Final results from a big study comparing two drugs for preventing breast cancer in high-risk women reveal surprises that challenge the government's claim that one is clearly better.

The study compared the old standby, tamoxifen, to raloxifene, a newer drug so far approved only for preventing the bone disease osteoporosis. The government contends raloxifene is safer.

At a news conference in April, the National Cancer Institute, which paid for the $88 million study, said both drugs were equally effective at lowering the risk of serious forms of breast cancer. But raloxifene users had 36 percent fewer uterine cancers and 29 percent fewer blood clots, making it a safer choice, government researchers said.

However, data made public on Monday show that the uterine cancer results were not statistically significant. This means the actual number of cases differed so little that they could have happened by chance.

Scientific standards have long held that such results only suggest trends and are not definitive, certainly not to the extent that government scientists portrayed them to be.

Furthermore, so few blood clots occurred in the study that some doctors don't believe that result proves raloxifene is better. Also, it isn't known whether raloxifene's cancer-prevention benefit will last years after women stop taking the pills, as tamoxifen's is known to do.

"There is some genuine controversy here," said Dr. Len Lichtenfeld, deputy chief medical officer of the American Cancer Society. Not everyone agrees "that there was a clear winner in this study," he said.

The news generated heated discussions at a meeting of the American Society of Clinical Oncology, where the study's results were to have been reported first, so experts could review them as they were released to the public. Instead, the cancer institute hastily called the press conference and didn't disclose in materials sent to reporters that some key results were not statistically significant.

"It needs to be publicly vetted because it's not clear either way" which drug is better, said Dr. Roy Herbst, a University of Texas M.D. Anderson Cancer Center doctor who helped run the oncology meeting, the world's largest cancer conference.

The cancer institute's prevention chief, Dr. Leslie Ford, defended her characterization of raloxifene as a clear winner and the way the news came out. The press conference was called to give women in the study the first word of its results without them "leaking out" as happened with two earlier high-profile women's health studies, Ford said.

Other doctors questioned the urgency, because these drugs are taken for five years to prevent long-term breast cancer risks — a very different situation from a drug to treat a disease that could have an immediate life-or-death impact.

Tamoxifen has been used for decades to treat and prevent breast cancer. It blunts estrogen, which fuels the growth of most tumors that occur after menopause, but also acts like estrogen elsewhere in the body and has previously been shown to raise the risk of blood clots and uterine cancer. Raloxifene, sold as Evista by Eli Lilly & Co. for osteoporosis, is believed less likely to cause these problems.

The study tested the drugs in nearly 20,000 postmenopausal women at high risk of breast cancer because of gene mutations, family history or other reasons. (Raloxifene's safety in premenopausal women is unknown.)

Results were released Monday at the cancer meeting and will be in the June 21 issue of the Journal of the American Medical Association.

The results:

--Invasive breast cancers: 168 among the 9,745 raloxifene users; 163 among the 9,726 who took tamoxifen.

--Non-invasive breast cancers: 57 in the tamoxifen group; 80 in raloxifene users.

--Uterine cancers: 36 among the 4,732 on tamoxifen; 23 in the 4,712 on raloxifene. (Half of the women in the study had had a hysterectomy and therefore were not at risk of uterine cancer).

--Less serious forms of breast cancer developed in 57 women on tamoxifen and 80 on raloxifene.

All of these results were a draw, statistically, meaning neither drug was better than the other.

Tamoxifen users probably would have had more uterine cancers as time went on, because they had more cases of abnormal uterine growths requiring hysterectomies, said Joseph Costantino of the University of Pittsburgh, the study's statistician.

Blood clots in the legs or lungs developed in 141 of the 9,726 on tamoxifen and in 100 of the 9,745 on raloxifene — a result that did meet statistical muster.

The difference of 41 clots among nearly 20,000 women is extremely small, and the real message is that both drugs are relatively safe, the cancer society's Lichtenfeld said.

"It's not necessarily going to change my practice" of prescribing tamoxifen, said Dr. George Somlo, co-director of the breast program at City of Hope Cancer Center in Duarte, Calif.

"The winners are the women because they now have choices" of two good drugs for preventing breast cancer, said the cancer institute's Ford.

Price probably won't drive decisions, unless Lilly raises the cost of Evista should it win Food and Drug Administration approval for breast cancer prevention. Generic tamoxifen costs $100 a month and Evista, as sold now for osteoporosis, $75.

In a separate analysis reported on Monday, Dr. Patricia Ganz of UCLA's Jonsson Cancer Center, compared quality of life among 973 tamoxifen and 1,101 raloxifene users in the study and found no difference in how they rated their overall physical and mental health.

However, weight gain, pain during sex, joint pain and vaginal dryness were more common in raloxifene users. Hot flashes, vaginal bleeding and discharge, bladder problems and leg cramps were worse for women on tamoxifen.

"There is a subtle difference between tamoxifen and raloxifene, with the tamoxifen women maintaining their sexual activity a little bit better ...," Ganz said.

Problems like hot flashes have kept many women from taking tamoxifen despite its known benefits, said the "father" of the drug, researcher V. Craig Jordan of the Fox Chase Cancer Center in Philadelphia.