Tamoxifen (TAM) has been prescribed for decades and aromatase inhibitors (AIs) have been used since the early 2000s in preventing subsequent breast cancer. However, outside of clinical trials, the effectiveness of AIs is not established. We examined the long-term risk of subsequent breast cancer among survivors treated with TAM and AIs in a large health plan. The study included 22,850 survivors, diagnosed with initial breast cancer (stages 0-IV) from 1996 to 2006, and followed 13 years maximum. We compared the risk of subsequent breast cancer in those who used TAM and/or AIs versus nonusers (the reference group). Hazard ratios (HR) adjusted for patient, tumor, treatment, and health-care characteristics were estimated using Cox models with time-dependent drug use status. Women who used TAM/AIs had a large reduction in risk of subsequent breast cancer compared with nonusers. While confidence intervals (CI) for all hormone treatment groups overlapped, women with high adherence (medication possession ratio ≥80%) who used AIs exclusively and had positive ER or PR receptor status had the greatest risk reduction (HR = 0.34, 95% CI: 0.28-0.41), followed by those who switched from TAM to AIs (HR = 0.39, 95% CI: 0.30-0.49), and those who used TAM exclusively (HR = 0.42, 95% CI: 0.36-0.47). Women with high adherence had the greatest risk reduction in subsequent breast cancer, but the results were not substantially different from women who took the drugs less regularly. Compared with nonusers, the reduction in subsequent breast cancer risk ranged from 58% to 66% across the hormone treatment groups and degree of adherence.
Keywords: AI; breast cancer; comparative effectiveness; tamoxifen.