ABSTRACT
Data were collected from all identified randomized clinical trials of adjuvant tamoxifen (or cytotoxic) therapy for early breast cancer that began before 1985. Randomization required tamoxifen to be compared with no treatment, or tamoxifen plus treatment A versus treatment A. In 28 trials of adjuvant tamoxifen, nearly 4000 of 16 513 women had died. Twenty-five of the trials involved tamoxifen use for 2 years or less. The data on individual patients were subjected to novel overview statistical analysis. There was a clear reduction in mortality by about one-fifth with tamoxifen (p 0.000 1) at 5 years: 68.2% of women on no tamoxifen survived compared with 73.0% on tamoxifen. This difference was accounted for by the difference in women aged 50 years or more. For younger women there was no demonstrable difference, but there were relatively few women in this comparison (n = 1062). It was notable that for those younger women receiving tamoxifen alone there was a strong trend to a benefit, but those receiving chemotherapy as well obtained no benefit, possibly because of the ovarian effects of cytotoxics.
