GAITHERSBURG, Md. -- An FDA advisory panel met Friday to discuss how to analyze trials designed to test devices used to treat depression, but didn't offer the FDA any clear-cut guidance.
The Neurological Devices Panel -- an outside group of neurologists, psychiatrists, and statisticians -- was tasked with advising the FDA on what guidelines researchers should have to adhere to when conducting trials on the safety and efficacy of devices to treat depression.
That includes devices used in electroconvulsive therapy, vagus nerve stimulation, repetitive transcranial magnetic stimulation, and cranial electrotherapy stimulation.
Although the agency has clear expectations when it comes to drug trials, it lacks such a requirement for approving devices to treat depression. For instance, the FDA requires a company to submit two positive and well-controlled trials for anti-depressants, but has no similar requirement for devices.
Major depressive disorder effects more than 14 million people in the U.S. each year, and more than 16% of all Americans will suffer from depression at some point in their lives, according to the FDA.
Between 15% and 30% of depressed people do not respond to anti-depressants such as selective serotonin reuptake inhibitors (SSRIs), serotonin and norephinephine reuptake inhibitors, and monoamine oxidase inhibitors (MAOIs), according to an FDA staff presentation.
Those patients may be candidates for device trials, but the panel agreed that it would like to see scientific proof that the patients didn't respond to medication before a potentially dangerous device was implanted in their bodies.
The panel agreed that it's not good enough to simply ask a patient whether they failed to respond to medication -- rather, a lead-in study to a device study should take place in which patients who did respond to the study drug would "exit" the study and not be eligible for device implantation. Those who failed to respond to medication would receive a device.
Another issue the panel discussed was the percentage of patients that would need to respond to treatment in order for the trial to be considered a success. The panel didn't agree on a figure, but one panelist said even if fewer than half of patients saw an improvement in their depression symptoms, the given device could be considered effective.
"This is a very sick group and a 40% response rate could be very clinically significant," said William McDonald, MD, chief of geriatric psychiatry at Emory University in Atlanta.
Among the other questions the panel grappled with: How many different drugs does a depressed patient need to try, unsuccessfully, before being allowed to enroll in a clinical trial testing a device to treat his or her depression? Again, the panel had no clear advice.
Devices to treat depression carry side effects that can be far greater than those that come with medications, including intracranial hemorrhage, stroke, paralysis, infection, surgical complications, and suicide.
The panel spent much of the day considering how many suicides would be acceptable in a clinical trial for a depression-treating device. The FDA was hoping the panel would provide a concise number such as 5% or 7%. But the panelists were uneasy about placing a figure on "acceptable" suicides.
The panel said each and every suicide in a clinical trial should be examined to determine its cause, but specific trials shouldn't be shut down if a certain number of patients commit suicide.
"No, I would not necessarily stop a trial if there was one suicide, particularly in a population prone to suicide," said panelist Sarah Lisandby, MD, chair of the Duke University Department of Psychiatry.
The FDA will now mull over the panel's advice. While the agency isn't actively drafting trial guidelines for depression devices, the information might eventually help device makers who are performing depression trials have a clearer idea of what the FDA is looking for.