Objective Human epididymis factor-4 (HE4) has been studied in the field of ovarian cancer with promising results during the last decade. The purpose of the present meta-analysis is to investigate the diagnostic accuracy of HE4, in terms of sensitivity and specificity, as a minimally invasive method to predict optimal debulking in women that experience ovarian cancer.

Methods We used the Medline (1966–2017), Scopus (2004–2017), EMBASE, ClinicalTrials.gov (2008–2017) and Cochrane Central Register of Controlled Trials CENTRAL (1999–2017) databases in our primary search, along with the reference lists of electronically retrieved full-text articles.

Results Eleven articles were finally retrieved that enrolled 1065 women with ovarian cancer. Three of those were excluded because data were insufficient to introduce them in the meta-analysis. The sensitivity of HE4 for the prediction of optimal debulking was 0.81 (95% confidence interval [CI] 0.74–0.86) and the pooled specificity 0.80 (95% CI 0.75–0.84). The diagnostic odds ratio was 13.88 (95% CI 7.18–26.84) and the area under the curve, 0.86±0.03. After selecting only studies that enrolled patients with advanced disease who had primary debulking surgery, we observed that the pooled sensitivity reached 0.81 (95% CI 0.72–0.88), and the pooled specificity 0.70 (95% CI 0.62–0.78), with an area under the curve of 0.88±0.04.

Conclusions Human epididymis factor-4 seems to be promising as a minimally invasive method to predict optimal cytoreduction among patients with ovarian cancer. However, interpretation of our findings should be approached with caution because of the significant heterogeneity of included studies.