Updated data request: Ovarian cancer risk factors by histologic subtypes and development of an ovarian cancer risk prediction model in the Ovarian Cancer Cohort Consortium (OC3)
I am resubmitting this request for two primary reasons 1) the co-PI, Dr. Shelley Tworoger has changed institutions, thus the data coordinating center is changing so we need to renew the data use agreement for the appropriate PI/institution, 2) we are requesting updated ovarian cancer outcomes (those available to extramural researchers since the original dataset creation) and updated information on exposures ascertained during follow-up (non-baseline questionnaire data), and information about biologic specimen availability. **It was suggested by Neal Freedman that we add Dr. Tworoger formally to this request so that the DTA can be taken care of via PLCO directly with Moffitt, rather than including as part of a combined DTA between NCI and Moffitt that also includes AARP and BCDDP.
Examining risk factors by tumor subtype. One broad research goal of the OC3 is to examine whether associations of putative ovarian cancer risk factors differ by ovarian cancer subtype. Thus far, we have defined subtypes by tumor histology/grade, dominance (as a surrogate for cell of origin), and aggressiveness (tumors fatal within three years vs. all others). Risk factors that have been analyzed include contraception history, reproductive history, postmenopausal hormone therapy, family history of ovarian cancer, anthropometric variables, analgesic use, and several biomarkers, including androgens, C-reactive protein, and insulin-like growth factors. We observed unique patterns of risk factor associations across subtypes. These results support that pre-diagnostic factors may influence ovarian cancer development and aggressiveness and that considering multiple tumor characteristics simultaneously may provide a clearer picture of disease etiology. Future research in the OC3 will continue to explore multi-faceted approaches to characterizing tumor heterogeneity (e.g., tumor immune marker profiles) and the associations of tumor subtypes with known and suspected risk factors.
Risk prediction. Although there are several known ovarian cancer risk factors, the ability to identify women at high risk remains limited. Thus, a major goal of the OC3 is to improve ovarian cancer risk prediction. On-going OC3 research is developing a risk prediction model for ovarian cancer overall. However, given the unique risk factor profiles of different ovarian cancer subtypes observed in previous OC3 research, and the poor performance of the model in predicting serous cancer (the most deadly subtype), the OC3 is focused on determining whether risk prediction models for ovarian cancer can be improved by accounting for differential associations by cancer phenotype. Research is currently underway to develop risk prediction models by tumor subtype.
Survival. Outside of surgery and chemotherapy, few factors have been associated with improved survival after diagnosis with ovarian cancer. An important goal of the OC3 is to conduct research to improve understanding of the impact of pre- and post-diagnosis exposures, and their interactions with tumor subtype, on survival. For example, we plan to characterize tumor immunosuppressive signatures related to poor prognosis, and examine their relationship with the trajectory of inflammation-related exposures before and after diagnosis. The long-term goal this research is to help focus efforts to develop novel cancer therapeutic strategies, and aid discovery of biomarkers for tailored treatment
Data repository expansion for future research aims. An important goal of the OC3 is to create an infrastructure with a core dataset of important variables for ovarian cancer epidemiology that will be available for future efforts to study ovarian cancer risk. Therefore, the OC3 plans to expand its data repository by obtaining funding to include dietary factors, updated exposure data from follow-up questionnaires, and biomarker information (both plasma/serum markers, including high-throughput omics data, and genetics, as available) that can be used to identify new risk factors as well as early detection markers.
Nicolas Wentzensen
Shelley Tworoger (Shelley.Tworoger@moffitt.org)