Estrogen Metabolism Pathways in Pregnancy and Subsequent Breast Cancer Risk: A Prospective Follow-up Study

Background: In the years following pregnancy, breast cancer risk is elevated, particularly for hormone receptor negative (HR-) tumors. Exposure to high maternal circulating estrogens, when the breast is vastly remodeling in structure and morphology, has been associated with risk, particularly for HR- tumors. Estrogen metabolite profiles in nonpregnant women, notably the ratio of 2:16 hydroxylation (OH) pathway metabolites, are associated with postmenopausal breast cancer development; whether estrogen metabolism during pregnancy influences subsequent HR- breast cancer risk is unknown. Methods: Weconducted a population-based case-control study in women 19-39 years identified in the Finnish Maternity Cohort Biobank and linked with the Finnish Cancer Registry to identify breast cancer diagnoses within 20 years of pregnancy. Estrogens and metabolites were measured using highly reliable and sensitive LC-MS/MS methods in serum collected during the first and second trimesters of pregnancy. Included were invasive, ER-/PR- breast cancer cases (n=449) and controls (n=449) matched on maternal age at index pregnancy, parity, calendar year of serum collection, gestational week of blood collection, and number of freeze/thaw cycles. Associations between estrogens and breast cancer risk were estimated using odds ratios (ORs) with 95% confidence intervals (CIs) from conditional logistic regression models. Results: The median years of follow-up between blood collection and breast cancer diagnosis/control selection was 9 (range 0-19). Total estrogens were positively associated with ER-/PR- breast cancer (OR associated with a doubling of total estrogens 1.16; 95% CI 1.02-1.32), as were metabolites in the 16-pathway including estriol [OR 1.11; 95% CI 1.01-1.22], 16-epiestriol [OR 1.11; 95% CI 1.01-1.21)], 17-epiestriol [OR 1.06; 95% CI 1.01-1.13], and total 16-hydroxylation pathway metabolites [OR 1.11; 95% CI 1.00-1.24]. There was no clear association with the ratio of 2:16 hydroxylation pathway metabolites. Some associations differed by parity, age at diagnosis, and gestational timing of blood collection, but interactions were not statistically significant. Results were similar when restricted to cases occurring within 15 years since pregnancy. Conclusion: This prospective study demonstrated positive associations of estrogen metabolites in pregnancy and risk of ER-/PR- breast cancer, but the magnitudes varied by metabolite. No strong or consistent pattern for one metabolic pathway emerged suggesting that total estrogen concentrations during pregnancy are associated with subsequent HR- breast cancer development, regardless of how they are metabolized.