Abstract LB152: Protocol for clinical validation of an ovarian cancer screening test using blinded retrospective evaluation of prospectively collected samples from a screening trial

Abstract Objectives: Our objective is to design a clinical validation study of an ovarian cancer (OC) biomarker test leveraging blinded samples from the UKCTOCS trial, thus minimizing bias. Our aim is to demonstrate the safety and diagnostic effectiveness of an OC screening test and present an attractive alternative to a large, multicenter and multi-year prospective trial. Methods: We propose a blinded retrospective evaluation of an extracellular vesicle-based blood test for OC screening in a prospectively collected cohort of asymptomatic postmenopausal women. The UKCTOCS trial enrolled postmenopausal women aged 50-74 years with no increased risk of familial ovarian cancer from the UK between 2001 and 2005. 50,624 eligible participants were randomized to the multimodal screening arm, in which annual blood samples were collected for CA125 biomarker testing. Each woman in the MMS arm underwent annual CA-125 measurements as a Level 1 screen, which was interpreted using the Risk of Ovarian Cancer Algorithm (ROCA) to calculate the risk of having ovarian cancer based on age specific prevalence and the trajectory of CA125 signal across all measurements undertaken in the course of the trial. Each risk result was classified as either normal, intermediate requiring repeat CA125 in 3 months, or elevated. Women with elevated risk or three consecutive intermediate risk results had transvaginal ultrasound and repeat CA125 in 6 weeks. Those with persistent abnormalities underwent further clinical workup with a subset resulting in surgery. Trial participants underwent annual screening until 2011.A retrospective study of banked serum samples from annual (Level 1) screens nested in the prospectively defined cohort of the MMS arm will be undertaken. The cohort will include 1. Cases - All women with screen detected invasive epithelial ovarian cancer (iEOC) and those with screen negative (interval) iEOC diagnosed within 1 year of the last annual test. 210 women meet these criteria of whom 200 (95.24%; 200/210) have sufficient banked sample volume remaining for testing. 2. Controls - All women who attended screening and were not diagnosed with ovarian cancer during the screening phase or the follow-up phase in the MMS arm. Of the 49,536 women, annual samples are available from 49,412 (99.7%; 49,412/49,536) for this study. A selection of approximately 2,500 subjects will be randomly selected from the available population after demographic enrichment for black and Asian women to approximate the US population. All samples are stored in straws in liquid nitrogen in a commercial biorepository. The blinded samples will undergo their first thaw, and the required aliquot will be packed and shipped on dry ice with temperature loggers to ensure sample integrity. Samples will be tested in a clinical laboratory that conforms to good clinical practice (GCP). The testing will be performed using a frozen design of extracellular vesicle-based OC screening test by trained operators. All clinical data associated with samples tested in this study will remain blinded until testing has been completed. The blinding key will be maintained by the co-investigators, University College London (UCL). Primary endpoints of sensitivity, specificity, PPV, and NPV will then be reported based on a pre-specified statistical analysis plan. Citation Format: Michael J. Smith, Brendan Manning, Troy B. Hawkins, Tara Maddala, David F. Ransohoff, Christine D. Berg, Steven J. Skates, Toumy Guettouche, Sophia Apostolidou, Aleksandra Gentry-Maharaj, Dawn Mattoon, Usha Menon. Protocol for clinical validation of an ovarian cancer screening test using blinded retrospective evaluation of prospectively collected samples from a screening trial [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2025; Part 2 (Late-Breaking, Clinical Trial, and Invited Abstracts); 2025 Apr 25-30; Chicago, IL. Philadelphia (PA): AACR; Cancer Res 2025;85(8_Suppl_2):Abstract nr LB152.