Ovarian cancer is the fourth leading cause of cancer deaths in women in the United States. There are three potential approaches to decreasing ovarian cancer deaths: (1) more effective treatment of advanced stage cancers, (2) better screening tests that allow early detection while cancer is confined to the ovary, and (3) prevention of the disease. All of these avenues should be explored, but we believe that prevention may represent the most promising approach. Although it is not widely appreciated, factors such as pregnancy and oral contraceptive (OC) use, which reduces the number of times a woman ovulates during her life, strikingly decrease the incidence of ovarian cancer. Women who use OCs for more than 5 years or have three or more children decrease their risk of ovarian cancer by about 50'. Funding from the Department of Defense (DOD) Ovarian Cancer Research Program in 1998 provided initial support for our multidisciplinary program project that focuses on genetic/molecular epidemiology and chemoprevention. The observation that OC use is strongly protective against ovarian cancer suggests that chemoprevention may be an effective means of decreasing mortality. Elucidation of the diverse etiologic factors involved in ovarian carcinogenesis is needed to define high-risk subsets of women in whom to target chemoprevention. An understanding of the etiologic heterogeneity of ovarian cancer also could facilitate prevention, as the optimal approach may vary between individuals. In the North Carolina Ovarian Cancer (NCOC), study we have enrolled 300 newly diagnosed ovarian cancer cases in a 48 county region of North Carolina as well as 350 age- and race-matched controls. Subjects undergo an in-home interview regarding risk factors for ovarian cancer, and blood and tumor tissue are collected. Previously, we found a strong association between high lifetime ovulatory exposure and alteration of the p53 tumor suppressor gene in ovarian cancers, and we are attempting to confirm in the NCOC study whether specific genetic alterations define molecular signatures of distinct etiologic pathways. This work will be completed with funds provided in the initial DOD grant and a subsequent National Cancer Institute grant. About 10' of ovarian cancers are attributable to mutations in the BRCA1 or BRCA2 ovarian cancer susceptibility genes, but polymorphisms in genes involved in pathways implicated in the etiology of this disease also may affect susceptibility. We recently have accrued sufficient samples in the NCOC study to begin examining the role of genetic polymorphisms in ovarian cancer susceptibility. The mechanism underlying the protective effect of OCs against ovarian cancer remains unknown. However, we have shown that the progestin levonorgestrel stimulates apoptosis (programmed cell death) in the ovarian epithelium, which suggests that the protective effect of OCs may be due to clearance of premalignant cells. We have initiated chemoprevention trials using levonorgestrel in chickens, one of the few species with a high spontaneous incidence of the disease, and found a 35' decrease in the rate of ovarian cancer.

In the next 4-year funding period, our plans are to continue accrual in the NCOC study until we enroll 820 cases and 820 controls. In Project 1, we will continue to examine whether polymorphisms in genes involved in pathways relevant to ovarian carcinogenesis affect ovarian cancer susceptibility. We also will examine the interaction between genotype and epidemiologic risk factors such as OC use, and we will examine a significant African American population since allele frequencies vary between racial groups. In Project 2, we will continue to develop progestins as ovarian cancer chemopreventives in the chicken model by comparing the efficacy of various progestins and will explore the efficacy of vitamin D analogues. We also will explore further the role of the TGF-[BETA] pathway in progestin-mediated apoptosis. In addition, in an Idea Grant, we will explore the interaction of progestins with the A and B isoforms of the progestin receptor with the goal of developing selective progestin receptor modulators.

Chemoprevention of ovarian cancer with progestins is a promising strategy for decreasing ovarian cancer mortality. An understanding of the role of genetic polymorphisms in ovarian cancer susceptibility would facilitate identification of a subset of high-risk women for whom to target chemoprevention trials. We will continue to study the efficacy of various progestins in an avian model, while also striving to understand the mechanisms that underlie their protective effect, as this could lead to the development of more selective means of exploiting this pathway for ovarian cancer prevention in the future.