SAN ANTONIO — Two new statistical analyses of Women's Health Initiative data persuasively indicate that the recent abrupt decline in breast cancer incidence in the United States is attributable to a dramatic drop in the use of estrogen-plus-progestin menopausal hormone therapy, and not—as skeptics have argued—to less utilization of mammography.
Academic fencing over causality aside, the practical take-home message from the latest Women's Health Initiative (WHI) data analyses is that the breast cancer risk imparted by hormone therapy rises sooner and more steeply than previously recognized, and it swiftly declines after HT is discontinued, Dr. Rowan T. Chlebowski said at the San Antonio Breast Cancer Symposium.
“I think the good news for women here is that the risk rapidly dissipated in just a year or year and a half,” said Dr. Chlebowski, a medical oncologist at the Los Angeles Biomedical Research Institute in Torrance, Calif.
The WHI was an extremely large National Institutes of Health-sponsored study on the prevention of cardiovascular disease and other chronic diseases. The HT clinical trial portion was halted prematurely in July 2002 upon detection of an increased breast cancer rate in its treatment arm.
In response to this announcement, the annual number of prescriptions for menopausal HT in the United States plunged from 60 million in 2001 to 25 million in 2003.
This development was temporally related to a sudden 8.6% decrease in the national annual age-adjusted incidence of breast cancer (20,000 fewer cases per year) beginning in 2003.
The reduction has mostly affected the 50- to 69- year-old age group, and mostly has involved fewer estrogen receptor-positive tumors (N. Engl. J. Med. 2007;356:1670–4).
Critics argued that this breast cancer decline might be an artifact resulting from less use of mammography once women stopped taking HT. But in San Antonio, Dr. Chlebowski presented two new analyses—involving a total of nearly 57,000 WHI participants—that ruled out a change in mammography utilization as a significant causal factor.
One analysis involved 15,387 participants in the WHI randomized clinical trial who were assigned to 0.625 mg/day of conjugated equine estrogens plus 2.5 mg/day of medroxyprogesterone acetate, or to placebo. The breast cancer risk was 26% greater in the HT group, compared with placebo, during the 5.6-year intervention period, then declined by 27% post intervention.
During the final year before the intervention was halted, the annualized incidence of breast cancer was 0.61% in the HT arm and 0.41% with placebo. In the first year post intervention, the rates were 0.44% in the HT group and 0.36% in the control group.
Tellingly, there was virtually no difference in the proportion of patients in the two study arms who had a mammogram in the year before the study was halted, nor in the period beginning 2 years before, within 1 year after, or 2 years after, Dr. Chlebowski noted.
The second analysis involved 25,328 women who were not using menopausal HT at entry into the nonrandomized WHI observational study and 16,121 others who were taking estrogen plus progestin at entry.
During 2001, the HT users had a highly significant 79% increased risk of being diagnosed with breast cancer, compared with HT nonusers, after adjustment for age, body mass index, ethnicity, education, physical activity, Gail risk score, alcohol use, smoking status, family history of breast cancer, and other potential confounders. During 2002, the HT users had a 65% increased risk.
However, during 2003—after all WHI participants had been instructed to stop taking HT—women who were HT users at entry into the observational study had only a nonsignificant 18% increased breast cancer risk compared with HT nonusers at entry. In 2004, the women who had been on HT at entry had a 20% increased incidence of breast cancer, again statistically nonsignificant. (See box.)
In each arm of the observational study, the percentage of patients who had a mammogram remained unchanged every year during 2001–2004.
Among nearly 2,000 WHI observational study participants who were diagnosed with invasive breast cancer, the 5-year mortality was roughly 4% regardless of whether they had been taking estrogen plus progestin or were HT nonusers, according to Dr. Chlebowski. This finding stands in marked contrast to a report from the California Teachers Study (also presented during the San Antonio conference), which concluded that breast cancer patients who had been on HT had a significantly better prognosis than did HT never-users.
Session cochair Dr. Peter M. Ravdin of the University of Texas M.D. Anderson Cancer Center, Houston, noted that the initial WHI report of an increased breast cancer risk in HT users prompted a change in guidelines for prescribing menopausal HT, with only short-duration therapy being recommended as safe. Will the new WHI findings lead to further change in the guidelines? he asked.
“I think these results will tend to focus attention on what ‘short duration’ means,” Dr. Chlebowski replied. “One part of the story is that the risk is greater than we thought with longer-duration therapy. We can see the incidence curve going up starting in year 2 of the observational study. … So there's now a strong incentive for women to take HT for shorter intervals. I think the threshold for starting HT should be higher; there should be more attention to the duration, and I think interventions other than HT for climacteric symptoms should be a research priority.”
Dr. Chlebowski disclosed that he is on the speakers bureaus for AstraZeneca Pharmaceuticals LP, Abraxix BioScience Inc., and Novartis, and serves as a consultant to AstraZeneca, Novartis, Genentech Inc., and Eli Lilly & Co.
FULL TEXT
