HRT and Risk of Breast Cancer: Another Look at the Data

NEW ORLEANS—Hormone replacement therapy (HRT) remains strongly associated with the development of breast cancer in the minds of many. But the discerning clinician should “go beyond P values and relative risk” and use the known data to make decisions regarding this issue, John C. Arpels, MD, said at the American Society of Breast Disease annual meeting.

His presentation was an attempt, he said, to bring clinical relevance to the lack of definitive data on whether HRT causes a significant increase in the risk of breast cancer. Dr. Arpels is associate clinical professor of OBGYN and reproductive sciences, University of California, San Francisco.

The existing evidence shows an overall 46% decrease in excessive mortality among HRT users. HRT users have an average 50% decrease in hip and Colles fractures plus a 90% decrease in vertebral fractures. The rate of coronary heart disease is reduced by 40% overall, and seven preliminary studies show 50% less senile dementia in those on HRT, he said.

In contrast, some studies show an increase in breast cancer risk among HRT users. However, no consistent trend of increased risk has emerged among 63 major published studies.

A recent re-analysis of the data from 51 epidemiologic studies by the Collaborative Group on Hormonal Factors in Breast Cancer showed a modest 58% increase in breast cancer risk, which became significant only with 15 years of use and actually translated into only 337 cases out of the total 52,700 breast cancers (Lancet 350:1047, 1997).

The authors of this study, who re-computed the raw data from these trials, calculate that this increase in risk projects to only 7 to 12 excess cases of breast cancer per 1,000 women aged 65 to 75 years using long-term HRT. Of further interest, this excess risk disappeared completely when HRT was discontinued for 5 years. Not until age 60 was there a significant difference between never-users and users of 5 to 10 years.

“One is left to speculate on how, once HRT has helped to initiate oncogenic transformation in breast epithelium, the cancer is ‘cured’ by the mere cessation of further hormonal use,” he said. “How, if you start the process then stop it, does the cancer just melt away? The doubling time for the average cancer is such that it may take 6 to 8 years to pick up a tumor, so cancer is occurring before women are on HRT.”

Also, he said, if cancer is occurring, “why did the study find a decrease in relative risk at years 10 to 14 of use as compared to years 5 to 9 of use (1.09 and 1.19, respectively)?”

A number of biochemical factors may also be confounding the association between HRT and breast cancer, Dr. Arpels pointed out. The presence of the P450 aromatase enzyme in breast tissue can convert androgens to endogenous breast estrogen against a serum gradient. And estrogen levels found in breast duct fluid can be 10- to 40-fold higher than estrogen levels in serum, even in women not on exogenous HRT. “So if estrogen is bad, the breast already has it,” he noted.

Additionally, among the different estrogen products, there are differences in nuclear retention times, bound vs bioactive fractions, and cofactors such as insulin-like growth factors. These observations should be factored into clinical trials seeking the true risk of HRT on breast cancer occurrence, he said.

Two large prospective trials—a European trial and the multicenter Women’s Health Initiative—may shed light on the association between HRT and breast cancer risk. Until then, Dr. Arpels said, it should be noted that 10 studies have shown a 10% to 40% increase in survival among women using HRT or oral contraceptives when their breast cancer was diagnosed, probably because of estrogen-related events that decrease the risk of metastatic spread.

These factors include antioxidative effects, enhancement of gap junction genes that maintain control of cell diversity, increases in humoral B-cell activation, and upregulation of the normal BRCA1 protein that prolongs tumors in a premalignant state.

Dr. Arpels also stressed that eight studies have found no difference in disease-free interval or death rate when HRT is given to breast cancer survivors.

The future holds the promise of agents similar to estrogen that may exert both anticancer as well as other favorable effects. A number of selective estrogen-receptor modulators (SERMs) as sequels to raloxifene (Evista) are under investigation. Plant biphenolic isoflavone phyto-estrogens are being studied as potential substitutes for steroidal estrogens and have shown antimitogenic biochemical properties in monkeys. And Dr. Arpels predicts “a great future” for targeted tissue-specific estrogens being developed in rodent models.

“We will probably have substitutes for estrogen long before we will know the answer to the question of whether estrogen causes breast cancer,” he commented.

[Editors’s Note: The Iowa Women’s Health Study of 37,105 postmenopausal women (JAMA June 9, 1999) showed no increase in the more common breast cancer types (DCIS or invasive ductal or lobular cancer) in HRT users. However, women taking HRT for 5 years or less had an 81% increased risk of medullary, papillary, tubular, and mucinous breast tumors, and those who used HRT for 5 years or more had an increased risk of 265%.]