BackgroundWe aim to determine the diagnostic accuracy of both serum and urinary human epididymis protein 4 (HE4) in the diagnosis of ovarian cancer.
MethodsElectronic databases and search engines such as PubMed, Cochrane Library, and Google Scholar were searched systematically by two independent reviewers to retrieve articles published from inception to June 11, 2022. The diagnostic accuracy of serum and urinary HE4 was computed using the random-effects model in terms of pooled sensitivity, pooled specificity, and diagnostic odds ratio (DOR) with 95% confidence interval (CI). To explain any source of possible heterogeneity, meta-regression and subgroup analyses were performed. Risk of bias assessment was conducted using the Quality Assessment of Diagnostic Accuracy Studies 2 (QUADAS-2) tools recommended by the Cochrane Library.
Result and conclusionThis meta-analysis included a total of 38 studies of serum HE4 involving 14,745 subjects and five studies for urinary HE4 involving 649 subjects. We observed acceptable pooled sensitivity, specificity, summary receiver operating characteristics (SROC), and diagnostic odds ratio (DOR) at 0.79 (95% CI: 0.75-0.82), 0.92 (95% CI: 0.87-0.95), 0.88 (95% CI: 0.85-0.91), and 43 (95% CI: 25-72), respectively, for serum HE4 for discriminating ovarian cancer. For urine HE4, the pooled sensitivity, specificity, SROC, and DOR were 0.80 (95% CI: 0.64-0.90), 0.93 (95% CI: 0.83-0.98), 0.94 (95% CI: 0.91-0.95), and 55 (95% CI: 15-198), respectively. Therefore, HE4 is a promising biomarker with a high degree of specificity and acceptable sensitivity for the diagnosis of ovarian cancer.
Registration numberThis meta-analysis was performed after the registration of the protocol in the PROSPERO database with registration number CRD42022324947.