A new federal study Wednesday dimmed hopes for a significant advance in stroke treatment but encouraged public health experts by demonstrating how to check the seemingly irresistible idea that the newest, most expensive medical tool is always best.

The Gateway-Wingspan system, a device modeled on stents used for heart patients, was approved in 2005 in the hopes of protecting thousands of patients at high risk for suffering a devastating stroke. Patients using it suffered more strokes and deaths than those simply aggressively treated with drugs and advice, the study found.

An independent panel monitoring the experiment halted the research prematurely in April when that pattern became clear. The New England Journal of Medicine published the results early online Wednesday because of the immediate public health implications.

“Without the trial, this procedure seemed destined to become the standard of care,” said H. Gilbert Welch, a professor of medicine at the Dartmouth Institute for Health Policy & Clinical Practice. “With it, we have another example in which the best medical care is not the most medical care.”

“We’re usually trying to play catch-up: The technology is out the door, and we’re trying to rein it back in,” said Walter J. Koroshetz of the National Institute of Neurological Diseases and Stroke. “This is an example of how to do it right.”

How it works

The Gateway-Wingspan gently widens a narrowed artery in the brain with a tiny balloon and then props the passageway open with a miniature expanding mesh scaffolding. As many as 90,000 Americans every year suffer the kind of stroke the treatment was designed to prevent.

“Intuitively, it made a lot of sense,” said Koroshetz, whose institute funded the study. “People thought we could fix these lesions in the brain just like we do with coronary arteries.”

The Food and Drug Administration approved the device based on a small, less rigorous study suggesting it might keep many survivors of strokes caused by narrowed arteries in their brains from suffering another attack that would severely incapacitate or kill them. Blacks and Asians are especially at risk. Regulators wanted to make what appeared to be a life-saving advance available quickly on a “compassionate” basis. Thousands of patients began to undergo the procedure.

But the Centers for Medicare & Medicaid Services (CMS), which administers the health-care system for the elderly, agreed to pay only for the procedures in studies verifying that it really worked, significantly limiting the proliferation of the procedure. Some powerful medical groups pressured CMS to pay for it unconditionally.

“You can see the respective societies and physicians advocating for reimbursement (some of their members do the procedure),” e-mailed Joseph P. Broderick of the University of Cincinnati College of Medicine, who authored an editorial accompanying the study. “CMS made the right call.”

In the meantime, the National Institutes of Health launched the $20 million study to evaluate the procedure on 764 patients ages 30 to 80 who had already suffered a stroke or mini-stroke. All the subjects received an aggressive regimen of drugs to lower blood pressure, cholesterol and blood clotting, along with detailed advice about how to exercise, eat better, quit smoking and control diabetes. That cost about $3,400 for the first year, said Marc I. Chimowitz, a professor of neuroscience at the Medical University of South Carolina, who led the eagerly awaited study. Half also received the Gateway-Wingspan, which added at least about another $20,000 for each patient, he said.

The study was halted after only 451 patients volunteered at 50 sites around the country when the surprising trend became clear. An interim analysis showed that 14.7 percent of those getting the Gateway-Wingspan suffered a stroke or died within 30 days, compared with only 5.8 percent of those getting drug-and-advice care. Within a year, 20.5 percent of the Gateway-Wingspan group had a stroke or died, compared with 11.5 percent of those who did not undergo the procedure.

“The stroke rate in the stenting arm was unexpectedly high, which was disappointing,” Chimowitz said. He noted that the procedure was becoming increasingly common at major medical centers in the United States, as well as in other countries. “These findings have worldwide implications.”

The researchers are unclear how to explain the results. But the procedure could disturb fragile plaque in arteries, breaking some loose and inadvertently causing a stroke. The death rate in the intensive medicine-and-advice group was also unexpectedly low.

“There’s the intuitive sense that if you make the artery look better, the patients will do better. This study shows this is not necessarily true. That’s why this is important,” Koroshetz said.

Seeing possible benefits

Some experts, however, remained confident the procedure might benefit certain patients, such as those who are sicker or have larger blockages.

“Initial studies with coronary artery stenting and carotid artery stenting also were not greatly successful,” said Michael J. Alexander of the Cedars-Sinai Medical Center in Los Angeles.

Stryker, which owns the system, said it “continues to support” the device for patients “who have failed medical therapy.”

“Stryker is committed to partnering with the NIH, FDA and study leaders to better understand and interpret the trial data results,” the company said in a statement.

The risks and benefits of stroke prevention procedures have long been controversial. A similar study is evaluating devices that suction out or pull blood clots from blocked arteries in the brain of acute stroke patients. Those techniques have proliferated before being fully evaluated.

“Without this trial, that’s probably what would have happened in this case — it would have spread all over,” Koroshetz said.