Abstract PO029: Oral contraceptive use and postmenopausal sex steroid hormone metabolism

Abstract
OBJECTIVE: Oral contraceptive (OC) use is common in the United States and high-profile articles have renewed research interest in OC use and cancer risk. Several mechanisms have been proposed to explain how OC use influences postmenopausal cancer risk; one of the leading hypotheses is that OC use fundamentally alters feedback in the hypothalamic-pituitary-gonad axis, leading to long term changes in sex steroid hormone metabolism. However, women who use OCs are also unique with respect to factors that may affect both hormone metabolism and cancer risk. No studies have comprehensively commented on hormone levels in postmenopausal women who did and did not use OCs. DESIGN: We examined differences in circulating markers of sex steroid hormone metabolism associated with prior use of OCs among a subcohort of postmenopausal women from the Women’s Health Initiative Observational Study (983 women with at least one intact ovary and not currently using menopausal hormone therapy). MATERIALS AND METHODS: Using highly sensitive liquid chromatography–tandem mass spectrometry assays, we measured over 30 markers of estrogen and androgen metabolism in study serum samples. We used linear regression (adjusted and weighted) to estimate geometric mean hormone levels, which were then converted into relative percent differences. We controlled for potential confounders and stratified by: lifetime ovulatory cycles, body mass index, and parity (the focus of this abstract). RESULTS: Women with a history of OC use (n=346) had lower levels of estrogen metabolites than women who never used OCs (n=637), particularly when limiting this comparison among parous women (n=158 nulliparous women, n=825 parous women). In parous women, OC use was associated with reductions in total estrone (-18.5%, 95% confidence interval [CI] -31.1, -3.7%) and in estrogen metabolism overall: 2-hydroxylation metabolites were reduced by -18.5% (CI -28.2, -7.7%), 4-hydroxylation metabolites by -19.9% (CI -29.4, -9.1%), and 16-alpha-hydroxylation metabolites by -24.0% (CI -34.2, -12.2%). Hormone differences associated with OC use among nulliparous women were imprecise. When we made comparisons to nulliparous women who did not use OCs, women who both used OCs and were parous had lower testosterone (-21.2%, CI -33.1, -7.2%) and higher dihydrotestosterone relative to testosterone (its metabolic precursor; 16.5%, CI% -0.6, 36.6). However, aside from lower total estrone (-25.5%, CI -44.6, 0.2%), estrogen metabolism was not different between these groups. CONCLUSIONS: We saw evidence suggesting estrogen metabolism, as reflected in circulating markers, may be lower in postmenopausal women who previously used OCs. These effects were strongest among parous women, likely due to sufficient sample size, rather than any synergistic effect attributable to both using OCs and giving birth. These results support the idea that OC users have lower estrogen levels, but this should be evaluated in other age groups, as differences in hormone metabolism early in the lifecourse may lead women to use OCs. Citation Format: Kara A. Michels, Sally B. Coburn, Garnet Anderson, Louise A. Brinton, Chu Chen, Roni T. Falk, Margery L. Gass, Ashley M. Geczik, JoAnn E. Manson, Ruth M. Pfeiffer, Kerryn Reding, Gloria E. Sarto, Nicolas Wentzensen, Robert A. Wild, Xia Xu, Britton Trabert. Oral contraceptive use and postmenopausal sex steroid hormone metabolism [abstract]. In: Proceedings of the AACR Virtual Special Conference: Endometrial Cancer: New Biology Driving Research and Treatment; 2020 Nov 9-10. Philadelphia (PA): AACR; Clin Cancer Res 2021;27(3_Suppl):Abstract nr PO029.