After rising for more than two decades, the incidence of breast cancer in the United States decreased sharply in 2003 and remained low for another year, researchers are reporting.

The researchers attribute the drop to a decrease in the use of hormone replacement therapy (HRT) that occurred at about the same time. Starting in mid-2002, millions of women stopped using HRT after an NIH study linked certain hormones to health risks, including breast cancer.

In a new analysis of the 2003 decrease, Dr. Donald Berry of the University of Texas M.D. Anderson Cancer Center and his colleagues say that the 2004 incidence figures reinforce their theory that a change in HRT use drove the decline in breast cancers.

Their findings appear in the April 19 New England Journal of Medicine.

Last December, the researchers reported a possible association between HRT use and breast cancer after the incidence rate declined nearly 7 percent between 2002 and 2003. But the 2004 rate was not yet available, leaving open the possibility that the 1-year decline was a statistical fluke.

"We now know that the decline persisted into 2004, and this suggests that it was a real decline and not a single-year anomaly," says co-author Dr. Kathy Cronin of NCI's Division of Cancer Control and Population Sciences (DCCPS).

The 2004 data also suggest that the incidence trend is following the HRT trend, she adds. The declines occur very close together in time and they take the same form, each starting with a steep drop before leveling off.

The decrease in breast cancers was seen only among women age 50 or older, which is the group most likely to use HRT. And the decrease was more evident in estrogen receptor-positive cancers, which may depend more on hormones for growth than estrogen receptor-negative cancers.

The researchers believe that withdrawing hormones may have had the effect of slowing the growth of some small cancers and preventing their detection on mammograms. But these cancers might eventually grow and be found later, leading to higher incidence rates in the future.

"We need to be clear to women that this is a decrease in breast cancer incidence, and we have no evidence that it will lead to a decrease in mortality," notes Dr. Berry. He points out that the tumors most affected by hormones are "exquisitely sensitive" to today's drugs.

Though other factors may have contributed to the decline, the authors found little evidence that changes in screening habits or prevention could explain large amounts of the 2003 decline.

For example, none of the drugs used to prevent breast cancer, such as tamoxifen, was taken by a substantial portion of postmenopausal women or showed a substantial change in use from 2000 to 2004.

Women who are still taking hormones should consider stopping them, says co-author Dr. Christine Berg of NCI's Division of Cancer Prevention. If hormones are needed to treat severe hot flashes and other symptoms, they should be used at the lowest dose for a limited amount of time, she adds.

With the 2004 data, Dr. Berry is more confident that HRT use may account for much of the unexplained rise in U.S. breast cancers that has occurred over the last 30 years or so. The increase is above and beyond what could be explained by mammography, and a number of factors are thought to play a role, including HRT use.

By Edward R. Winstead