Cover of Hormone Replacement Therapy and Breast Cancer

Hormone Replacement Therapy and Breast Cancer

Systematic Evidence Reviews, No. 14

, MD, MPH.

Oregon Health Sciences University Evidence-based Practice Center, Portland, Oregon

Structured Abstract


To evaluate and update the prior review evaluating the risk of breast cancer and breast cancer (BCA) death associated with the use of postmenopausal hormone replacement therapy (HRT) by reviewing the medical literature which has been published since the last US Preventive Services Task Force update.

Data Source:

All English language studies identified in the Medline database from 1992-2000 and all previously published meta-analyses. In addition, reference lists of key articles, letters, and editorials were reviewed for all related studies, including those pre-dating the database search.

Study Selection:

All studies that evaluated breast cancer incidence or mortality as a primary or secondary outcome in association with hormone replacement therapy published between 1992-2000. Studies evaluating the effect of hormone replacement therapy on breast density were also reviewed.

Data Extraction:

The following studies met inclusion criteria: 8 meta-analyses from the years 1988-1997, 1 nested case-control study, 14 case-control studies, and 15 cohort studies all evaluating breast cancer incidence, mortality, or both. Of the 15 cohort studies, 10 represented unique cohorts and of the 14 case control studies, 2 involved updates of the same case set. Data from each study were abstracted to prepared forms. When more than one study from the same population was reported, data from the most recent publication were reviewed. If data from the same population were analyzed by cohort and by case-control analysis, both results were reported if they evaluated different outcomes. In addition, several studies evaluating breast density and HRT were reviewed, and the best studies summarized.

Data Synthesis:

For ever or short-term use of estrogen, 7 of the 8 meta-analyses, 8 of the 11 case-control studies, and 6 of the 7 cohort studies evaluating incidence showed no increase in breast cancer with hormone replacement therapy. Of the original studies reviewed evaluating incidence, 12 of 19 showed no increased risk of breast cancer with long duration ERT or HRT use. However, 5 of the meta-analyses showed increased risk with duration over 5 years and 2 important cohort studies showed increased risk with longer duration use. Eleven original studies evaluated combined estrogen and progestin, and one showed increased risk of BCA with short-term use; 3 of the 5 evaluating duration with combined therapy showed increased risk that was statistically significant. Current use of ERT was associated with significantly increased risk of breast cancer in two of the best cohort studies; use of combined therapy was associated with increased risk in 3 studies. Six recent cohort studies (1992-2000) evaluated breast cancer mortality in association with hormone use: 1 showed increased risk of death, 4 showed decreased risk of death, and one showed no association. Several recent studies show that post-menopausal estrogen therapy is associated with increased breast density by mammography and that adding progesterone to estrogen results in even greater increases in breast density. Finally, there is evidence suggesting an important interaction between HRT and alcohol use and HRT and lower body weight.


The association of short-term hormone replacement therapy with the development of breast cancer is uncertain based on multiple studies with inconsistent findings. Among studies indicating increased risk, the risk is largely confined to current and long-term use (>5–10 years), and the risk is relatively small (RR 1.2–1.5). Reduced mortality is a fairly consistent finding among the studies evaluating breast cancer mortality and HRT use. The addition of progesterone to estrogen and current, as well as long-term, use may be associated with breast cancer risk above that of estrogen itself. Although the biological plausibility of an association between postmenopausal hormone use and breast cancer is high, the studies showing risk or benefit from the use of post-menopausal hormones are limited by the observational nature of the epidemiologic data existing to date. Data from randomized controlled trials are needed to validly evaluate the relationship.