PREDICT: The Prospective Early Detection Consortium for Ovarian Cancer
Abstract
Epithelial ovarian cancer (EOC) is a highly fatal disease, as generally it gets diagnosed at a late stage when chances of a cure are low. Mortality could be reduced significantly if EOC were diagnosed more often in earlier stages. Current strategies for ovarian cancer screening use a combination of a blood-based biomarker, CA125, plus a gynecologic examination by ultrasound imaging. A combination of biomarker and imaging is used because either method alone lacks specificity for ovarian cancer and would cause too many false-positive findings and invasive follow-up examinations among healthy women. Unfortunately, however, randomized trials showed no clear reduction in ovarian cancer mortality for women screened by CA125 plus ultrasound imaging, compared to unscreened controls. To improve screening, biomarkers with higher detection sensitivity than CA125 for earlier-stage tumors are required, to be used together with imaging. Promising types of blood-based biomarkers for ovarian cancer include tumor-associated protein antigens (i.e., atypical proteins produced by tumors that may leak into the bloodstream), auto-antibodies (AAbs) against such proteins formed by the immune system, alterations in levels of small noncoding RNA fragments (micro-RNAs), and circulating tumor DNA fragments (ctDNA). For each of these potential marker types, modern “omics” technologies (e.g., proteomics, genomics) allow measuring up to thousands of candidates in small blood volumes and hold great promise for discovery of detection markers by comparing large sets of candidate measurements in blood samples from cancer patients and cancer-free individuals. However, to date, omics-based discovery of novel EOC detection markers has been hampered by the use of sub-optimal study designs, based on serum measurements for patients with relatively advanced disease. Thus, markers identified in clinical case-control settings often failed to show useful detection discrimination upon further testing in prospective studies, suggesting that either they lacked tumor specificity (e.g., they may have been related to inflammation or other a-specific phenomena accompanying advanced cancer) or they had insufficient sensitivity for the detection of smaller and earlier-stage tumors. To overcome these problems and to identify markers that can detect disease before the usual time of diagnosis, we propose to conduct omics-based discovery studies in pre-diagnosis blood samples collected from EOC cases and controls in prospective cohorts, with large enough numbers of EOC cases and controls to allow the statistical selection of diagnostically informative markers. To achieve this, we will establish the Prospective Early Detection Consortium for Ovarian Cancer (PREDICT), a consortium of six of the world’s largest prospective cohort studies, assembling a biobank resource with blood samples collected less than or equal to 18 months prior to diagnosis from over 500 well-characterized cases of EOC and 1,000 cancer-free controls, with background risk factor information and previous measurements of CA125. The first specific aims for PREDICT will focus on the discovery and cross-validation of auto-antibody panels and micro-RNA profiles as early detection markers. For these two marker types, we performed extensive omics-based pilot studies and found first individual AAbs and miRNA signatures with potential utility for ovarian cancer detection. However, our omics-based discovery studies so far were based exclusively on clinical case-control comparisons, with limited numbers of EOC cases and controls. Thus, our first candidate markers identified may include false leads, whereas, on the other hand, many relevant AAb or miRNA markers also may have been missed. Therefore, we aim to perform prospective and larger studies for both extended omics discovery and subsequent validation of AAb-based detection panels and for prospective validation of miRNA signatures. In preparation for
Document Details
- Document Type
- DoD Grant Award
- Publication Date
- Jul 16, 2019
- Source ID
- W81XWH1910307
Entities
People
- Rudolf Kaaks
Organizations
- German Cancer Research Center
- United States Army