Abstract Background: Digoxin, a cardiac glycoside derived from Digitalis purpurea, has been used for centuries to treat heart failure and atrial fibrillation. Early evidence suggested that digoxin attenuates breast tumor aggressiveness and could become an antineoplastic drug. Two recent Danish studies did not support this hypothesis, showing an approximately 30% increased rate of breast cancer in digoxin users. One of these studies noted a stronger association with incidence of estrogen receptor (ER)-positive breast cancer than with incidence of ER-negative disease. Neither Danish study could account for mammography usage or for potential confounding by lifestyle-related breast cancer risk factors such as adiposity, postmenopausal hormone usage, and reproductive history, all of which hampers a causal interpretation of their findings. To address this evidence gap, we estimated the association between digoxin use and breast cancer incidence in the Nurses' Health Study (NHS) cohort. Methods: The NHS began in 1976 with the enrollment of 121,700 female U.S. registered nurses, who have completed biennial questionnaires to update demographic characteristics, lifestyle, dietary, and pharmaceutical exposures, and newly diagnosed illnesses. The current study followed cancer-free postmenopausal participants from 1994 (the year in which digoxin was first ascertained) until the end of 2010. Incident breast cancers were reported on questionnaires and confirmed by medical record review. We fit multivariable Cox proportional hazards models to estimate associations between time-varying digoxin use and breast cancer incidence. Competing risks models were used to assess heterogeneity of associations by tumor ER status. We conducted a secondary analysis to evaluate possible confounding by indication through use of an alternative reference group, comprised of digoxin-unexposed women who took other cardiac drugs (e.g., beta blockers). Results: Our study included 90,202 postmenopausal women who contributed approximately 1.05 million person-years of observation. There were 4,876 women who reported using digoxin during the follow-up period, among whom 144 cases of invasive breast cancer were diagnosed. Compared with non-users, digoxin users had lower average BMI, were more likely to have undergone regular screening mammograms, and were more likely to take aspirin, beta blockers and lipid lowering drugs. The exposed and unexposed groups were similar with respect to reproductive history, alcohol consumption, and use of postmenopausal hormones. Current digoxin use of greater than four years' duration was associated with a 49% increase in breast cancer incidence compared with never use (HRadj=1.49, 95% CI: 1.16, 1.91). This association was somewhat stronger for ER-positive breast cancer (HRadj=1.52, 95% CI: 1.14, 2.02) than for ER-negative breast cancer (HRadj=1.13, 95% CI: 0.53, 2.41). Current users with four or fewer years of exposure had a 31% higher incidence of ER-positive breast cancer (HRadj=1.31, 95% CI: 1.01, 1.69), and former users had about the same breast cancer risk as never users. The association was robust to adjustment for lifestyle-related breast cancer risk factors and to restriction to women who underwent regular screening mammograms. Furthermore, we saw no evidence of confounding by indication. Conclusions: This study confirms previously reported associations between current digoxin use and the incidence of breast cancer. We demonstrate that this association is not attenuated when lifestyle-related breast cancer risk factors and screening practices are accounted for. The considerable worldwide prevalence of digoxin use and the availability of alternative drugs may present an opportunity for primary prevention of a substantial number of breast cancer cases. Citation Format: Thomas P. Ahern, Rulla M. Tamimi, Bernard A. Rosner, Susan E. Hankinson. Digoxin use and risk of invasive breast cancer: Evidence from the Nurses' Health Study. [abstract]. In: Proceedings of the Twelfth Annual AACR International Conference on Frontiers in Cancer Prevention Research; 2013 Oct 27-30; National Harbor, MD. Philadelphia (PA): AACR; Can Prev Res 2013;6(11 Suppl): Abstract nr B55.
Read full abstract