Performance comparison of serum and urine biomarkers from independent samples for ovarian cancer screening

Abstract

This study compares the diagnostic performance of urine and serum multiple biomarkers for early diagnosis of ovarian cancer. The sample population includes 119 benign and 101 ovarian cancer patients. The marker combinations used to compare performance include 16 markers whose concentration values were obtained using the Luminex assay. In order to identify an optimal marker combination that could classify ovarian cancer and benign patients, the area under the curve (AUC) is used to evaluate 2-, 3-, and 4-marker combinations and the classification is performed by using logistic regression. In the case of urine samples, the best AUC values are 87.89% for the 2 protein markers combination, 90.22% for the 3 markers combination, and 92.43% for the 4 marker combination. In contrast, the best AUC values for serum sample are 92.4% for the 2 marker combination, 93.63% for the 3 marker combination and 94.63% for the 4 marker combination. This study confirmed that combining multiple biomarkers could improve diagnostic accuracy. Even though the urine sample shows relatively lower performance than serum, urine could be utilized more widely for its simple usability.