Background: Cyclooxygenase (COX) enzymes catalyze key reactions of prostaglandin biosynthesis. COX-2 is implicated in inflammation, and high expression has been associated with lower survival from several malignancies, including ovarian cancer. COX-1 is thought to contribute primarily to normal physiology, so cancer outcomes have been evaluated less frequently. However, COX-1 is highly expressed in ovarian cancer, and both COX enzymes have differences in expression across histologic subtypes of disease. This study was undertaken to evaluate COX-1 and COX-2 expression in relation to ovarian cancer survival, including differences between serous and non-serous disease subtypes.

Methods: A tissue microarray was created from primary ovarian, fallopian, or peritoneal cancers diagnosed at the Vanderbilt University Medical Center. COX-1 (Santa Cruz Biotechnology, Dallas, TX) and COX-2 (Thermo Fisher Scientific, Inc., Waltham, MA) immunostaining was performed; semi-quantitative measurement was conducted using the automated Ariol® SL-50 Platform. Outcomes and clinical characteristics, including age, stage, grade, histologic subtype, residual disease, and race were collected by medical record abstraction. Progression-free and overall survival were evaluated using proportional hazards regression to calculate Hazard Ratios (HR) and Confidence Intervals (CI).

Results: Among 191 epithelial ovarian cancer cases, 57 had documented disease progression, and 121 deaths occurred. COX expression was categorized into tertiles; significant differences were found in regards to histologic subtype, disease stage, and residual disease. In multivariate models among all cases, increasing tertiles of COX-2 expression was significantly associated with increased hazards of disease progression (HR: 1.43, 95% CI: 1.01-2.02) but not death (HR: 1.02, 95% CI: 0.79-1.30). Among serous subtypes, higher COX-2 expression was significantly associated with worse overall survival (HR: 1.53, 95% CI: 1.04-2.23). Among non-serous subtypes, higher COX-1 expression was associated with better overall survival (HR: 0.46, 95% CI: 0.24-0.90). Differences by histology were suggestive for COX-2 (P-value for interaction=0.090) and significant for COX-1 (P-value for interaction=0.013).

Conclusions: This single institution retrospective study demonstrates distinct associations between COX enzymes and ovarian cancer disease outcomes across histologic subtypes. If replicated, these findings imply that histologic subtype of disease may be important for the appropriate selection and application of COX inhibitors for ovarian cancer treatment.

Citation Format: Alicia Beeghly-Fadiel, Meral El Ramahi, Andrew Wilson, Spencer Keene, Oluwole Fadare, Marta A. Crispens, Dineo Khabele. The importance of cyclooxygenase 1 and 2 expression in ovarian cancer survival: Notable differences by histologic subtype. [abstract]. In: Proceedings of the AACR Special Conference on Advances in Ovarian Cancer Research: Exploiting Vulnerabilities; Oct 17-20, 2015; Orlando, FL. Philadelphia (PA): AACR; Clin Cancer Res 2016;22(2 Suppl):Abstract nr A63.