Breast cancer and hormone replacement therapy: putting the risk into perspective

Abstract

Data on hormone replacement therapy and breast cancerrisk come from a number of observational studies (mostlyAmerican studies). Those published up to 1995 werereanalyzed by the Collaborative Group on HormonalFactors in Breast Cancer (CGHFBC). They involvedpopulations where exceedingly high estrogen doses wereused as first-line therapy, and a progestin was added in aminority of women. Overall, the CGHFBC reanalysisfound that the relative risk increased by 0.023 for eachyear of use (with an absolute excess risk of two or six casesout of 1000 women treated for 5 or 10 years, respectively).Further American studies, published in 2000and involving populations where lower doses were used,showed a risk increase of 0.01 per year of estrogen-onlyuse. Both the CGHFBC reanalysis and the furtherstudies did not find an increase of risk in treated overweightwomen. Possibly, overweight women already havea maximal estrogenic stimulus on the breast due toextraglandular estrogen production. An additionalexplanation could be that oral estrogens, through theirhepatocellular effects, reverse some biological features ofobesity (e.g. decreased sex hormone binding globulin leveland increased insulin-like growth factor-I bioactivity) thatpotentially increase breast cancer risk, so balancing theestrogen stimulation. The CGHFBC reanalysis did notshow a substantial difference in breast cancer risk betweenthe majority using estrogen alone and the small minorityusing estrogen plus progestin. Conversely, Swedish studiesand the recent American studies suggest that the riskincrease could be higher with the addition of a progestin,compared with estrogen-only use. The biological effect ofprogesterone/progestins on the breast tissue is controversial.Even if the observed increase in risk could be partiallyascribed to non-progesterone-like effects of some progestins(e.g. opposing the hepatocellular effects of oral estrogens)and also (in the American studies) to use-bias, a detrimentalaction due to progesterone-like effects cannot beexcluded. However, the theoretical possibility exists thatlow doses of oral estrogens, plus a progestin providingprogesterone-like effects only, will be shown to be associatedwith a limited breast cancer risk increase.