FDA Device Panel Considers Design of Clinical Trials for Treatment of Depression

Emma Hitt, PhD

October 11, 2010

October 11, 2010 — The Neurological Devices Panel convened on Friday to advise the US Food and Drug Administration (FDA) about the design of clinical trials of devices to treat depression, but no firm recommendations were made.

Devices used to treat depression include electroconvulsive therapy, repetitive transcranial magnetic stimulation, and vagus nerve stimulation. These approaches are generally used in patients (about 20%-30%) who do not respond to antidepressant medications. In general, devices are implanted surgically and pose all the risks associated with surgery, including infection, stroke, and paralysis.

During the session, the panel considered several issues with respect to trial design involved in evaluating these devices.

One of the questions was what the panel would consider to be the maximal number of completed and attempted suicides in a general depression population and also in a treatment-resistant population before a trial is stopped.

The panel was unable to agree on a number (such as 5% per year of treatment) that would be acceptable.

"Clearly the panel is in agreement that this is a grave problem, that it needs to be examined very carefully, and that it doesn't lend itself — in the absence of data — to numerical stopping rules today," said panel chair Thomas G. Brott, MD, from the Mayo Clinic in Jacksonville, Florida.

Another question posed to the panel was regarding whether a third group (one not receiving device treatment) is needed to assess long-term safety since in trials of devices, both the active treatment and control groups receive treatment with a device sooner or later.

Dr. Brott pointed out that patients will often object to being placed in a group that is not expected to receive treatment and this is a "real problem" in clinical trial design.

"There are pluses and minuses to a third arm," said Dr. Brott as he summarized the panelists' discussion on the topic. "The value of a third arm would be greatest if there were randomization, but randomization to a third arm could threaten the integrity of the trial because of potential refusals to follow through with the third arm," he noted.

Another issue the panel discussed was the percentage of patients that would need to respond to treatment in order to consider a device effective in a trial. Again, no definitive numerical value was reached, but the panel seemed to agree that a response rate of less than 50% would be acceptable. The standard is 50%, but "even a 40% response rate could be very clinically significant," said William McDonald, MD, from Emory University in Atlanta, Georgia.

According to Michael Hoffman, a biomedical engineer with the FDA who presented information on the regulatory processes regarding neurologic devices, the information from Friday's panel may be used to "help create consistency and transparency in designing clinical trials for devices intended to treat depression."