Thousands of people treated for severe circulation problems in their legs may have died prematurely after getting artery-opening devices implanted in their legs that were coated with a crystalline formulation of the drug paclitaxel.

Researchers first raised the startling possibility late last year after an extensive review of data from 28 clinical trials.

Medtronic, Boston Scientific and other makers of these lucrative devices for peripheral artery disease (PAD) say they don’t see any evidence that their particular devices are causing higher rates of death.

This week, a panel of experts assembled by the U.S. Food and Drug Administration is expected to recommend whether to maintain the status quo, order more studies or pull the products from the market. The FDA typically follows the advice of its expert panels.

A lot of money could be riding on the outcome. Medtronic, for example, had been expected to clear $200 million in sales of its In.Pact Admiral drug-coated balloon this year.

Stock analyst Mike Matson, an industry observer at Needham & Co., said it’s unlikely that the FDA’s expert panel would recommend yanking the devices off the market. But sales this year have been lower than expected and some companies have already warned investors not to expect a rapid recovery.

Amid the uncertainty, hospitals across the country — including Abbott Northwestern in Minneapolis — have cut back on using the paclitaxel-eluting stents for the legs —an intervention that had been rapidly gaining in popularity before the specter of early death was raised. The Mayo Clinic recently imposed a voluntary moratorium on all paclitaxel-eluting devices for the legs.

The patients most likely to have paclitaxel applied above the knee in their legs are those who have a condition called critical limb ischemia, where blood circulation in the affected lower limb is so poor that the patient is at risk of amputation.

Critical limb ischemia in the legs is a severe form of PAD, which is caused by plaque buildup that blocks the flow of the blood.

The Centers for Disease Control and Prevention says more than 8 million Americans have PAD in the legs, with the incidence rising with age. About 2% of PAD patients aged 50 or older go on to develop critical limb ischemia, which has a cardiovascular mortality rate of about 25% one year after diagnosis, and a limb-amputation rate of 25% in that same time period, studies show.

In the past the most durable long-term treatment for critical limb ischemia was an artery bypass, an operation lasting more than three hours in which a blood vessel is harvested from elsewhere in the body and then connected above and below the diseased artery segment to bypass the blockage.

In 2012, Indiana-based Cook Medical introduced to the U.S. market the first minimally invasive stent that seemed to significantly improve “patency,” or the ability of a vessel in the legs to stay open after initial treatment.

In 2014, the FDA approved thin medical balloons that can be inserted into a blocked artery in the leg and inflated to clear the blockage, including Medtronic’s In.Pact Admiral and the Lutonix balloon, which was developed in Minnesota and is now sold by the medical supplier BD.

Each of these devices is coated with the drug paclitaxel, a chemotherapy agent widely used on medical devices for the blood vessels because it can reduce inflammation and help prevent the need for a re-operation. The balloons are one-time-use devices that leave a coating of the drug on the vessel, while stents are metal mesh tubes left permanently in the vessel that slowly release the drug.

A five-year analysis of outcomes from the patients treated with the Cook Medical device, called the Zilver PTX found that those who had paclitaxel-coated stents for PAD had a 40% reduction in the relative risk for needing a new stent or having the target vessel reclose within five years, compared to patients treated without the drug. An analysis of In.Pact Admiral patients found they had half the rate of needing a new procedure after three years than patients treated with a plain balloon.

Hoping to replicate or exceed those results, Boston Scientific rolled out its Eluvia paclitaxel stent for the legs last September.

Eden Prairie-based SurModics is working to enroll patients in a clinical trial of its SurVeil paclitaxel-coated balloon, for which Abbott Laboratories has already agreed to be the global distributor. Philips now sells the Stellarex balloon, which was sold by Covidien to satisfy antitrust regulators ahead of Medtronic’s 2014 acquisition of Covidien.

But a critical December 2018 study in the Journal of the American Heart Association sent shock waves through the industry.

The study analyzed 28 past randomized controlled trials of paclitaxel devices for PAD and found an enhanced risk of death years after the procedure.

Among 4,663 patients, the all-cause death rate was 2% in both paclitaxel and non-paclitaxel groups, but for the 2,316 patients with two years of data, the crude risk of death jumped to 7% for the paclitaxel patients and only 4% for control patients.

The difference was even more stark among the 863 patients who had five years of data, with a 15% risk of death from any cause after getting paclitaxel in the legs, vs. 8% mortality in the control groups. (The data do not apply to paclitaxel on coronary stents.)

By January, the FDA had begun to try to replicate the JAHA paper’s findings. It analyzed or used comparative data on the Zilver and Eluvia stents, and the Lutonix, Stellarex and In.Pact Admiral balloons. Among the 975 people with five years of data, the FDA found a 50% increased risk of death among paclitaxel-treated patients, 20% vs. 13%.

But neither the JAHA study nor the FDA analysis proves paclitaxel caused the premature deaths — they can only show an association.

While theories have been floated, there’s no widely accepted theory about a biological mechanism of action. Doses of the drug vary widely among the devices, among other factors.

The companies that make the devices said they see no signs of the mortality trend in their own patients. All five have submitted their own in-depth analyses of their in-house data, which they consider more reliable than the JAHA figures because they contain important “patient-level” information like actual causes of death, like cardiovascular-related causes, cancers and infections.

Since the JAHA study was published, Medtronic and Cook Medical have both acknowledged that past papers on these devices misreported the number of deaths. Both companies said the additional deaths didn’t change the overall findings, but the corrections didn’t inspire doctors’ confidence.

“I appreciate the transparency in retrospect, but it definitely makes you a little bit concerned,” said Dr. Jason Alexander, a vascular surgeon with Abbott Northwestern Hospital in Minneapolis, “even though the response of industry may be entirely valid, and may even be correct. We don’t know the answer to that.”