Abstract 1761: Serous ovarian cancer risk factors by grade: Evidence for etiologic heterogeneity

Abstract

Abstract Background: Growing evidence suggests that high grade serous ovarian carcinomas may arise from the fallopian tube, while low-grade serous ovarian cancer may develop on the ovaries. Previous studies have suggested differences in ovarian cancer risk factors by histologic subtypes, but few have evaluated differences in risk factors for high- vs. low-grade serous cancers. In the Ovarian Cancer Cohort Consortium (OC3), a pooling project of individual-level data from 23 prospective cohort studies, we evaluated associations of hormonal, reproductive, demographic and lifestyle factors, and family history of cancer with for serous ovarian cancer subtypes. Methods: Among 16 studies that abstracted grade information from pathology reports, 3,095 serous ovarian cancers were identified during follow-up (125 well differentiated, 506 moderately differentiated, 1671 poorly differentiated, 793 undetermined). We used competing risks Cox proportional hazards regression to compute relative risks (RRs) and 95% confidence intervals (CIs) for differences in association by grade. Models were stratified on study and year of birth and adjusted for age, parity and OC use; subtype heterogeneity was evaluated by likelihood ratio tests. Results: Although sample sizes were small for low-grade tumors, there was evidence for heterogeneity in the associations for oral contraceptive (OC) use, age at menopause, endometriosis, and family history of ovarian cancer. For example, each 5-year increase in OC use was associated with a 21% decrease in risk of low-grade serous ovarian cancer (RR: 0.79; 95% CI: 0.62-1.00), but only a 10% decrease in risk of high-grade serous cancers (RR: 0.90; 95% CI: 0.84-0.96; p-heterogeneity = 0.09). Consistent with a prior report in the Ovarian Cancer Association Consortium, endometriosis was associated with increased risk of low-grade serous tumors (RR: 3.77; 95% CI: 1.24-11.48), but not with high-grade serous tumors (RR: 1.11; 95% CI: 0.70-1.74; p-heterogeneity = 0.12). Conclusion: Our results demonstrate heterogeneous associations of risk factors with subtypes of serous ovarian cancer, supporting the idea that the high- and low-grade serous tumors develop through different pathways. Despite the small sample size for low-grade serous tumors, most risk factors were more strongly associated with low-grade tumors compared to high-grade serous carcinomas, suggesting that risk prediction may be more challenging for the most fatal subtype. Identifying subtype-specific risk factor and biomarkers is important both for better understanding ovarian cancer etiology and for targeted development of novel prevention approaches. These results underscore the importance of consortial projects to evaluate rare subtypes (low-grade serous cancers) for the better understanding of etiologic heterogeneity of this deadly disease. Citation Format: Elizabeth M. Poole, Nicolas A. Wentzensen, Britton Trabert, Shelley S. Tworoger, The Ovarian Cancer Cohort Consortium. Serous ovarian cancer risk factors by grade: Evidence for etiologic heterogeneity. [abstract]. In: Proceedings of the 107th Annual Meeting of the American Association for Cancer Research; 2016 Apr 16-20; New Orleans, LA. Philadelphia (PA): AACR; Cancer Res 2016;76(14 Suppl):Abstract nr 1761.