Ovarian cancer morbidity and mortality could be improved by better methods of primary and secondary prevention. In turn, primary prevention requires better biomarkers of risk, especially those which might translate into strategies for intervention;and secondary prevention requires highly sensitive and specific early detection biomarkers. From specimens obtained months or years prior to ovarian cancer diagnosis, we have accumulated exciting data on early detection and risk biomarkers. We evaluated 28 biomarkers in specimens from the Prostate, Lung, Colorectal, and Ovarian (PLCO) cancer screening trial and showed there is no better screening biomarker than CA125 for detecting preclinical cases within 6 months of diagnosis, underscoring the need to understand CA125 "negative" cases. However, adding HE4, CA72.4, and beta-2-microglobulin (B2M) plus epidemiologic variables like ovulatory cycles, endometriosis, body mass index, and family history of breast cancer improved upon CA125 alone in identifying cases more than a year remote from diagnosis. In work with specimens from the Nurses'Health Study (NHS) taken at least 3 years from diagnosis, we tested a new paradigm for ovarian cancer pathogenesis through events that either increase or decrease immunity to the important epithelial cell and cancer marker, MUC1 (in the same family as CA125). The level of anti-MUC1 antibodies tracked with epidemiologic events raising or lowering risk and a higher level of antibodies correlated with lower ovarian cancer risk in women less than age 64. In other work, we developed an assay to detect anti-CA125 antibodies, found higher levels in ovarian cancer cases with normal CA125 at diagnosis, and hypothesized that immune complexes may shield CA125 from its conventional assay. We now wish to validate these findings in pre-clinical specimens from the European Prospective Investigation into Nutrition and Cancer (EPIC) with an estimated 816 cases and 2024 matched controls. Using all cases and controls, we will first identify key epidemiologic risk factors and develop a risk prediction model. In the same set of specimens, we will measure MUC1 (CA15.3) and MUC16 (CA125) free antigens, anti-MUC1 and anti-MUC16 antibodies, and immune complexes involving mucin antigens and antibodies and determine how they relate to epidemiologic factors, age at entry, and risk for ovarian cancer by remoteness of the blood from diagnosis. In the 196 EPIC cases diagnosed within three years of blood draw and 784 matched controls, we will then measure the additional early detection markers HE4, CA72.4, and beta-2-microglobulin. We will evaluate and refine an early detection algorithm with the particular goal of determining whether the addition of epidemiologic risk factors improve an early detection model and whether mucin-related risk biomarkers can help identify the CA125 or CA15.3 negative case. The goal of this study is to understand how mucin-immunity relates to ovarian cancer pathogenesis and whether markers based upon it can improve performance of the current best early detection biomarkers.
This proposal takes the current best data regarding early detection biomarkers and innovative ideas on risk biomarkers for ovarian cancer and evaluates them in specimens from the large and unique EPIC cohort with the goal of advancing risk stratification and general population screening. Our early detection model includes epidemiologic risk factors and mucin-related markers including one for CA125 bound in an immune complex and shielded from its conventional assay. Our risk algorithm will be based upon an immune model for ovarian cancer that is novel and explains risk better than existing etiologic models.