Abstract B31: Oral contraceptive use in relation to breast cancer subtypes in African American women: Results from the AMBER Consortium

Abstract

Abstract Introduction. Evidence on oral contraceptive (OC) use in relation to breast cancer subtypes is sparse and the findings are inconsistent. We investigated duration and recency of OC use in relation to molecular subtypes of breast cancer in a pooled analysis of data from African American women. Methods. The African American Breast Cancer Epidemiology and Risk (AMBER) Consortium is a collaborative research project which pools data, samples, and tumor tissue from several studies. The present analysis was based on data from the Black Women's Health Study, the Carolina Breast Cancer Study, and the Women's Circle of Health Study. Pathology reports and/or cancer registry data were abstracted to classify cancers by molecular subtype. Exposures of interest were duration and recency of OC use. Never use was defined as never having used OCs or use that lasted less than 1 year. Duration was categorized as <5 years, 5-9 years, and ≥10 years; recency was categorized as <5 years ago, 5-9 years ago, and ≥10 years ago. Multivariable analyses controlled for age, study, time period, geographic area, age of menarche, parity, age at first birth, lifetime duration of breastfeeding, history of breast cancer in a first-degree female relative, menopausal status, age at menopause, female hormone use, body mass index (BMI), educational attainment, alcohol consumption, and cigarette smoking. Polytomous unconditional logistic regression was used to calculate odds ratios (ORs) and 95% confidence intervals (CIs) in pooled data and in each study separately. The likelihood ratio test was conducted to examine potential interaction. Between-study heterogeneity was assessed by performing meta-analyses with a random-effects model. Results. The study population included 14,106 African American women, of whom 1,839 had estrogen receptor-positive (ER+) breast cancer, 1,037 had estrogen receptor-negative (ER-) breast cancer, 495 had triple negative breast cancer (ER-, progesterone receptor-negative, and HER2 receptor-negative), and 10,061 were controls. ORs for ≥10 years duration of OC use were elevated for both ER+ and ER- breast cancer (ER+, OR=1.30, 95% CI 1.11-1.52; ER-, OR=1.59, 95% CI 1.31-1.94). OC use within the past 5 years was also associated with risk of both ER+ (1.57, 95% CI 1.27-1.94) and ER- (1.72, 95% CI 1.33-2.22) breast cancer. ORs for TN breast cancer were similar to those for ER- breast cancer. Results were consistent across age groups, including for early-onset breast cancer (diagnosed before age 40). The only evidence of multiplicative interaction was for OC use with BMI and the interaction was statistically significant for ER- breast cancer (p<0.01). For both ER+ and ER- breast cancer, OC use was not associated with risk among women with BMI <25. For ER- breast cancer, OC use was associated with increased risk among both overweight and obese women, whereas for ER+ breast cancer, the association was present only among obese women. Results from a meta-analysis of study-specific results were similar to results from the pooled data, with a p-value >0.05 indicating no statistically significant heterogeneity. Conclusion. Our results suggest that oral contraceptive use, particularly long duration and recent use, is associated with an increased risk of both ER+ and ER- breast cancer, and that obesity may modify the association. Potential mechanisms for this interaction will be discussed. Citation Format: Traci N. Bethea, Lynn Rosenberg, Chi-Chen Hong, Melissa A. Troester, Kathryn A. Lunetta, Elisa V. Bandera, Pepper Schedin, Virginia F. Borges, Andrew F. Olshan, Christine B. Ambrosone, Julie R. Palmer. Oral contraceptive use in relation to breast cancer subtypes in African American women: Results from the AMBER Consortium. [abstract]. In: Proceedings of the Sixth AACR Conference: The Science of Cancer Health Disparities; Dec 6–9, 2013; Atlanta, GA. Philadelphia (PA): AACR; Cancer Epidemiol Biomarkers Prev 2014;23(11 Suppl):Abstract nr B31. doi:10.1158/1538-7755.DISP13-B31