Category 3
Cancer Prevention and Control

Background: Ovarian cancer is the most fatal gynecologic malignancy. Four major histotypes are distinguished which have substantial biological and clinical differences. Previous studies have established that inflammation-related exposures are associated with ovarian cancer risk but have not widely assessed histotype-specific effects and interactions between exposures.

Goal: We are conducting a large analysis in the Ovarian Cancer Association Consortium (OCAC) to evaluate inflammation-associated risk factors and their interactions with ovarian cancer histotypes.

Objectives: The objective of this analysis is to determine whether ovarian cancer risk factors act independently to influence ovarian cancer histotype-specific risk, or if risk factors act jointly to influence risk.

Approach/Methods: The OCAC includes mainly ovarian cancer case-control studies with excellent histotype and covariate information. We chose six a priori exposure combinations based on known associations with ovarian cancer, plausible biological interaction and potential as modifiable risk factors. Multivariable logistic regression was used to assess primary exposure effects in histotypes, interaction and joint effects between exposures in a core set of 7 studies (n=7,529 cases; 10,491 controls) that captured data for all inflammation-related exposures of interest. P-values for interaction between modifiers and risk factors were calculated by Wald test of the beta coefficient for the interaction term.

Results: Five individual risk factor associations were heterogenous by histotype (p<0.0001): aspirin (reduced risk) was associated with serous tumors; endometriosis (increased risk) and tubal ligation (decreased risk) were most strongly associated with endometrioid and clear cell tumors; high LOC (increased risk) was most strongly associated with clear cell tumors; high BMI (increased risk) was associated with endometrioid and mucinous tumors. We did not observe interactions between inflammation-related exposures both overall and within histotype (pint 0.10-0.98).

Importance to Public Health: Inflammation-related risk factor associations vary by cancer histotype. Despite expectations from biological priors, statistical interactions between inflammatory exposures were not observed. Our ability to elucidate interaction effects may be complicated by a lack of detailed information for timing of exposures or the potential existence of multiple inflammatory pathways. An integrated assessment of inflammatory risk factors in ovarian histotypes is critical to better understand possible actionable strategies to lower ovarian cancer risk.