Tamoxifen is proven to prevent reoccurrence of breast cancer by almost half. The price it exacts for that progression-free survival (PFS) benefit can be life-threatening uterine cancer, stroke, deep vein thrombosis, pulmonary embolism, or deep vein thrombosis. More common are potentially debilitating hot flashes, eye damage, and a variety of menopausal annoyances. However, what is absolutely clear from long-term studies is that there is no actual survival benefit.
Tamoxifen is a selective estrogen receptor modifier (SERM). It was at one time the standard of care for estrogen-receptor positive breast cancer. It was prescribed for up to five years, but studies suggested the benefits plateaued out at that point. However, this was revised, and it is often given for 10 to 15 years now. Oncologists recognize the peculiar phenomenon of very late breast cancer recurrence up to 20 years after diagnosis and apparently successful treatment. To be a true chemoprevention tamoxifen would have to have benefit over a span of decades.
In fact, the death rate in women on Tamoxifen outstrips placebo or that of women with similar breast cancers who do not receive Tamoxifen. The excess deaths in the Tamoxifen group in two large studies did not reach statistical significance, but there was certainly no overall survival benefit.
The common interpretation of this outcome is that while Tamoxifen lowers the number of women experiencing a relapse at 10 to 15 years, those who do relapse tend to have a more aggressive and fatal form of the cancer. Put plainly, this drug does not prevent breast cancer deaths.
Numbers needed to treat (NNT) to produce a response, namely PFS, varies from the same as the number of patients needed to harm (NNH) to several-fold less, depending on the risk category of the breast cancer and morbidity or mortality in question. This may give the impression that Tamoxifen has clinical benefit. However, if you examine life-threatening risks against lack of life-saving benefit, the value of this drug looks starkly different.
References:
Kinsinger, L et al. Systematic Evidence Reviews, No. 8. Rockville (MD): Agency for Healthcare Research and Quality (US); 2002 Jul
http://www.ncbi.nlm.nih.gov/books/NBK42583/
http://www.ncbi.nlm.nih.gov/books/NBK42585/
