Hormone replacement therapy (HRT) uses reproductive hormones to ameliorate unpleasant symptoms of menopause, such as hot flushes, lack of sex drive and depression. Estrogen is the main hormone used to alleviate symptoms; however, estrogen-only therapy increases risk of endometrial cancer and is recommended only in women with hysterectomy. Progesterone is added to reduce endometrial cancer risk for women with a uterus and this estrogen-progesterone cocktail is known as combination HRT.
In 1993 to 1998, the Women’s Health Initiative (WHI) had investigated the potential of long-term HRT to prevent coronary heart disease and osteoporotic fractures. The original placebo-controlled trial was intended to have a nine-year follow-up. However, the study that used combination HRT had to be halted after an average follow-up of only 5.2 years due to significant increases in the risk of breast cancer and cardiovascular disease. Ancillary studies have established that women on combination HRT undergo a significant increase in breast density, a known risk factor of breast cancer—but is that what underpinned the increased cancer incidence seen by WHI?
The publication from Celia Byrne and colleagues provides the first evidence in that direction. This nested case-control study looked at 174 women who developed breast cancer within the WHI cohort, as well as 733 disease-free controls. Women were followed up for an average of 7.9 years. Breast density was measured by either Cumulus or Madena at one year after trial entry.
Women on HRT had a mean 9.49% increase in density, while women in the placebo arm saw a density decline of 0.65%. Women taking HRT (but not placebo) underwent a 3% increase in breast cancer risk for every 1% increase in density. Women who took HRT and underwent the greatest increase in density were at greatest risk; women in the top quintile underwent an increase in density of 19.32% or greater, and were at 3.61-fold greater risk of breast cancer, compared to women in the bottom quintile of density change. Interestingly, a woman’s baseline density also affected whether she was at increased risk (again, only if she took HRT). A 1% increase in density lead to a 4% increase in risk for women whose baseline density was below 10%. The same 1% density increase conferred an 8% increase in risk for women whose baseline density was between 10% and 25%. However, a 1% increase in density in women with baseline density of 25% or greater conferred no additional risk. It is speculated that women with very high baseline density are already at high risk, and thus do not experience a further risk increase. However, density may be harder to measure accurately in these ranges and this effect could thus be due to measurement error.
Most interestingly, HRT takers had 28% greater odds of having breast cancer compared to placebo (although this increase did not quite reach significance in this cohort). Strikingly, this increase in risk completely disappeared once you controlled for the change in density. This suggests that HRT mediates its breast cancer-promoting effects essentially entirely through increased density.
The implication of these results is that an increase in density in response to HRT can be used as a marker to indicate a likely increase in breast cancer risk. However, the method for measuring breast density is important. Broad categorizations of density, such as BI-RADS, are unlikely to be sufficiently sensitive to detect the density changes observed in this study. A continuous density measure is thus much preferable to assess these density alterations in response to HRT, and uncover if a woman is likely to be at increased risk.
