Validation of a risk assessment decision rule for epithelial ovarian cancer

点击翻译按钮获取中文摘要

DESCRIPTION (provided by applicant): Our goal is to develop a risk-classification tool to identify post-menopausal women who are at high or elevated risk for epithelial ovarian cancer (EOC). We will develop and validate a risk-assessment decision rule based on serum markers CA125 and HE4 as well as epidemiologic risk factors using a split-sample design and data from the Women's Health Initiative (WHI) Observational Study (OS), the WHI Clinical Trial (CT) and the Prostate, Lung, Colon and Ovary (PLCO) trial. A decision rule developed as preliminary work identified 10%- 13% of all post-menopausal women as elevated risk and predicted 26%-58% of cases. Surgical prevention consisting of prophylactic removal of fallopian tubes (FTs) and ovaries is recognized as the best approach to prevent epithelial ovarian cancer (EOC) and especially high grade serous cancer (HGSC) in high-risk women. We will develop a tool to categorize women into risk classifications that have clear clinical implications including recommendations for surgical prevention, imaging, surveillance and routine care. In the first 2 aims we will focus on development of the epidemiologic portion of the model because the serum marker portion of the decision rule has already been developed based on extensive published work. In Aim 3 we will validate the final decision rule that combines epidemiologic risk factors with serum markers. We operationally define high risk as relative risk (RR) of at least 6, and elevated risk as RR of at least 2. Aim 1: Using a randomly selected fraction of women participating in the WHI OS, WHI CT and the PLCO trial, develop decision rules to identify women at elevated risk for EOC and HGSC using epidemiologic variables alone. Using the remaining women from each study, validate each decision rule within each population. Decision rules developed in this aim will be specific to each study population in order to take advantage of all of the epidemiologic data collected by each study. Aim 2: Cross-validate the decision rules developed in Aim 1 to identify the rule that best identifies women at elevated risk for EOC and HGSC across the 3 different populations. This will require refinement of the decision rules to accommodate differences in data collection across the cohorts. The study fraction used for development will be used to identify the best common rule, and cross-validation will employ the remaining validation fraction of women participating in the WHI OS, WHI CT and the PLCO trial. This aim will identify the best epidemiologic decision rule for use in the overall decision rule to be validated in Aim 3. Aim 3: Using nested case-control study serum marker data as well as epidemiologic data from each study, validate the best decision rule from Aim 2 in combination with the serum marker component of the rule to identify women at elevated- and high-risk for EOC and HGSC in each of those populations. In this aim we will validate the final decision rule that incorporates both epidemiologic risk factors and serum markers.