However, the drug Zometa did seem to help certain post-menopausal women. Its maker, Novartis AG, is considering further study, but will suspend plans to expand it beyond its current use as a treatment for patients whose cancer has spread to the bone.
"Ten years of work and to have essentially a negative study is disappointing, particularly on a tremendous wave of enthusiasm for this based on some positive trials in the past," said the study's leader, Dr. Robert Coleman of the University of Sheffield in England. He presented results at a cancer conference in San Antonio.
Bone drugs called bisphosphonates, sold as Fosamax, Boniva and Actonel, have long been sold for treating osteoporosis. Those are daily pills; Zometa, sold as Reclast for osteoporosis, is given as an infusion twice a year.
Hopes that these drugs could also prevent cancer soared after a study two years ago found Zometa cut the risk of cancer recurrence by 30 percent in younger women forced into early menopause by hormone treatments they received.
The excitement grew last year, when a large study found that women who were not cancer patients and were taking daily bisphosphonate pills to prevent bone problems were about one-third less likely to develop breast cancer.
The new study was meant to be definitive. It tested Zometa in 3,360 women of all ages in seven countries who had breast cancer that had spread to lymph nodes. All received standard cancer treatments, and half also got periodic infusions of Zometa for five years.
After five years of followup, about 400 women in each group had died or suffered a recurrence.
However, among the roughly 1,100 women who were at least five years past menopause when the study began, Zometa cut the risk of recurrence by about 27 percent and improved survival odds by about 29 percent.
"We don't believe that's just a chance finding," Coleman said. It fits with the benefit seen in the earlier study of younger women. It may be that Zometa works best when no or little estrogen is present, he said.
Side effects are a concern: 26 women on Zometa - 1 to 2 percent of the group - developed confirmed or suspected cases of jawbone decay, a serious problem long linked to bisphosphonates. Blood clots in the lung also were more common among those on Zometa, although not significantly so, Coleman said.
"That makes the case that these drugs should not be given without a clear indication of their value. They're not totally innocuous," he said.
The study was sponsored by Switzerland-based Novartis, and Coleman consults for the company. Zometa costs more than $1,000 per infusion.
Herve Hoppenot, president of Novartis Oncology, said the company would withdraw applications in the U.S., Europe and elsewhere to expand Zometa's use to premenopausal breast cancer patients like those in the earliler successful study.
"We need to decide what to do from here," he said. The new study's result "was a surprise, I would say."
Zometa's role in cancer prevention remains uncertain, said Dr. Peter Ravdin of the University of Texas Health Science Center at San Antonio, one of the organizers of the cancer conference.
"There are some indications that in some patients it may still have value," but the side effect profile "certainly means that this drug shouldn't be given without confidence that it will cause benefit," he said.
Studies testing other bisphosphonate drugs for breast cancer will have results in a year or two.
Breast cancer is the most common major cancer in women. About 207,000 new cases and nearly 40,000 deaths from it are expected in the United States this year.
The cancer conference is sponsored by the American Association for Cancer Research, Baylor College of Medicine and the UT Health Science Center.