysis model. Impacts of publication bias on the pooled effect sizes were assessed using “Duval and Tweedie nonparametric trim and fill method”. Sensitivity analysis and meta-regression test were applied to deeply trace the heterogeneity sources among eligible studies. A total of 14 studies with 1342 cancer cases were included. The combined effect sizes showed that MALAT-1 expression profiling conferred an estimated sensitivity of 0.69 (95% CI: 0.62–0.75) (I² = 84.01%, P < 0.001), specificity of 0.85 (95% CI: 0.79–0.90) (I² = 87.95%, P < 0.001) and AUC (area under curve) of 0.83 in distinguishing cancer patients from noncancerous contrasts. Moreover, stratified analysis depending on cancer type manifested that elevated MALAT-1 harbored a promising efficacy in the diagnosis of pulmonary tumors (AUC = 0.90), digestive system tumors (AUC = 0.84), gynecologic cancers (AUC = 0.84) and nasopharyngeal carcinoma (AUC = 0.84), particularly in confirming the subtype of squamous carcinoma (AUC = 0.91) and non-small cell lung carcinoma (AUC = 0.88) in lung cancer. Other analyses based on test matrix and ethnicity also presented robust results. Collectively, elevated MALAT-1 could be developed as an auxiliary molecular marker to aid in cancer diagnosis.