More negative fallout from a landmark government study suggests breast cancer linked to estrogen-progestin pills may be fast-growing and hard to detect, clarifying risks for millions of women still using hormone treatment.

"Hopefully, it will convince women to reconsider," said Dr. Susan Hendrix of Wayne State University in Detroit, a co-author of the new analysis. "We've got to find a better way to help women with their menopausal symptoms."

Some previous studies suggested breast tumors might be less aggressive in hormone users; other studies indicated the opposite. Previous research also suggested that hormones might make breast tissue more dense, hindering the detection of tumors.

Seeking more definitive answers, the researchers took a closer look at data from the government's landmark Women's Health Initiative (search) study, which was halted last summer after it was found that estrogen-progestin pills (search) raise the risk of heart attack, strokes and breast cancer.

While last summer's findings led many women to stop taking hormones, Hendrix said an estimated 3 million women still use them, primarily to relieve hot flashes and other symptoms of menopause.

The latest findings appear in Wednesday's Journal of the American Medical Association.

The analysis involved 16,608 women ages 50 to 79 who used either combined hormone treatment or dummy pills for an average of five years.

As of January, breast cancer had developed in 245 women who used the combined hormone treatment and in 185 women who had taken dummy pills.

Hormone users' tumors were larger at diagnosis, 1.7 centimeters on average versus 1.5 centimeters in placebo women. Tumors had begun to spread in 25.4 percent of hormone users, compared with 16 percent of placebo women.

The researchers said this appears to mean that in women on estrogen-progestin, the tumors both grow faster — that is, they are more aggressive — and escape detection longer.

Overall, women on both hormones faced a 24 percent increased risk of breast cancer (search) — equal to eight extra cases of cancer per year for every 10,000 women taking the pills.

The increased risk did not appear in the first two years of treatment. But Hendrix said the tumors may have been present early on but were not detected until later because of hormone-induced breast density.

The new analysis did not examine breast density. But researchers think progestin may be the culprit because it can cause breast cells — both normal and abnormal — to proliferate, an effect that may be accentuated when the hormone is combined with estrogen.

Wyeth Pharmaceuticals, maker of the Prempro pills used in the study, said hormones remain an appropriate therapy when used at the lowest possible dose for the shortest possible time.

The latest analysis is by far the most conclusive, said Dr. Peter Gann, an associate professor of preventive medicine at Northwestern University who was not involved in the study.

It "further worsens the news for long-term hormone replacement therapy. It suggests the excess breast cancer risk is not trivial," Gann said.

Last summer's Women's Health Initiative findings shattered long-held beliefs that hormones are good for women's hearts. Last month, another analysis of data from the study found that instead of sharpening the mind, hormones may double the risk of Alzheimer's and other forms of dementia.

A second, smaller study in Wednesday's journal also confirmed a link between combined hormone treatments and breast cancer and suggested estrogen-only treatment may be safer.

The study involved 975 Seattle-area women ages 65 to 79. The greatest breast cancer risk was in women who used estrogen-progestin for at least five years, even if they took the progestin component only some days a month.

Those who used estrogen alone, even for 25 years or longer, showed no appreciable increased risk, according to the study, led by Dr. Christopher Li of Fred Hutchinson Cancer Research Center in Seattle.

Estrogen alone is recommended only for women with hysterectomies because it can cause uterine cancer unless balanced by progestin.

The researchers said more definitive answers will come from the continuing estrogen-only part of the Women's Health Initiative study.