Logistic models with five serum biomarkers for preoperative differentiation of benign ovarian tumors from ovarian carcinoma

Abstract

Ovarian carcinoma (OC) is one of the leading causes of mortality in women with oncological diseases of the reproductive system. The disease is often diagnosed at an advanced stage due to the lack of characteristic symptoms and effective screening strategies. Early differentiation of benign ovarian tumors (BOT) from OC is of key importance for choosing the optimal surgical tactics and improving the long-term prognosis. Classical diagnostic methods, including imaging studies and determination of tumor markers such as CA125, often do not provide sufficient sensitivity and specificity. Therefore, multivariate models are being developed that combine biomarkers and clinical parameters for a more accurate risk assessment. By determining the concentrations of five serum proteins – transthyretin, transferrin, apolipoprotein А1, β2 microglobulin and CA125 and monitoring their diagnostic abilities using logistic regression, we created formulas that can be implemented in a computer program. Their application after a gynecological examination and the detection of an ovarian tumor would be useful in preoperative differential diagnosis.