Prognostic role of autophagy-related proteins in epithelial ovarian cancer: a meta-analysis of observational studies.

INTRODUCTION: Autophagy has been reported to play important roles in tumorigenesis. This meta-analysis was conducted in order to investigate the association between autophagy-related proteins, beclin 1 and LC3, with prognostic value and clinical characteristics of ovarian cancer.

EVIDENCE ACQUISITION: PubMed, EMBASE and CNKI databases (up to August 2016) were searched for eligible studies using the following key words: "beclin 1", "beclin-1", "Becn 1", "Becn-1", "Atg6", "LC3", "ovarian cancer", "ovarian carcinoma", "ovarian neoplasm", "ovarian tumor", "ovary cancer", "ovary carcinoma", "ovary neoplasm", "ovary tumor". The random effects model was used to combine effect sizes. Higgins I2 value was used to examine heterogeneity between studies. Sensitivity analysis was used to investigate the heterogeneity and the stability of the results. Egger's test and Harbord's test were used to examine publication bias.

EVIDENCE SYNTHESIS: Eight studies including a total of 872 patients were eligible for this meta-analysis. The results showed expressions of both beclin 1 (OR=0.498, 95% CI: 0.298-0.832) and LC3 (OR=0.308, 95% CI: 0.142-0.668) were associated with FIGO stage (FIGO I/II versus FIGO III/IV). However, no association was found between the expression of beclin 1 and overall survival (HR=0.678, 95% CI: 0.434-1.058), age (OR=0.965, 95% CI: 0.473-1.967), lymph node metastasis (OR=0.517, 95% CI: 0.218-1.227), pathological classification (OR=0.895, 95% CI: 0.649-1.233) and differentiation (OR=0.571, 95% CI: 0.261-1.247). Similarly, no relationship was found between LC3 with age (OR=0.947, 95% CI: 0.399-2.248), pathological classification (OR=0.821, 95% CI: 0.410-1.644), and differentiation (OR=0.432, 95% CI: 0.118-1.580).

CONCLUSIONS: Autophagy-related proteins might be important markers of ovarian cancer progression. However, due to the limited number of original studies, more well-designed, large-scale, high-quality studies are needed, which provide new, valuable information for further investigation.