Introduction/Background Differentiation of benign and malignant ovarian tumors before surgery is difficult with currently used biomarker based algorithms such as the Risk of Malignancy Index (RMI), and these algorithms are prone to subjective interpretation. This leads to misclassification and incorrect referral to oncologic centers. Therefore, there is an urgent need for more discriminating and objective biomarkers. Recently, ‘Tumor-Educated Platelets (TEPs)’ have shown to be able to distinguish patients with and without a localized malignancy with an accuracy of 96% (Cancer Cell. 2015 Nov 9;28(5):666–676). However, TEPs have never been investigated in clinical early stage ovarian cancer. Therefore, we determined the accuracy of TEPs to discern ovarian malignancies from benign ovarian tumors in patients with an ovarian tumor referred to oncologic centers.

Methodology The RMI was calculated according to Jacobs et al (Br J Obstet Gynaecol. 1990 Oct;97(10):922–999). The TEP score was calculated according to Best et al (Cancer Cell. 2015 Nov 9;28(5):666–676) using a training and evaluation cohort. Pathology results were used for the definitive diagnosis of benign or malignant disease.

Results The training cohort consisted of 41 women (20 FIGO stage I&II ovarian carcinomas (OC), and 21 healthy controls). The evaluation cohort consisted of 43 women (21 OC, and 20 benign tumors (BT)). When using a RMI cut-off level of 200, the accuracy of differentiating OC from BT was 44%. When using TEPs, the accuracy went up to 80% (figure 1 and tables 1 and 2).

Conclusion Implementing TEPs as a standard test before referral, is likely to lead to an increase in correct classification of ovarian tumors. Prospective studies will have to confirm whether this results in a decrease of incorrect referrals of patients with an ovarian tumor to an oncologic center.

Disclosure Nothing to disclose.

• Download figure
• Open in new tab
• Download powerpoint

Abstract EP457 Figure 1