ARTICLE IN BRIEF
Experts welcome a new Institute of Medicine report that calls on the FDA to revise its fast-track approval process for medical devices.
Various nerve and brain stimulation devices for treating epilepsy, depression, chronic pain, nerve and muscle disorders, and brain monitoring have been approved by the FDA. But a new report from the Institute of Medicine questions the basis for the approval process and suggests that an entirely new system is needed.
When Congress approved the 510(k) exemption process for medical devices in 1975, manufacturers for the first time could use results of safety and efficacy analysis of similar devices to seek approval for medical devices by meeting a “substantial equivalence” benchmark.
This approach has failed, said the IOM report authors, largely because many of the devices used as the basis for comparison were never thoroughly reviewed for safety or effectiveness in the first place.
The IOM report authors urged the FDA to begin collecting the necessary information to create a new regulatory framework to replace the 510(k) process.
The IOM's conclusions are “right on target,” said Gary Franklin, MD, an adjunct research professor of neurology at the University of Washington in Seattle, who conducts research on outcomes of medical therapies for occupational musculoskeletal disease and predictors of disability in workers' compensation.
“There is no legal basis for allowing these devices to be approved without the same safety and efficacy testing required of all others, especially moderate-risk devices,” he told Neurology Today in a telephone interview, noting that more than 95 percent of medical devices currently on the market have been 510(k) approvals, including moderate-risk and implanted devices.
“I am glad that the IOM report calls for looking at the entire lifespan of a device,” he said. “In a way, pre-market approvals are the most troubling to me. The standards are not anything close to what we require of new drugs.”
In addition to nerve and brain stimulation devices, 510(k) approvals have been granted for hydrocephalic shunts, neurothrombectomy, and artificial embolization devices.
A number of adverse events have been reported in the medical literature by researchers studying patients with some of these devices.
In a 2009 article published in the open access journal the Public Library of Science, the advocacy group Public Citizen pointed to fast track approval of surgically implanted vagus nerve stimulation devices in the brain for treatment-resistant depression.
According to the paper, the only randomized, controlled trial of one device did not demonstrate a statistically significant benefit. Moreover, the trial was not randomized or blinded, it relied on less-rigorous follow-up data, patients in the treated and control groups were recruited at different times, and both groups were permitted to modify their antidepressant medications and/or receive electroconvulsive therapy (ECT).
Implanted brain and nerve stimulation devices have also had sporadic problems. For example, deep brain stimulation devices for treating dystonia and other conditions have resulted in post-operative hemorrhage, infection, paresis/asthenia, and hemiplegia or hemiparesis. Electrode terminal lead movement has also been reported in the months following placement. Far more serious problems have been reported in patients with implanted devices when they undergo MRI, including coma and permanent neurological impairment.
Patients with external spinal stimulation devices have also experienced instances of over-stimulation and shock when they were exposed to electromagnetic impulses emitted by security systems, metal detectors, and antitheft devices. Nine such adverse events have been reported to the FDA, which has advised patients to turn their devices off in airports and similar areas where they might be exposed to electromagnetic signals.
Perhaps the most serious problem to date was the subject of a 2005 FDA safety alert cautioning patients with implanated neurostimulation devices to avoid diathermy therapy — heating tissue electromagnetically or ultrasonically for various therapeutic purposes, including muscle relaxation and pain relief. The agency warned that diathermy can cause implanted electrodes to become very hot, causing tissue damage that can lead to permanent injury or even death. Two patients died in 2001 after receiving diathermy.
Some hydrocephalic shunts granted 510(k) status have also been associated with problems that only surfaced after the fact, including shunt obstructions, mechanical failure, the need to lengthen or replace the catheter, and, of greater concern, over-draining, in which CSF flows more quickly from ventricles than it can be produced, causing them to collapse or resulting in blood vessel damage. In rare instances, this has resulted in subdural hematoma, a type of life threatening brain hemorrhage.
BETTER TESTING, MONITORING
“The entire system needs to be re-examined,” Dr. Franklin said. “Right now, post-marketing safety and efficacy review of these products is pretty woeful. There is no systematic method for collection of either population-based outcome or adverse event data.”
There are currently two methods for reporting adverse events, he continued. The first is the FDA MAUDE database, which is accessible by the public. “But MAUDE is a passive system and relies on physicians or patients to send in a report, and it includes few details on the circumstances of, say, a reported death related to use of an implant.”
The second method is serious adverse event data that the FDA requires manufacturers to report, but Dr. Franklin noted that these data are “not summarized in any systematic way, nor is it publicly accessible.”
Dr. Franklin said some examples of post-marketing surveillance that would be very useful include requiring details on all deaths and hospitalizations related to life-threatening infections from any type of implant, as well as rates of explantation of all implantable devices, including the timeframe and reason for taking them out.
Richard A. Deyo, MD, MPH, the Kaiser Permanente Professor of Evidence-Based Family Medicine and director of the community and practice-based research program at Oregon Health and Science University in Portland, said he hopes the FDA will move quickly to develop a new system to replace the current equivalency approval process, but with so many interests involved, he doubts change will come quickly.
“I think it may happen gradually, and may even involve the Congress,” said Dr. Deyo.
He told Neurology Today in a telephone interview that for spinal fusion procedures — one of his areas of expertise — the fast-track approval process has been especially troubling.
“From my perspective, the current process is outdated. What I find most striking is that permanent devices that are implanted in the body are subject to less stringent approval standards than drugs, when drugs can simply be halted if there is a problem,” he noted, “It seems strange to say that most devices can be approved based, for example, on an ‘equivalent’ device approved back in 1976.”
Dr. Deyo added that insurers, including Medicare, have sometimes taken a tougher stance with regard to certain devices than the FDA. Artificial discs approved as an alternative to spinal fusion devices are one example of this, he said, even though the discs were approved by the more rigorous Pre Market Approval (PMA) mechanism.
“I find it interesting that some insurers are more skeptical than the FDA about device effectiveness and they are playing a playing a role in what I view as a very awkward situation. Patients and their doctors want the newest devices, but carriers are more cautious about providing coverage,” he said.
“Medicare looked at the artificial discs and decided that there was too little evidence to cover them in patients over the age of 65, where there have been no trials. Many other insurers then followed suit. That was the beginning of the push-back.”
He also noted current difficulties in postmarket surveillance, because Medicare data often does not show the specific type of device that was used in a procedure.
“Other insurer claims data are similarly problematic. It is challenging work collecting data on specific [510(k)] devices, but there should be some feasible solutions.”
A BALANCED APPROACH
S. Claiborne Johnston, MD, PhD, director of the Clinical and Translational Science Institute and associate vice chancellor of research at the University of California, San Francisco, also agreed that changes are necessary.
“The [current 510(k)] system is flawed because it creates two tiers of evidence, although it has had a positive influence in helping the medical device field evolve,” he told Neurology Today in a telephone interview.
“If you set the bar too high,” patients who might benefit from devices will suffer, Dr. Johnston noted, for example requiring the same level of evidence required of new drugs. “‘First do no harm,’ does not mean we have to be positive that there are no risks. If that were the case, we would never do any surgery,” he commented.
Although randomized trials should be required of certain potentially harmful devices, he said, not all 510(k) devices fall into that category.
“I hope we can take a careful and balanced approach to changing the 510(k) program instead of demonizing industry; maybe [we need] a system that rewards evidence-based devices.”
Whatever changes eventually get made, Dr. Johnston said not to expect them soon. “This is not going to change overnight.”