Literatures selection. The literature screening process is provided in Figure 1. First of all, 126 articles were retrieved from three databases, and 65 duplicate articles were excluded after screening. After that, we carefully read the titles and abstracts of the remaining 61 articles, and excluded 23 irrelevant studies articles and 10 articles that did not meet the inclusion criteria. Then, we read the full text of the remaining 28 literature and excluded 8 literature that did not have available data. Finally, 20 articles were included in this meta-analysis.
Characteristics of included studies. The characteristics of the 20 included studies in the meta-analysis were shown in Table I. The publication date of these studies involving 1226 patients was between 2016 and 2023. Three of the studies used multivariate analysis as the prognostic analysis method, and others were only provided Kaplan-Meier survival curves. The included literature covered twelve types of malignant tumors, including colorectal cancer, breast cancer, melanoma, bladder cancer, colon cancer, esophageal squamous cell carcinoma, lung caner, pancreatic ductal adenocarcinoma, gastric cancer, hepatocellular carcinoma, acute myeloid leukemia and glioma.
Relationship between DUXAP8 and prognosis. Among 20 eligible studies, 15 studies reported results of OS toward DUXAP8 expression with a total of 967 patients, two reported PFS, and two reported DFS. Therefore, OS was used as an outcome indicator for patients with different DUXAP8 expression levels. Due to heterogeneity between included studies (I2 = 41.8%, p = 0.045), a random-effects model was used. Meta-analysis showed that high DUXAP8 expression was significantly associated with poor OS (HR = 2.50, 95%CI: 1.90-3.28, p < 0.001, Figure 2).
To further analyze heterogeneity between included studies and evaluate the association between aberrant DUXAP8 expression and OS, subgroup analyses according to the sample size, the type of cancer, the morphology of cancer and the source of HRs were conducted (Table Ⅱ). In the stratified analysis by the type of cancer, high DUXAP8 expression was significantly associated with poor OS in patients with digestive system tumor (HR = 2.66, 95%CI: 2.05–3.44, p < 0.001) and non-digestive system tumor (HR = 2.50, 95%CI:1.90–3.28, p = 0.024). The similar result was also observed in the stratified analysis by the type of cancer and the source of HRs. As for the cancer morphology, high DUXAP8 expression was significantly associated with poor OS in solid tumors (HR = 2.72, 95%CI: 2.21–3.35, p < 0.001) including 14 studies reported results of OS toward DUXAP8 expression. No significant heterogeneity across studies was detected in this subgroup analysis. The forest plot of this subgroup analyses is shown in Figure S1-S4.
Associations of DUXAP8 expression with clinicopathological features. In a pooled analysis based on different clinicopathological characteristics of the included patients (Table Ⅲ), higher DUXAP8 expression was associated with poorer TNM staging (OR = 2.60, 95%CI 1.96-3.46, p < 0.001, supplement ), lymph node metastasis (OR = 2.76, 95%CI: 1.71-4.46, p = 0.044) and poorer histological grade (OR = 2.20, 95%CI: 1.42-3.40, p = 0.019). The forest plot of this process is shown in Figures S5-S9.
Publication bias and sensitivity analyses. The publication bias for DUXAP8 expression and OS was assessed using a Begg’s funnel plot. No significant publication bias was observed among those studies (Pr > |z| = 0.051, z = 1.77) (Figure 3). The results of sensitivity analysis showed that no matter which included study was removed, the pooled results would not be significantly changed (Figure 4), indicating the stability of the results.
DUXAP8 was upregulated in PCa tissues and cells. Initially, we investigated the changes in the expression of DUXAP8 in PCa samples based on TCGA database, and it showed that DUXAP8 was upregulated in PCa tissues compared with normal prostate tissues (Figure 5A-B). Additionally, we also searched the expression of DUXAP8 in PCa cell lines (LNCap, 22Rv1, PC-3 and DU145) and normal human prostate cell line RWPE-1 by qRT-PCR. As shown, the expressions of DUXAP8 were significantly increased in PCa cells compared to normal prostate cell lines (Figure 5C).
DUXAP8 was associated with poor prognosis in prostate cancer. To further explore the relationship between DUXAP8 expression level and long-term prognosis of PCa patients, Kaplan-Meier analysis and Log-rank test were performed based on TCGA patients. Results showed that patients with higher DUXAP8 levels had shorter BCR than those with lower DUXAP8 levels (Figure 5D). In addition, we also investigated the relationship between DUXAP8 level and clinicopathological characteristics of PCa patients. Chi-squared test showed that high DUXAP8 expression in PCa was significantly correlated with postoperative pathological stage (p = 0.035), Gleason score (p = 0.009), clinical T stage (p = 0.003) and lymph node metastasis (p = 0.024), but not with age and distant metastasis (Table Ⅳ).
DUXAP8 knockdown inhibited cell proliferation, invasion and migration of PCa cells. The qRT-PCR results showed that the DUXAP8 levels were significantly down-regulated, with the knockdown efficiency reaching about 70% (Figure 6A). EdU staining results demonstrated that DUXAP8 knockdown suppressed the cell proliferation of DU145 and PC-3 cell lines compared with the control group (Figure 6B-C). The wound healing rate was significantly decreased in the DUXAP8 knockdown group following wound injury (Figure 6D-E). Moreover, transwell assay demonstrated that DUXAP8 knockdown could decrease the number of invaded and migrated PCa cells (Figure 6F-G).
To further investigate the impact of DUXAP8 knockdown on tumor growth in vivo, a subcutaneous xenograft tumor model by injecting DUXAP8 stable knockdown PCa cells was used. Results from qRT-PCR showed an obvious reduction of DUXAP8 expression in sh-DUXAP8 group compared with sh-NC group (Figure 7A). As expected, the volumes and weights of tumors in the sh-DUXAP8 group were significantly smaller than those in the control group (Figure 7B-D).