Common genetic variants and ovarian cancer screening in PLCO
1. Use a polygenic risk score to identify subsets of the population at lower and higher risk of ovarian cancer and evaluate screening performance in these subgroups, and
2. Evaluate genetic predictors of CA-125 levels in a general population of women undergoing ovarian cancer screening.
Identifying high-risk populations for screening can reduce overscreening and overtreatment compared to general population screening. Population-based screening for ovarian cancer did not show a mortality benefit in PLCO. However, screening may be more favorable in a high-risk subset of the population. We propose to generate a polygenic risk score (PRS) derived from single nucleotide polymorphisms (SNPs) previously associated with ovarian cancer. Among PLCO participants, this PRS will be used to assign women to risk groups, and ovarian cancer screening will be evaluated among various risk groups.
CA-125 is used as a biomarker for ovarian and other tumors, but levels are highly variable among healthy women, and CA-125 levels may remain low even among some women with prevalent ovarian cancer. Genetic variation may influence baseline CA-125 levels in healthy individuals, but genetic associations with CA-125 levels have not been determined and may have implications for future screening guidelines.
1. Performance of ovarian cancer screening in women stratified by polygenic risk score (PRS). A PRS will be derived from an independent GWAS dataset from the Ovarian Cancer Association Consortium (OCAC) using all previously identified SNPs significantly associated with ovarian cancer overall or subtype-specifically. This PRS will be used to identify PLCO participants in the top and bottom quartile of genetic risk for ovarian cancer. The performance of screening (sensitivity, specificity, predictive values) will then be evaluated among women in the risk-stratified subsets of the population. This study will improve our understanding of ovarian cancer screening test characteristics among women at both low and high risk.
2. Prediction of cancer antigen 125 (CA-125) levels using common genetic variants. Among PLCO screening arm participants who were genotyped, GWAS data will used to search for SNPs associated with CA-125 levels in healthy individuals. Women with an ovarian cancer diagnosis within three years of the CA-125 measurement will be excluded, and outcomes will be adjusted for age. Additional potential covariates may include but are not limited to race, BMI, parity, menopausal status, family history of ovarian cancer, smoking. Findings will be validated using additional subjects from the European Prospective Investigation into Cancer and Nutrition (EPIC), the Nurses’ Health Study (NHS), and New England Case Control Study (NECC).
None