RESUMO
INTRODUCTION: The value of serum human epididymis protein 4 (HE4) in guiding referral decisions in patients with an ovarian mass remains unclear, because the majority of studies investigating HE4 were performed in oncology hospitals. However, the decision to refer is made at general hospitals with a low ovarian cancer prevalence. We assessed accuracies of HE4 in differentiating benign or borderline from malignant tumors in patients presenting with an ovarian mass at general hospitals. METHOD: Patients with an ovarian mass were prospectively included between 2017 and 2021 in nine general hospitals. HE4 and CA125 were preoperatively measured and the risk of malignancy index (RMI) was calculated. Histological diagnosis was the reference standard. RESULTS: We included 316 patients, of whom 195 had a benign, 39 had a borderline and 82 had a malignant ovarian mass. HE4 had the highest AUC of 0.80 (95%CI 0.74-0.86), followed by RMI (0.71, 95%CI 0.64-0.78) and CA125 (0.69, 95%CI 0.62-0.75). Clinical setting significantly influenced biomarker performances. Applying age-dependent cut-off values for HE4 resulted in a better performance than one cut-off. Addition of HE4 to RMI resulted in a 32% decrease of unnecessary referred patients, while the number of correctly referred patients remained the same. CONCLUSION: HE4 is superior to RMI in predicting malignancy in patients with an ovarian mass from general hospitals. The addition of HE4 to the RMI improved HE4 alone. Although, there is still room for improvement, HE4 can guide referral decisions in patients with an ovarian mass to an oncology hospital.
