Objectives There is an unmet need to improve accurate detection of malignancy in patients with pelvic masses. Our objective was to obtain summary estimates of HE4 accuracy for diagnosing malignancy and to compare performance with CA125, in different clinical settings.

Methods We searched PubMed, Ovid and Scopus using terms for ‘pelvic masses’ and ‘HE4’, to identify studies that evaluated HE4 for diagnosing malignant ovarian masses. Screening, data extraction and quality assessment were done independently by two authors. We performed meta-analysis of HE4 and CA125 accuracies using a random-effects bivariate logit-normal model.

Results In the 17 eligible studies, OC prevalence ranged from 15% to 71%. All studies seemed to have recruited patients in specialized settings. A meta-analysis of 7 HE4-studies resulted in a mean sensitivity and specificity (95% CI) of 79.4% (74.1%-83.8%) and 84.1% (79.6%-87.8%), for cut-off values of 67–72 pmol/L. Based on 8 studies, the mean sensitivity and specificity of CA125 was 81.4% (74.6%-86.2%) and 56.8% (47.9%- 65.4%), respectively, at a cut-off of 35U/mL. Given a 40% OC prevalence, the positive predictive value (PPV) for HE4 was 76.9% (71.9%-81.2%) versus 55.6% (50.2%-60.9%) for CA125. At a 15% prevalence, the negative predictive value (NPV) was 95.8% (95% CI: 94.4%-96.7%) and 94.4% (95% CI: 92.3%-96.0), respectively.

Conclusions HE4 had higher specificity and similar sensitivity compared to CA125. At high prevalence in specialized settings, PPV is higher for HE4. At low prevalence in general settings, NPV of HE4 is similar to CA125. Prevalence and setting are important variables that should always be reported in biomarker research.