Get the best of the Magazine’s award-winning stories and features right in your e-mail inbox every Sunday. Sign up here.
IN 2006 A CLINICAL TRIAL came to the attention of Dr. Andrew Cole, a neurologist and the director of the Epilepsy Service at Massachusetts General Hospital. The trial was to establish the effectiveness of a new device for treating epilepsy, the NeuroPace RNS System. The device intrigued him, because of its potential for addressing the most challenging cases of epilepsy. It also happened to be a new approach to fighting disease — one that may be as revolutionary as drugs were in the 20th century.
Epilepsy has multiple causes, among them genetics, head trauma, and brain tumors. Regardless, its signature seizures are the result of malfunctioning electrical connections in some part of the brain. Drugs can help, but they don’t work for about 20 percent of epileptics. And for a subset of those patients, conventional surgery isn’t even an option. Seizures can come from areas of the brain involved with speech, vision, memory, or movement of the hand. For these patients, “surgery carries too many risks,” says Cole.
Which is why MGH joined the trial. The battery-powered device — the size of a wispy-thin domino — could be embedded into the skull and connected with tiny wires to the glitchy part of a patient’s brain. It would function as a sort of cerebral pacemaker, stabilizing the misfiring neural links that were causing the seizures.
The trial results convinced Cole, and since the device was approved by the FDA in 2013, MGH surgeons have used it in 25 patients. While it’s no silver bullet — a few patients reported minimal or no impact — Cole says 60 percent of them have seen the number of seizures cut in half, while a smaller group reported even better outcomes.
The device, made by Silicon Valley-based NeuroPace, is an example of bioelectronic medicine, which uses electricity to treat disease. Previous epilepsy neurostimulators worked like an alarm clock, stimulating the brain on a set schedule. The NeuroPace implant tracks and records the electrical activity in the brain and sends electric impulses only when it senses the onset of a seizure. Even better, it has none of the side effects of a drug-based treatment.
These advantages have the health care industry excited about electricity. “The body is filled with nerves and those nerves are conducting electricity,” says Imran Eba, a partner at Action Potential Venture Capital in Cambridge, a fund started by pharma giant GlaxoSmithKline (GSK) that so far has invested more than $100 million in seven bioelectronic medicine startups. “If we can understand how those electrical signals travel through the nervous system to control, say, how much insulin the pancreas produces, then we can start treating disease in a totally different way.”
HUMANS HAVE USED DRUGS TO TREAT ailments since forever. We used to extract them from plants and shrubs: Providing pain relief with the juice of the opium poppy plant — the source of morphine — dates back to 3000 BC. “Nature’s pharmaceuticals,” such concoctions have been called. Even after pharmacology emerged as a scientific field in the second half of the 19th century, drugs often built on centuries-old wisdom: Aspirin, the blockbuster drug of 1899, is a synthesized version of a treatment for fever that used the extracts of white willow bark.
Drugs became foundational to health care over the course of the 20th century. “If you are sick or broken you either take a drug or get a medical device or both,” says Anthony Arnold, a Boston Scientific veteran and now CEO of SetPoint Medical, a pioneer of the bioelectronic approach to disease. If you don’t believe him, just count how many pills the members of your family swallow every day, or the number of drug ads you see while watching TV. Yet drugs are expensive — Arnold points to Humira, a best-selling drug for a range of inflammatory diseases, that costs about $50,000 a year per patient. And because drugs enter the bloodstream, they affect the entire body and can have unwanted and sometimes dangerous side effects. That’s led Arnold and others to search for new, non-pharmaceutical treatments. Think immunotherapy, which re-engineers T cells, harnessing the power of our immune system to combat leukemia, for example. Or ingestible robotic medicine — the idea of swallowing a tiny bot that could treat you internally — being explored in Daniela Rus’s laboratory at MIT.
Bioelectronic medicine, otherwise known as electroceuticals or neuromodulation, may be the fastest-growing alternative to drugs. It has attracted more than $1 billion in investment in the past three years. Pharmaceutical and biotech heavyweights like Boston Scientific, GSK, and Medtronic (not to mention Google’s parent company Alphabet) have put money both into research labs working on the fundamental science underlying this emerging medical field and startups trying to bring new treatments to market.
We do, of course, already use electricity to treat some conditions: Pacemakers have steadied abnormal heart rhythms for decades and more recently, doctors have used so-called deep brain stimulation probes — like Boston Scientific’s Vercise system — to treat Parkinson’s disease. Both approaches involve major surgery. Bioelectronics true believers envision a time when doctors will inject a grain-sized device into a patient and that device won’t just stimulate or modulate the nervous system — it will turn individual neurons on and off to prevent disease.
This new era of bioelectronic medicine began when Dr. Kevin Tracey, a neurosurgeon at the Feinstein Institute for Medical Research in New York, discovered the “inflammatory reflex” — the neural circuit that regulates the immune system’s inflammatory response. In 2007 Tracey teamed up with Dr. Shaw Warren, an infectious disease specialist at Massachusetts General Hospital Research Institute, to launch SetPoint to commercialize bioelectronic treatments for inflammatory diseases (neither cofounder is directly involved with the company today).
Progress hasn’t been lightning fast — last year, almost a decade after its founding, SetPoint published the results of a clinical trial of its treatment for rheumatoid arthritis. This painful inflammatory joint condition affects more than 1 million Americans, who mask the pain and reduce inflammation with drugs. The SetPoint treatment aims at eliminating the cause of inflammation, excess production of certain proteins. Doing so involves implanting a small device in a patient’s neck onto the vagus nerve, also known as the “superhighway” of the nervous system, because it extends from the brainstem down to the abdomen. SetPoint’s implant sends electrical pulses to the nerve, resetting protein production to normal levels. Of the 17 patients in the clinical trial, 12 showed significant improvements and several saw their symptoms disappear, results promising enough for the company to move forward. A treatment may be on the market in three years, says Arnold. Meanwhile, SetPoint is building on its understanding of how the neural system controls inflammation to develop treatments for Crohn’s disease and multiple sclerosis.
SetPoint is one of many young companies with bioelectronic treatments in the trial stage. NeuroSigma is testing a patch that, when worn on the forehead overnight, stimulates the trigeminal nerve. Primarily responsible for detecting facial sensations like touch, the trigeminal also distributes these signals to nerves in parts of the brain associated with attention and mood. The NeuroSigma patch would offer a non-drug alternative to treat ADHD. Researchers at the Feinstein Institute — which remains a leading center for bioelectronic research — are shooting for the improbable: giving patients paralyzed by spinal cord injuries control over their limbs again.
And bioelectronic treatments for some common ailments are already available. Boston Scientific, among others, offers several implants that treat lower back pain by electrically interfering with pain signals headed up the spinal cord to the brain. To treat central sleep apnea, a breathing disorder that can cause everything from poor sleep to congestive heart failure, Respicardia’s Remede implant prompts the phrenic nerve, which runs from the neck down through the lungs, to contract the diaphragm and keep a person breathing regularly.
Not all electroceuticals involve surgery. There are FDA-approved options on the market for depression, anxiety, and insomnia that use a neurostimulating headband or hand-held device. And for the aggressive brain cancer glioblastoma, patients can wear a cap that creates electric fields to slow or stop the tumor’s growth.
WHILE BIOELECTRONIC MEDICINE presents an opportunity, it also offers significant obstacles. The next generation of bioelectronic medicine is taking shape in labs like those of Dan Freeman, an electrophysiologist at Cambridge’s Draper Labs. He spends his days working on “sub-millimeter, inductively powered neural stimulators” and “microscopic magnetic stimulation of neural tissue.” In other words, the engineering challenges of building a device small enough to be injected into the human body, pass muster with the FDA, and not require surgery every time it needs a new battery. Finding the right materials and power sources is an ongoing question, but Freeman’s team recently developed a stimulator 2½ millimeters long and with wires 0.02 millimeter in diameter — that can be powered wirelessly and deliver sufficient charge to a nerve to activate it.
Then there’s the sheer scale of the nervous system, which contains thousands of different kinds of neurons or cells across a network with trillions of connections. Ed Boyden, professor of neuroscience at the MIT Media Lab, and his collaborators are trying to understand what different cells do and how to control them. They also want to see if we can get the nervous system to control our immune system. “If you could turn the immune system on, could you attack a cancer? If you could turn it off, could you calm down an autoimmune disease?” Boyden says.
To answer these questions, Boyden and his collaborators invented a new field: optogenetics. The idea is to create light-sensitive molecules that can be inserted into neurons and, from there, tracked and controlled. “When you shine a light you can turn those nerve cells on or off,” says Boyden. “Groups in academia and industry are using the tool to discover patterns of neural activity.”
There are other engineering challenges, too, but bioelectronic pioneers also need to bring about a culture shift within health care. Doctors are used to thinking about biochemistry and molecules, says Dr. Ian Cook, a professor of psychiatry at UCLA and the chief medical officer at NeuroSigma, the company developing a bioelectronic treatment for ADHD. “This idea of thinking about magnetic fields and electrical signals is a bit out of the box,” he says, and it will take time for physicians to understand and become comfortable with the new treatments. For the average patient, the idea that the brain has both chemical and electrical elements is new. “People were familiar with electrocardiograms being used to monitor the heart for decades before the first pacemaker was implanted,” says Cook.
And then there’s the question of who will pay for these treatments. An electroceutical implant that treats the same chronic condition as a drug should eliminate the ongoing annual costs of the drug. It’s not yet clear, though, what upfront development costs will be for these devices, which will still have to go through the same studies and trials as drugs. Insurance companies, of course, base decisions on a cost-benefit analysis. If bioelectronic treatments offer a novel or more targeted, less expensive treatment than a drug, insurers will probably pay for it. Most insurers — even Medicare — already cover the deep brain stimulation for Parkinson’s and the NeuroPace implant for epilepsy. Newer treatments will have to prove their effectiveness. It will take years for bioelectronic medicine to reach its full potential.
This story has been updated to correct a reporting error. The diameter of a wire in an experimental stimulator developed by Dan Freeman’s team was misidentified; the correct diameter is 0.02 millimeters.