Combining the biomarkers macrophage inhibitory factor, osteopontin and prolactin with CA-125 improves early detection of ovarian cancer

Abstract

<h3>Objectives:</h3> To determine the combined efficacy of MIF, OPN, PROL and CA-125 in distinguishing ovarian cancer from healthy compared to CA-125 alone as a single biomarker. <h3>Methods:</h3> Serum samples from 153 ovarian cancer patients (37 Stage I/II and 116 Stage III/IV) and 279 healthy age-matched controls were utilized in this study. The ovarian cancer group (n=153, average age=57.1 yr) was composed of women with newly diagnosed ovarian cancer following discovery of pelvic mass. Mean concentration of MIF, OPN, PROL and CA-125 was measured using a novel, fully automated, multi-analyte immunoassay platform, Ella. The serum samples were randomly divided into two different cohorts (n=153 and n=279) to be interchangeably used as training and testing sets for statistical analysis. Four statistical classification models, k-nearest neighbors, logistic regression, random forest, and support vector machines were used to calculate accuracy, sensitivity, specificity, positive-predictive value (PPV), and negative-predictive value (NPV) of the four protein biomarker panel and CA-125 alone. <h3>Results:</h3> In distinguishing serum of ovarian cancer patients from healthy controls, the four-protein biomarker panel yielded an average accuracy of 91% compared to 85% when using CA-125 alone across the four classification models (p=3.224e-09). Further, the four-protein biomarker produced statistically significantly higher sensitivity (median of 76%), specificity (median of 98%), PPV (median of 91.5%), and NPV (median of 92%), compared to CA-125 alone in distinguishing ovarian cancer from healthy controls (Wilcoxon p-values: 2.22e-07, 0.02, 1.27e-2, and 3.629e-07, respectively; Figure 1). Importantly, the improved performance of the four-protein biomarker over CA-125 alone is maintained in the detection of early-stage ovarian cancer (Stage I/II) from health controls. <h3>Conclusions:</h3> Combining MIF, OPN, and PROL to CA-125 resulted in improved detection of ovarian cancer compared to CA-125 alone. This study is another step forward towards developing a useful serologic biomarker panel for early detection of ovarian cancer that is both highly accurate and specific. The multi-analyte platform, Ella, is fully automated, highly reproducible and a useful tool in future ovarian cancer biomarker discoveries.