Demographic, reproductive, and clinical factors influence CA125 levels in healthy women and those with ovarian cancer; therefore, a single threshold for CA125 screening may not be appropriate. Using detailed questionnaire data collected from over 2000 healthy women participating in the New England Case Control (NECC) study, we will confirm previously identified associations and identify novel factors such as diet, medication, comorbidities, or single nucleotide polymorphisms (SNPs) associated with CA125 levels in healthy women. Next, we will determine predictors of pre-treatment CA125 in cases participating in the NECC and other Ovarian Cancer Association Consortium (OCAC) studies. Based on previous analyses and novel findings from the research proposed here, we will determine which factors improve the ability of CA125 to distinguish women with and without ovarian cancer in the NECC and validate our findings using detailed questionnaire data and prediagnostic blood samples or CA125 measurements from Nurses’ Health Study (NHS) and Prostate Lung Colon and Ovarian Cancer Screening Trial (PLCO). We also have the unique opportunity to measure CA125 in paired follicular and luteal samples from 40 white women at three timepoints as well as 40 black and 40 Asian women at a single timepoint from the NHS cohort, which will allow us to explore the influence of race, menstrual phase, and change in CA125 over time. Thus, we plan to delineate heterogeneity in CA125 levels that could be used to define the normal range of CA125 for each individual and refine screening protocols. We will accomplish this goal through the following specific aims:
1. Identify key predictors of plasma CA125, including previously identified and novel factors (e.g., diet, chronic diseases, medications, SNPs), separately in women with and without ovarian cancer in the NECC.
a. In 2,026 population-based controls from the NECC study, measure plasma CA125 and confirm known predictors of CA125 and evaluate novel factors.
b. In 4,417 cases from NECC and nine other studies participating in the Ovarian Cancer Association Consortium, assess if predictors from aim 1a predict pre-operative CA125.
2. Define personalized CA125 cutpoints based on individual characteristics and test the performance of CA125 as ovarian cancer screening test utilizing these cutpoints in the NECC study.
a. Develop an algorithm with a defined set of predictors that influence CA125 to determine personal cut-points for ovarian cancer screening.
b. Compare the discriminatory ability of a single CA125 cutpoint (>35) to personalized cutpoints in NECC to differentiate cases and controls by comparing receiver operator characteristic curves.
c. Validate the discriminatory ability of personalized covariate-dependent CA125 cutpoints to identify women who will develop ovarian cancer within 4 years using prospectively collected samples from two cohorts (NHS, PLCO).
3. Compare plasma CA125 values in paired follicular/secretory samples from 40 white, 40 black, and 40 Asian premenopausal women.
a. Assess the difference in plasma CA125 by menstrual cycle phase.
b. Compare plasma CA125 values across race categories for follicular and secretory samples.

Shelley Tworoger, PhD Brigham and Women's Hosp
Daniel Cramer, MD, ScD Brigham and Women's Hosp
Bernard Rosner, PhD Brigham and Women's Hosp
Nicolas Wentzensen, MD, PhD, NCI