A panel of doctors and medical experts told the Food and Drug Administration Thursday that regulators should not halt sales of paclitaxel-containing medical devices for the legs in the U.S., even though there may be an increased long-term risk of death from using them.

Instead, the FDA's Circulatory System Devices Panel gave a nonbinding recommendation that regulators should push for better long-term clinical data than was seen during the two-day review in Maryland.

FDA officials didn't immediately announce Thursday night what they will do next, or when they will do it.

From the data presented, the panel did not find any dose relationship with mortality, nor any pattern of adverse events that suggested a biological mechanism of action. The panel said the products' labels should be updated to reflect potential risks, allowing doctors and patients to decide for themselves what to do.

The controversy centers on a set of five devices approved in the U.S. to treat peripheral artery disease (PAD) above the knee in the legs by mechanically opening blocked blood vessels and then leaving behind varying amounts of the anti-proliferative drug paclitaxel to prevent vessel re-closure.

Such devices account for hundreds of millions of dollars in annual sales for Medtronic, Cook Medical, Philips, BD and Boston Scientific, all of which make paclitaxel-eluting leg stents or paclitaxel-coated balloons for PAD. The Centers for Disease Control and Prevention said about 8.5 million Americans have PAD, including as many as 20% of people older than 60.

The product segment has grown rapidly since the first such device was approved in the U.S. in 2012 because of the devices' positive effects. Patients with extreme leg pain found that the paclitaxel devices not only prevented the need for repeat procedures, but even prevented amputations in extreme cases.

But growth in paclitaxel devices for above-the-knee PAD came to a halt shortly after a study was published in the Journal of the American Heart Association last December.

The article reported that these complex PAD patients seemed to be dying earlier than people who were treated in randomized trials with the same procedures without the drug. But the exact amount of increased risk is still unclear, and FDA panel members seemed divided over what should be included on an updated product label.

The JAHA paper analyzed five-year data in studies from Medtronic and Cook and found a 14.7% death rate among the paclitaxel patients and 8.1% death rate among the control patients. In March the FDA expanded the analysis to include long-term randomized trials from Philips and BD, and found a lower but still elevated risk with paclitaxel.

Device companies and physician-researchers argued that it wasn't fair to pool data from such different devices into single studies, and they noted that much larger analyses of real-world Medicare billing and outcome data concluded that the mortality rate was essentially even in the two groups.

On Wednesday, the FDA panel members agreed that there was a clear "mortality signal" in the randomized controlled trials that could not be ignored. None of the trials individually was large enough to detect the signal, which is why it wasn't clear until the studies were combined, panel members said.

Thursday concluded with a recommendation that further data be gathered and more information given to the public, because there is a chance that the signal will turn out to be statistical "noise." But no one on the panel advocated halting sales of the devices.

"I know there are high risk factors and I think the patient or the consumer has the right to be aware of the risk-benefit that they have," said Louisiana registered nurse and FDA panel consumer representative Patricia Daigle. "But ultimately the choice should be up to the consumer."

Amid a hearing with many complaints about missing long-term clinical trial data, Medtronic's In.Pact Admiral drug coated balloon had the lowest rate of missing data of any of the devices studied at the meeting. The company, run from offices in Fridley, had once been expected to surpass $200 million in sales this year.

On Thursday, Medtronic noted that the device remains a first-line strategy to treat PAD, with three out of four DCB patients remaining "intervention-free" for five years.

"We take our responsibility to patients and physicians very seriously and look forward to further collaboration on a response to the panel's recommended next steps," Medtronic spokeswoman Julia Baron Fuller said via e-mail.

Joe Carlson • 612-673-4779