Fine-Tuning Deep Brain Stimulation

Despite advances in both brain science and technology, DBS devices have not improved much since the original prototypes were implanted 25 years ago, says Philip Starr, M.D., a neurological surgeon at the University of California, San Francisco (UCSF).

“This is a device that works very well for most people, but there are a few problems that could be addressed by more advanced devices,” he says.

Those issues, says Nick Langhals, program director for neural engineering at the National Institute of Neurological Disorders and Stroke (NINDS), include the fact that they are always turned “on,” which can lead to DBS devices becoming less effective over time (through habituation), as well stimulation-induced side effects like bradykinesia, an involuntary slowing of movement, or dyskinesia, an overall impairment of voluntary movement.

“There is potential here, using disease-related biomarkers, to fine-tune the device, not only across an individual stimulating electrode site but also to maybe steer the current in different directions based on biological signatures that are detected from the brain,” he says. “The latest generation medical devices now offer the ability to not only stimulate the brain but also sense activity from other areas. This offers academic investigators the opportunity to explore, gather data, and come up with better tuning for these devices to help patients avoid problems.” Perhaps a device could be set to off, and then when its sensors register the onset of a symptom signal the device turns on until the symptom subsides.

To test the idea of such an adaptive closed-loop stimulation, Starr and colleagues used the newest Medtronic DBS device to both monitor and stimulate brain activity. By placing the stimulating electrode in the basal ganglia, as the older models do, and adding a sensing electrode in the primary motor cortex, the system could alter stimulation levels in real-time depending on the activity it is reading from the brain. By monitoring activity in primary motor cortex, a brain area that shows a distinct pattern of activity during dyskinesia, the device could reduce stimulation to basal ganglia when it read that pattern, reducing the unwanted side effect. Starr and colleagues implanted this device in two patients with Parkinson’s disease, demonstrating that the adaptive device was as effective as traditional devices, and showing that it  self-tuned to help with the disorder all while controlling for dyskinesic side effects. The results were published in the May 9 issue of the Journal of Neural Engineering.

“Previous work showed that you can detect a brain oscillation right in primary motor cortex that correlates strongly with dyskinesias, an unwanted side effect from excessive stimulation,” says Starr. “It’s a strong signal—so we thought it was a good candidate to use for adaptive stimulation. And we found that it did help with the dyskinesia. This is a technical demonstration of the ability of a new type of totally implanted sense-and-stimulation device that can function in an adaptive mode.”

Beyond Parkinson’s disease

Having a device that can both monitor and modulate activity is a boon to patients and researchers, says Langhals. More researchers are using machine learning algorithms on the data recorded from these devices to come up with new approaches to old problems.

“This offers the potential to find new pattern signatures in the brain that can be used as biomarkers—and perhaps used to better fine-tune brain stimulation in the future,” he says. “These new devices let us take advantage of this large amount of data and mine that for these signatures that we might not have known to look for, or even otherwise have seen.”

Starr agrees. To follow this feasibility study, Starr and colleagues plan to test the device in patients with so-called “brittle” fluctuations, who require changing levels of stimulation throughout the day to help manage their symptoms—as well as to find new ways to adapt stimulation to deal with other side effects related to Parkinson’s disease.

“It’s quite hard to program devices using conventional techniques that can help these patients and we hope to show that adaptive stimulation works better with fewer dyskinesic side effects,” he says. “In addition, we are also working on adaptive stimulation algorithms to treat depression and anxiety in Parkison’s patients.”

Helen Mayberg, a neuroscientist at Mount Sinai School of Medicine who has championed DBS to treat depression in patients with major depressive disorder (MDD), says Starr’s results, while preliminary, are very encouraging.

“We are constantly watching where [researchers] are going in the Parkinson’s field because they are showing us what is possible,” she says. “They do have a tremendous advantage over those of us who are trying to treat depression because they have such a strong biomarker—a signal that they can successfully manipulate and track. But there is a lot to learn from this that we may be able to apply to our own efforts.”

Langhals thinks Starr’s feasibility study is a strong first step to devices that can help the neuroscience community develop better therapies for a variety of neurological and neuropsychiatric disorders.

“The whole mission of the BRAIN Initiative is to develop the next generation of tools and technologies to not only better understand how the brain works—but also to develop new strategies to treat these different disorders,” he says. “This study, and many like it, are successfully pushing the envelope, coming up with new ways to help us treat the problems that so many patients face every single day.”