Methods: Using a Monte Carlo model, we compared the quality-adjusted life expectancy (QALE) associated with breast cancer chemoprevention with tamoxifen or raloxifene versus usual care in 60-year-old white women at average risk of invasive breast cancer (5-year probability 1.9%). The model considered risks of invasive breast cancer (including ER status), endometrial cancer, DVT, PE, and stroke. Meta-analysis of four randomized controlled trials (BCPT, MORE, RUTH, STAR) provided relative risk (RR) estimates. SEER data were used for cancer incidence and mortality. Quality of life estimates were from the literature. Results are undiscounted.

Results: Because of trade-offs between risks and benefits, no strategy was consistently preferred in the base case. However, with a 4% 5-year risk of invasive breast cancer, 5 or 10 years of either drug increased QALE by 71 and 72 days, respectively. The gains increased to 106 and 113 days with a 6% 5-year risk of breast cancer. Even among women at 10-fold increased risk of endometrial cancer, 5 years of raloxifene therapy increased QALE by only 13 days compared with tamoxifen because of the low baseline incidence of endometrial cancer. In women with both a 6% 5-year risk of breast cancer and a 10-fold endometrial cancer risk, raloxifene provided 127 and 15 additional quality-adjusted life days compared with placebo and tamoxifen, respectively. The results were sensitive to the disutility associated with chemoprevention (base case=0.995), with usual care favored for values falling below 0.985. For raloxifene, benefit was mitigated most by its risk of stroke, followed by DVT/PE, and then endometial cancer. In contrast, endometrial cancer was the most important risk for tamoxifen, followed by stroke, and then DVT/PE.

Conclusions: Based on our model, breast cancer chemoprevention appears to be a toss-up versus usual care for average risk 60-year-old women. For women at higher risk of breast cancer, chemoprevention provides a modest survival advantage. The risk of endometrial cancer appears to have little impact on longevity estimates, but the benefit of breast cancer chemoprevention may be offset by patient preferences regarding their quality of life taking these medications and by their risk of stroke.