SAN ANTONIO – Five years of tamoxifen provided 20 years of breast cancer prevention to some at-risk women who took it prophylactically.
However, their 20-year all-cause mortality was no different from those taking placebo (182 vs. 166 deaths), nor was their mortality from breast cancer (31 vs. 26, respectively), Jack Cuzick, Ph.D., said at the San Antonio Breast Cancer Symposium.
Dr. Jack Cuzick
“While we saw clear, lasting benefits of tamoxifen in reducing breast cancer incidence, uncertainty with respect to mortality remains,” said Dr. Cuzick, the John Snow professor of epidemiology at Wolfson Institute of Preventive Medicine at Queen Mary University, London.
He suggested that, in light of the small number of deaths, the study was not sufficiently powered to detect any significant survival difference. But women in the IBIS-1 trial will continue to be observed, and future analyses could clarify the issue, he added.
“Although 20 years seems like a long follow-up time, it is actually too early to make any clear statement about mortality,” he said. “However, we are concerned about an excess emergence of ER-negative tumors, which we saw after 10 years.”
The study was simultaneously published on Dec. 13 in Lancet Oncology
IBIS-1 randomized 7,154 healthy pre- and postmenopausal women to 5 years of either 20 mg daily tamoxifen or placebo. These women were aged 35-70 years at baseline and presented with an increased breast cancer risk attributable to a family history.
IBIS-1 was first reported in 2002, when 4-year follow-up found a 32% reduction in breast cancer risk associated with tamoxifen. But it also found a significant increase in deaths among that group (25 vs.11 in placebo), many of which were attributable to more endometrial cancer (11 vs. 5).
In 2007, the investigators published the trial’s 8-10 year findings. At that time, tamoxifen was associated with a 27% decreased risk of any breast cancer, and a 34% decreased risk of developing an ER-positive cancer. It found a consistent prophylactic benefit for tamoxifen; most of the risk reduction, however, occurred in the 5-year active treatment phase.
The significant increase in death had continued (25 vs. 11), although that report indicated that no specific cause, including endometrial cancer, drove that finding.
At the meeting, Dr. Cuzick discussed the latest findings of IBIS-1, which has now followed the cohort for up to 20 years (median, 16 years).
“We have seen a continued separation of the cancer incidence curves, with a 20-year incidence of 7.8% vs. 12.3%,” Dr. Cuzick said. “We saw that the 30% overall reduction was maintained; the number needed to treat to prevent one breast cancer was 22, which is very favorable when compared to prevention strategies for other diseases.”
The incidence of ER-positive tumors was reduced, compared with placebo, he added (4.9% vs. 8.3%), with a number needed to treat of 29. However, Dr. Cuzick said, there was a slight increase in the incidence of ER-negative tumors after 10 years.
“This is likely because these are tumors that would have appeared earlier as ER-positive tumors, but under tamoxifen, they were held back for some time and eventually broke out as ER-negative tumors.”
Hormone therapy was allowed in IBIS-1, and about 50% of women were taking hormones during at least part of the study. They did not experience the same level of benefit form tamoxifen as those who were not taking hormone therapy (12% vs. 38%). “This is very clear evidence that the benefits are substantially greater in those who are not using concurrent [hormone therapy] when on tamoxifen,” he said.
However, tamoxifen was associated with side effects and risks, including an increased risk of developing other cancers. Most concerning was the 45% increase in the risk of endometrial cancer, which accounted for five deaths in the treatment group; there were no deaths from endometrial cancer in the placebo group.
“We had hoped that endometrial cancer might not translate into such a large morality increase as there were no deaths due to this in the 8-year follow-up. We do need to be aware of this.”
Recurrent breast cancer was the single largest cause of death, but the between-group difference was not significant (26 vs. 31). Cardiovascular deaths were similar between the groups, and all that did occur, did so during the treatment period.
Nonmelanoma skin cancers were 39% more likely in the tamoxifen group, although they caused no deaths. “This was a very large increase and a surprise; we really don’t understand it,” Dr. Cuzick said.
There were 12 fewer colorectal cancers in the tamoxifen group. “That there was a slight reduction here is of some interest, as we did not see at all with placebo.”