RESUMEN
INTRODUCTION: Our primary objective was to test the hypothesis that human epididymal protein 4 (HE4) and risk of ovarian malignancy index outperform the CA 125 and risk of malignancy index tests in categorizing a pelvic mass into high or low risk of malignancy in a Swedish population. Furthermore, cut-off values needed to be defined for HE4 and ROMA in premenopausal and postmenopausal women prior to their introduction to clinical practice. A third objective was to investigate the correlation between HE4 levels in serum and urine. MATERIAL AND METHODS: Women with a pelvic mass scheduled for surgery were recruited from nine hospitals in south-east Sweden. Preoperative blood samples were taken for analyzing CA125 and HE4 as well as urine samples for analyzing HE4. RESULTS: We enrolled a total of 901 women, of whom 784 were evaluable. In the premenopausal and postmenopausal groups, no significant differences were found for sensitivity, positive and negative predictive value, either for RMI vs ROMA or for CA125 vs HE4 using a fixed specificity of 75%. Cut-off values indicating malignancy were established for HE4 and ROMA in premenopausal and postmenopausal women. We found no correlation between HE4 concentration in serum and urine. CONCLUSIONS: We could not confirm that ROMA had diagnostic superiority over RMI in categorizing women with a pelvic mass into low-risk or high-risk groups for malignancy in a Swedish population. We have defined cut-off values for HE4 and ROMA. The lack of correlation between serum and urine HE4 obviates the introduction of urine HE4 analysis in clinical diagnostics.