Thumbs Down From FDA Panel on Plastic Meniscus Replacement

An FDA advisory committee voted 6-2, with one absention, that a novel, synthetic knee meniscus replacement’s benefits do not outweigh its risks, and thus is not suitable for marketing.

Members of the Orthopaedic and Rehabilitation Devices Panel of the FDA’s Medical Devices Advisory Committee said the efficacy data were too weak, and a nearly 50% rate of device failure according to prespecified criteria was too great to ignore.

“[They have] cherry-picked patients and cherry-picked surgeons, and if this is the best they can do…,” said panelist Paul A. Manner, MD, of the University of Washington in Seattle, “in actual practice, it’s going to be substantially worse.”

The product, called NUsurface, is a rubber-like plastic disc intended for patients with persistent knee pain following medial meniscectomy. “Meniscus replacement has the potential to fill the gap between minimally invasive meniscus repair and total knee replacement,” according to the manufacturer, Tennessee-based Active Implants. It’s made of polycarbonate-urethane and “does not require fixation to bone or soft tissues”; it is simply inserted into the medial compartment and is held in place by the joint’s natural shape.

At least initially, most patients who still experience significant pain after meniscectomy are given non-opioid painkillers, steroids, hyaluronic acid injections, and/or physical therapy, and may be urged to lose weight as well. Surgical options include meniscal allograft transplant, resorbable scaffold implants, and unloading osteotomy.

The NUsurface implant has been available in Europe since 2008, and Active Implants has been going back and forth with the FDA for that long. The current application represents Active Implants’ fifth attempt to get the product onto the U.S. market: two 510(k) applications were rejected, and two subsequent applications under the FDA’s “de novo classification” system were also denied. Now the company is pushing a third de novo request that, it hopes, address the agency’s previous objections.

In briefing documents released prior to the meeting, FDA staff had questioned what they saw as high rates of treatment failure: lack of clear improvement in pain and function in more than one-third of patients, and revision surgeries needed for close to 20%. Active Implants had identified a number of factors associated with poor outcomes and is proposing labeling to minimize use of the device in patients at higher risk for failure.

Two of the restrictions would rule out patients with large meniscal extrusions and warn against implantation in those with tibial spine heights less than 11 mm. Much of Active Implants’ presentation to the advisory committee was devoted to explaining these.

With these restrictions in place, rates of surgical failure fell by half, from about 33% to 16%, in the company’s clinical trial data. “Study success” rates (a composite of improved pain and function plus no second surgery) likewise increased from 45% to 51%, while the rate in conventionally managed controls went from 23% to 16%, thus emphasizing the implant’s efficacy.

During the meeting, there was some disagreement about whether the need for second procedures reflected actual failure from the patients’ perspective. In most of these cases, the second procedures involved giving patients another NUsurface implant. Trial participants who spoke during the meeting’s public comment portion included three who underwent such revisions, and all said they had no complaints about it and remained very happy with their outcomes.

But most panel members would have none of it. “Any second surgery should be considered a failure,” said Manner, a sentiment repeated by others when asked about the trials’ specification that further surgeries would be considered failure.

Active Implants’ proposal to limit use of the product to patients lacking meniscal extrusions 5 mm or greater, and to those with tibial spine heights of at least 11 mm, was also sharply critiqued. Not because committee members believed extrusions and short spaces between knee bones weren’t important, but because measuring these would be too difficult in routine practice, given that MRIs are needed.

The panel’s radiologist, Ty Subhawong, MD, of the University of Miami, said that it’s unusual even for MRI specialists to quantify tibial spine heights, let alone orthopedic surgeons. And chairperson Harvey Smith, MD, of the University of Pennsylvania in Philadelphia, argued that MRI has a lot of variables that make such precision difficult.

Notably, Active Implants’ trials had each MRI image evaluated by two raters, and they did not agree in about 20% of those from patients assigned to receive the NUsurface product.

Some members also worried that, once the product reaches the open market, surgeons may fudge or ignore the restrictions to implant it more widely. An Active Implants representative, however, promised to provide training to emphasize the importance of adhering to the stated limitations.

Other criticisms included the company’s statistical methods, including use of “last observation carried forward” to account for missing data. (Attrition among control patients was considerable over the 2-year study period.) One panel member noted that this method is not appropriate for patients with a progressive disease such as osteoarthritis.

Also, some panelists complained that 2 years was not long enough to judge the implant’s performance adequately. Col. Melvin Helgeson, MD, of Walter Reed National Military Medical Center in Washington, suggested that time to undergo total knee replacement (the treatment of last resort) would have been a better measure.

In explaining their votes, most of those voting against the product said they didn’t believe tweaking the proposed indications or otherwise modifying the label instructions would make a difference.

But two panelists ultimately accepted the sponsor’s main argument, repeated by clinicians and patients speaking during the meeting’s public comment portion: that the benefits outweigh the negatives. Subhawong was one of them, saying there seemed to be significant (if “small”) net benefit for the target population.

The other yes vote came from John S. Kirkpatrick, MD, of the Orlando VA Medical Center in Florida, who said the benefit-risk balance “tilted slightly” in the device’s favor. But he also urged that the FDA require a postmarketing study to rigorously track the rate and timing of revision surgeries.

Meanwhile, the Active Implants representative urged the agency “not to move the goal posts” after having agreed on how the company should conduct the trials. He also said the FDA should consider NUsurface’s 15-year experience in Europe that demonstrates the product’s overall safety and effectiveness.

The FDA is not required to follow its advisors’ recommendations, but it usually does.