Objective At present, considerable efforts have been made to identify new cancer-specific markers for ovarian cancer (OC) diagnosis and the kallikrein-related peptidases (KLKs) family is one of the most studied candidates. This meta-analysis aims to evaluate the pooled diagnostic value of serum KLK measurement for diagnosing OC.

Methods The Cochrane Library, PubMed, Excerpt Medica Database were searched for all relevant literature. The Quality Assessment for Studies of Diagnostic Accuracy tool was applied to assess the quality of enrolled studies. Statistical analysis was conducted by using Stata 13.0 software and Meta-Disc.

Results A total of 15 studies from 13 articles were considered eligible for inclusion in the present analysis. The following pooled parameters were calculated by using the bivariate model: sensitivity of 0.582 (95% confidence interval [CI], 0.517–0.644), specificity of 0.909 (95% CI, 0.833–0.952), positive likelihood ratios of 6.367 (95% CI, 3.330–12.172), negative likelihood ratios of 0.460 (95% CI, 0.388–0.546), diagnostic odds ratio of 13.831 (95% CI, 6.460–29.614), respectively.

Conclusions Kallikrein-related peptidase seems to be a promising candidate biomarker in diagnosing OC, but the associated poor sensitivity of KLK individually may limit its value in clinical application. To resolve this problem, the combination of KLK and other markers may offer improved performance than a single marker.