Ovarian cancer remains the most lethal of the gynecological cancers. Although relatively uncommon, afflicting ϳ1 of 70 women in the United States, the high mortality rate makes this disease a major health concern. The high mortality rate arises from the lack of an effective screening approach combined with inadequate therapeutic approaches for advanced disease. Indeed, fewer than 25% of ovarian cancers are identified at an early curable stage. Other than family and personal history of cancer or

more »

... the inheritance of abnormalities in the BRCA1 or BRCA2 or other ovarian cancer predisposition genes, there are few indicators that a women has an elevated risk of development of cancer. Thus, it is difficult to identify women warranting prophylactic surgery or intensive screening. The high mortality rate combined with the lack of effective screening approaches make ovarian cancer a strong candidate for chemoprevention (1). In this issue of "Clinical Cancer Research," Rodriguez-Burford et al. (2) explore the potential mechanisms underlying epidemiological observations that ovarian cancer occurs at a lower frequency in women exposed to NSAIDS. 2 Cramer et al. (3) in a population-based, case-control study demonstrated that a 6-month exposure to acetaminophen at least 1 day/week, but not aspirin or ibuprofen, resulted in a statistically significant 2-fold decrease in the lifetime risk of developing ovarian cancer. Moysich et al. (4) in a hospital-based, case-control study also demonstrated a time-and dose-dependent association of acetaminophen, but not aspirin use, with decreased risk of developing ovarian cancer. Rosenberg et al. (5) in a large-scale, casecontrol study failed to confirm the association of moderate use of acetaminophen and decreased risk of ovarian cancer demonstrated by Cramer et al. (3) but did identify a statistically significant decrease in development of ovarian cancer with prolonged acetaminophen use, at least Ն4 days/week for 5 years. Tavani et al. (6) has failed recently to detect an association of aspirin use with prevention of ovarian cancer. In contrast to these reports, Akhmedkhanov et al. (7) demonstrated a 2ϳ3fold decrease in epithelial ovarian cancer associated with aspirin use. Taken together, although controversial, these epidemiological observations suggest that an improved understanding of the mechanism(s) by which NSAIDS may decrease the development of ovarian cancer could lead to improved approaches for chemoprevention of this deadly disease. The studies in this issue of "Clinical Cancer Research" identify several of the problems associated with attempting to use cell models to understand the mechanisms underlying epidemiological studies of chemoprevention of epithelial ovarian cancer. At this point, there is no readily manipulatable mammalian model of epithelial ovarian cancer. Indeed, the only extant animal model with reproducible epithelial ovarian cancer is a chicken model (8). This model, while contributing to our understanding of chemoprevention of ovarian cancer, is far from optimal. The mouse models of human cancer consortium, as well as members of the ovarian cancer research community, are investigating a number of approaches to develop practical animal models; however, at the present, it is necessary to rely on ovarian cancer cell lines, cultures of normal ovarian epithelium, or normal epithelium with enforced expression of genes, such as the large T antigen of SV40, telomerase, and oncogenes to attempt to explore the mechanisms underlying chemoprevention of epithelial ovarian cancer. These established cell culture systems and, in particular, ovarian cancer cell lines may not adequately reflect the transformation processes targeted by chemoprevention agents in vivo. The studies by Rodriguez-Burford et al. (2) demonstrate that acetylsalicylic acid, acetaminophen, and a Cox-2 inhibitor, NS-398, can decrease the growth of fully transformed human epithelial ovarian cancer cells. The Cox-2 agent both decreased cell proliferation in established cell lines and induced apoptosis in freshly isolated ovarian cancer cells. This information implicates cellular proliferation and survival in the action of NSAIDS on the development of epithelial ovarian cancer. Although all three NSAIDS decreased the proliferation of ovarian cancer cell lines, concentrations far above the therapeutic range and well above the maximal tolerated dose and, in particular, the chemopreventative doses of the inhibitors were required to demonstrate antiproliferative effects in vitro (2). This is compatible with the observation that in contrast to bowel cancer cells, which are very sensitive to the effects of NSAIDS likely as a consequence of overexpression of Cox2 (9 -12) and dependent on Cox2 function for growth and survival, human epithelial ovarian cancer cells appear to express Cox1 and Cox2 at low levels (2, 13). Thus, either a prolonged low-dose exposure to the NSAIDS has effects on Cox2 function in normal or partially transformed ovarian epithelium, or NSAIDS exert their effects through alternative mechanisms. The observation that very high concentrations of NSAIDS are required to decrease the proliferation of ovarian cancer cell lines and that ovarian cancer cells express low levels of COX1 and COX2 (2, 13) suggests that the effects of NSAIDS on ovarian cancer prevention may be attributable to mechanisms independent of inhibition of COX1 and COX2 as reflected by the effects of NSAIDS on the proliferation or survival of established ovarian cancer cell lines. In this aspect, it is important to note that the epidemiological data most strongly implicate acetaminophen as compared with aspirin or ibuprofen in the prevention of ovarian cancer (2-7). Acetaminophen is a relatively poor inhibitor of cyclooxygenase and does not have