Learning From Our Mistakes

As 2012 comes to a close, it’s worth noting that there were many positive developments during the year. Among the most noteworthy were the emergence of the retinal prosthesis market in Europe and a favorable FDA panel vote for Second Sight’s Argus II implant. The maturation of the hypoglossal nerve stimulation market and the emergence of neurotech products for Alzheimer’s disease are also good signs.

But despite these positive trends, the neurotechnology industry is still prone to failure in clinical trials that are significant for the overall health of the industry. Speaking at the 2012 meeting of the North American Neuromodulation Society, Brain Kopell of Mount Sinai Medical Center in New York offered his assessment of recent failed clinical trials. “We keep doing the same things over and over again,” he lamented, citing as examples failed trials for occipital nerve stimulation for headache, DBS for epilepsy, CVRx’s neurostimulation for hypertension, and Medtronic’s use of VC/VS DBS for depression. Kopell served briefly as chief medical officer for Northstar Neuroscience and was intimately aware of many of the causes of that company’s failed trial for cortical stimulation for stroke rehabilitation, which led to the demise of the well funded firm.

Kopell and other speakers at the meeting expressed regret that there is no published data from many of these failed trials. “We’re never going to learn if we bury the results,” he said. While commercial firms may seem to have little motivation for publishing data from failed trials, the results can be extremely valuable to future investigators and startups, particularly when the trial failure is the result of poor trial design and not ineffective science. Kopell also believes that overreliance on randomized controlled trials creates problems for the neurotechnology industry. “The RCT assumes we know how to randomize in the first place to produce biologically distinct subgroups,” he said. Instead, Kopell thinks that audited registries, which allow analysis of patient characteristics, would serve the industry better.

As a new era of comparative effectiveness looms for the healthcare industry, neurotech vendors will not only have to demonstrate that their devices are safe and effective, but also that they are cost-effective relative to other available therapies. This will place renewed pressure on investigators to select their trial participants very carefully, a point emphasized by pain clinician Joel Saper in another session at NANS [see conference report, p7]. “Even the most troubled of our patients deserve treatment,” he said. “But we have to beat the placebo in these studies. If we’re forced to do neuromodulation only in the end stage, we’re not going to do well.”

Hopefully, neurotech vendors will heed the advice from observers such as Kopell and Saper and avoid making the same mistakes in the future.

James Cavuoto
Editor and Publisher