The selection of research objects and their characteristics
The processes of identifying and selecting studies are presented in Figure 1. Through the preliminary search, a total of 153 studies were screened out. By reading the titles and abstracts of these articles, the repetitive or irrelevant studies were excluded. After a detailed review of the remaining 16 studies that might meet the inclusion criteria, 5 articles with 781 patients were finally included in the meta-analysis. OS was documented in all studies. According to the NOS literature quality assessment scale, the quality score of all the studies was 7, indicating that the quality of the studies was relatively high (Table 1).
Association of PKM2 expression with OS
PKM2 has been reported as a key molecule in the metastasis of cancer, with it being over-expressed in various cancer tissues in comparison with paired normal adjacent tissue (NAT). However, there has been a lack of summary of different studies on PKM2 in ESCC that can provide the reader with extensive information on the clinical impact of PKM2 in ESCC. Consequently, the expression of PKM2 in ESCC was meta-analyzed by us. Eventually, the results showed that over-expression of PKM2 was associated with poor prognosis in patients with ESCC (HR=1.72, 95%CI: 1.41~2.09, P<0.01) (Figure 2, A).
Association of PKM2 expression with clinicopathological features
PKM2 positivity and over-expression were significantly different between ESCC and its paired normal controls (OR=21.18, 95%CI: 6.46~69.47, P<0.01) and significantly different between lymph node metastasis and non-metastasis (OR=2.38, 95%CI: 1.68~3.35, P<0.01) (Figure 2, B and C). PKM2 positivity and over-expression were significantly associated with clinical stage I/II and clinical stage III/IV (OR=3.29, 95%CI: 2.27~4.77, P<0.01) and significantly associated with T classification (OR=2.92, 95%CI: 2.05~4.16, P<0.01) (Figure 2, D and E). But PKM2 positivity and over-expression were not significantly associated with differentiated degree (OR=1.40, 95%CI: 0.79~2.48, P=0.25) (Figure 2, F). Together, these data indicate that PKM2 over-expression could significantly promote lymph node metastasis, clinical stage and T classification in tissues of ESCC.
Heterogeneity analysis
Heterogeneity among the studies was analyzed by χ2 test and I2 test, and heterogeneity was found in correlation analysis of PKM2 expression between ESCC and NAT (P<0.05, I2=86%) and degree of differentiation (P<0.05, I2=63%). Therefore, the random effects model was used to analyze PKM2 expression between ESCC and NAT and differentiated degree, and fixed effects model was used for other correlation analysis.
Publication bias
Begg’s funnel plots were performed to access the publication bias of the literatures. The sharps of the funnel plots did not reveal any evidence of obvious asymmetry (Figure 3). For the accuracy of the results, Egger's test was used to further evaluate publication bias and the results showed that there was no publication bias among the studies(P>0.05, Table 2).