Statin use and breast cancer: prospective results from the Women's Health Initiative.

BACKGROUND Despite experimental observations suggesting that 3-hydroxy-3-methylglutaryl coenzyme A inhibitors (statins) have antitumor activity, clinical studies have reached mixed conclusions about the relationship between statin use and breast cancer risk. METHODS To investigate associations between potency, duration of use, and type of statin used and risk of invasive breast cancer, we examined data for 156,351 postmenopausal women who were enrolled in the Women's Health Initiative. Information was collected on breast cancer risk factors and on the use of statins and other lipid-lowering drugs. Cox proportional hazards regression was used to calculate hazard ratios (HRs) with 95% confidence intervals (CIs). Statistical tests were two-sided. RESULTS Over an average follow-up of 6.7 years, 4383 invasive breast cancers were confirmed by medical record and pathology report review. Statins were used by 11,710 (7.5%) of the cohort. Breast cancer incidence was 4.09 per 1000 person-years (PY) among statin users and 4.28 per 1000 PY among nonusers. In multivariable models, the hazard ratio of breast cancer among users of any statin, compared with nonusers, was 0.91 (95% CI = 0.80 to 1.05, P = .20). There was no trend in risk by duration of statin use, with HR = 0.80 (95% CI = 0.63 to 1.03) for < 1 year of use, HR = 0.99 (95% CI = 0.80 to 1.23) for 1- < 3 years of use, and HR = 0.94 (95% CI = 0.75 to 1.18) for > or = 3 years of use. Hydrophobic statins (i.e., simvastatin, lovastatin, and fluvastatin) were used by 8106 women, and their use was associated with an 18% lower breast cancer incidence (HR = 0.82, 95% CI = 0.70 to 0.97, P = .02). Use of other statins (i.e., pravastatin and atorvastatin) or nonstatin lipid-lowering agents was not associated with breast cancer incidence. CONCLUSIONS Overall statin use was not associated with invasive breast cancer incidence. Our finding that use of hydrophobic statins may be associated with lower breast cancer incidence suggests possible within-class differences that warrant further evaluation.