Abstract
BACKGROUND: Metallothioneins (MTs) were reported to be associated with many kinds of tumors' prognosis, although MTs expression varied greatly among tumors. To assess the prognostic value of Metallothioneins (MTs) in different kinds of tumors, comprehensive literature search was conducted to perform a meta-analysis. METHODS: Eligible studies were identified by PubMed, MEDLINE, Web of Science (WOS), the Cochrane Library of Systematic Reviews, EMBASE, China National Knowledge Infrastructure (CNKI), WANFANG database and SinoMed database up to December 2017, which was designed to assess the prognostic value of MTs in different kinds of tumors. The main endpoint events were overall survival (OS) and disease-free survival (DFS). Hazard ratios (HRs) and its variance were retrieved from the original studies directly or calculated using Engauge Digitizer version 4.1. Random or fixed effects model meta-analysis was employed depending on the heterogeneity. Publication bias was evaluated by funnel plots, Begg and Egger tests. RESULTS: A total of 22 studies were enrolled in this meta-analysis, including 2843 tumor tissues (1517 were MTs negative/low, and 1326 were MTs high). Results showed that there was significant association between MTs expression and tumors' OS (HR = 1.60; 95%CI 1.34∼1.92, P < .00001). Subgroup analysis showed that high level of MTs expression was associated with prolonged OS in liver cancer (HR = 0.65, 95%CI 0.48∼0.89, P = .007), but it was on the contrary in the tumor of ovary (HR = 1.47, 95%CI 1.01∼2.14, P = .04), bladder (HR = 1.71, 95%CI 1.21∼2.42, P = .002), intestine (HR = 3.13, 95%CI 1.97∼4.97, P < .00001), kidney (HR = 3.31, 95%CI 1.61∼6.79, P = .001). However, there was no significant association between MTs expression and OS in breast (HR = 1.02, 95%CI 0.69∼1.51, P = .93). CONCLUSIONS: MTs could be taken as a potential prognostic biomarker for tumors, and uniqueness of MTs prognostic value in liver cancer deserved further study.