Sixteen women given Zometa died versus 26 of the others — a difference that could have occurred by chance alone but an encouraging trend that doctors hope will mean better survival as the groups are followed for a longer time.

There were no big differences in serious side effects, though minor ones like fever and bone and joint pain were more common among women given Zometa. Two percent of all study participants developed a rapid heartbeat, but only three were hospitalized — two on Zometa and one of the others.

The study was sponsored by Zometa's maker, Swiss-based Novartis, and British-based AstraZeneca PLC, which makes Arimidex, the brand name of anastrozole. Gnant consults for the companies and several other breast cancer drugmakers.

With doctor fees for the infusion, a Zometa treatment can run more than $1,200. The other large study is testing it in 3,360 pre- and postmenopausal women with cancer that has spread but not extensively.

Experts stressed that the results so far are only in women who were made menopausal by hormone-blocking treatments — not women who develop breast cancer after natural menopause.

For now, using Zometa to prevent breast cancer recurrence should be confined to those who develop breast cancer before menopause, said Dr. Eric Winer of Dana-Farber Cancer Center in Boston.

"This is a treatment that doctors should talk to a patient about" because of these encouraging new results, Winer said.

In other news at the conference, women with advanced breast cancers who were given Avastin plus Taxotere were a little less likely to have their cancers progress than women given Taxotere alone. However, side effects including high blood pressure were more common for those taking both drugs. Taxotere treatment is more common in Europe and Asia; in the United States, doctors are more likely to use Taxol.

In the study of 736 women, 44 percent of those given just Taxotere had their tumors shrink versus 55 percent of those also given a lower dose of Avastin and 63 percent of those given a higher dose.

Avastin, marketed by California-based Genentech and Swiss-based Roche Holding AG, recently won federal approval for breast cancer — against the recommendations of outside advisers. The approval was based on measurements like those in this study — cancer progression, rather than overall survival. The new study was too short to show any differences in survival.