Endometrial cancer is clearly a hormonally responsive
tumor, with a critical role played by
estrogens unopposed by
progestins. Numerous epidemiologic studies have shown substantial risk increases associated with use of unopposed
estrogens, especially among thin women. This risk, however, can be reduced if
progestins are added to the
therapy. The manner in which
progestins are prescribed is a critical determinant of risk. Most studies show that women who have ever used
progestins continuously (>25 days/months) are at somewhat reduced risk relative to non-users (meta-analysis relative risk, RR, based on observational studies=0.78, 95 confidence intervals, CI, 0.72-0.86). The reduced risk in greatest among heavy women. In contrast, women who have ever used
progestins sequentially for <10 days each month are at increased risk, with meta-analysis results showing on overall RR of 1.76 (1.51-2.05); in contrast,
progestins given for 10-24 days/month appear unrelated to risk (RR=1.07, 0.92-1.24). These risks were based on varying patterns of usage, with little information available regarding how
endometrial cancer risk is affected by duration of use, type and/or dose of
estrogen or
progestin, or mode of administration. Effects may also vary by clinical characteristics (e.g., differences for Type I
vs. II tumors). Further resolution of many of these relationships may be dependent on pooling data from multiple studies to derive sufficient power for subgroups of users. With changing clinical practices, it will be important for future studies to monitor a wide range of exposures and to account for divergent effects of different usage patterns. This article is part of a Special Issue entitled 'Menopause'.