Abstract 721: Minimally invasive test and composite biomarker for early detection of serous ovarian carcinoma

Abstract

Abstract Background: Current screening programs for early detection of high grade epithelial ovarian cancer (HGOvC) among high-risk populations have failed to show improvement in HGOvC mortality, therefore these women are offered risk-reducing bilateral salpingo-oophorectomy (RRBSO) at 35- 40 years. Stratification of high-risk population, especially BRCA mutation carriers, may enable personalized risk counseling and individualization of timing of RRBSO. In most cases, the precursor lesions of HGOvC arise in the epithelium of the fallopian tube (FT) fimbriae rather than intra-peritoneally. It is therefore plausible that proteins, RNA or DNA from early-stage tumor cells may be identifiable in fluid samples obtained from the lumen of the gynecological tract, thus making it possible to identify curable, early stage lesions. Aims: (1) Test the feasibility of uterine lavage as a minimally invasive test for early detection of ovarian cancer, and (2) Identify novel early-detection biomarkers in the uterine lavage fluid (UtLF). Methods: We developed a method for sampling of gynecologic tract fluid termed uterine lavage fluid (UtLF), which is a simple, reproducible, low-cost office procedure that can be performed routinely during gynecologic follow-up visits. We have already collected UtLF from 140 HGOvC patients and control women undergoing gynecologic surgical procedures for non-malignant indications. Deep proteomic profiling of UtLF is performed by isolation of microparticles from body fluids, followed by solubilization, trypsin digestion and high resolution mass spectrometric (MS) analysis (on the Q-Exactive MS). Machine learning algorithms have been used to extract a classifier that can predict the diagnosis of ovarian cancer. Results: Uterine lavage appears to be a feasible, low burden procedure. The MS approach has identified thousands of proteins in each UtLF specimen, in a high throughput manner. The label-free quantification algorithm (MaxQuant) enables a quantitative comparison between samples from cases and controls. We have derived a 20-protein classifier with an area under the curve (AUC) of Receiver Operating Characteristics (ROC) curve of 0.91 at 20% error. The composite biomarker has been applied to an independent validation set with a negative predictive value (NPV) of 92% and positive predictive value (PPV) of 45%. Conclusions: A minimally invasive technique of uterine lavage to collect unique diagnostic samples, coupled with state-of-the-art proteomics methods, results in a highly sensitive and specific composite biomarker which may be developed in to a screening tool for early detection of serous ovarian cancer in high-risk populations. Citation Format: Keren Bahar-Shany, Georgina D. Barnabas, Limor Helpman, Ariella Yakobson-Siton, Tamar Perri, Ram Eitan, Jacob Korach, Tamar Geiger, Keren Levanon. Minimally invasive test and composite biomarker for early detection of serous ovarian carcinoma [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2017; 2017 Apr 1-5; Washington, DC. Philadelphia (PA): AACR; Cancer Res 2017;77(13 Suppl):Abstract nr 721. doi:10.1158/1538-7445.AM2017-721