Inflammation, Anti-Inflammatory Drug Use and Risk of Ovarian Cancer

Abstract

Ovarian cancer is a highly fatal malignancy. Advancing knowledge of ovarian cancer etiology to improve prevention strategies is essential to reducing ovarian cancer mortality. My research investigated the role of inflammation in ovarian cancer etiology. First, I evaluated if non-steroidal anti-inflammatory drug (NSAID) use is associated with risk of ovarian cancer in the Nurses’ Health Study (NHS) and NHSII cohorts. Using Cox proportional hazards models, I evaluated associations for aspirin and non-aspirin NSAIDs, considering timing, frequency, quantity, dose, and duration. Then, to investigate the prostaglandin synthesis pathway as a possible mechanism underlying the association between NSAID use and ovarian cancer, I conducted two additional studies. In a case-control study nested in the NHS, NHSII, and Shanghai Women’s Health Study, I estimated the association between pre-diagnosis urinary PGE-M, a marker of prostaglandin E2, and ovarian cancer using unconditional logistic regression. In a case-control study based in the NHS, NHSII and New England Case-Control study, I used polytomous logistic regression to evaluate if the association between NSAID use and ovarian cancer differed by COX1 and COX2 expression, or infiltration with tumor-associated macrophages (TAMs). There was no evidence of an association between aspirin use and ovarian cancer risk when all aspirin products were evaluated together; however, the hazard ratio (HR) for current versus non-use of low-dose aspirin was 0.77 (95%CI=0.61-0.96). We observed a positive association for ≥10 years of non-aspirin NSAID use versus non-use (HR=1.34, 95%CI=1.06-1.70; p-trend=0.03), and a positive trend for tablets per week (p=0.03). Pre-diagnosis urinary PGE-M was not associated with risk of ovarian cancer (odds ratio [OR]= 1.09, 95%CI=0.71-1.66; p-trend=0.48). Additionally, the association between NSAIDs and ovarian cancer did not differ by tumor COX1 or COX2 expression (p-heterogeneity>0.05). We observed significant heterogeneity for current aspirin use (p-heterogeneity<0.001) and NSAID use (p-heterogeneity<0.001) with risk of ovarian cancer by infiltration with pro-tumorigenic M2-type TAMs. Overall, our results suggested a lower risk of ovarian cancer among low-dose aspirin users and an increased risk of ovarian cancer among women with heavy use of non-aspirin NSAIDs. There was no evidence that the prostaglandin synthesis pathway influences ovarian carcinogenesis, but immune function may be relevant.