Donald Baker receives magnetic stimulation of addiction brain circuits while viewing images of cocaine.

Brian Schneider

Brain-altering magnetic pulses could zap cocaine addiction

Luca Rossi tried to hang himself in a bedroom in Perugia, Italy, in 2012. Suspended by his belt from a wardrobe, he had begun to choke when his fiancée unexpectedly walked in. He struggled to safety, defeated even in this intended last act.

The 35-year-old physician had everything to live for: a medical career, plans for a family, and supportive parents. But Rossi* was addicted to crack cocaine. He had begun his habit not long after medical school, confidently assuming that he could control the drug. Now, it owned him. Once ebullient and passionate, he no longer smiled or cried. He knew he might be endangering his patients, but even that didn’t matter. He was indifferent to all except obtaining his next fix. “It pushes you to suicide because it fills you with your own emptiness,” he says. In the first months after his near suicide, Rossi didn’t drop his $3500-a-month habit. Early in 2013, he learned that his fiancée was pregnant. Frightened by impending fatherhood, he smoked even more. He didn’t—couldn’t—stop.

Then, in April 2013, Rossi’s father, a chemist, happened upon a local newspaper article describing work just published in Nature. Neuroscientists led by Antonello Bonci and Billy Chen at the National Institute on Drug Abuse (NIDA) in Baltimore, Maryland, had studied rats trained to seek cocaine compulsively—animals so powerfully addicted that they tolerated repeated electric shocks to their feet to get their fixes. The rats had also been genetically engineered so that their neurons could be controlled with light. When the researchers stimulated the animals’ brains in an area that regulates impulse control, the rats essentially kicked their habit. “They would almost instantaneously stop searching for cocaine,” Bonci says.

He and his Nature co-authors suggested that targeted stimulation of the analogous region in the human brain—an area in the prefrontal cortex, which sits behind the forehead—could help compulsive cocaine users. In that Italian newspaper article, Bonci explained that transcranial magnetic stimulation (TMS), a noninvasive method of triggering neural activity, might do the job.

Rossi’s father began a frenetic quest for help that led him to Luigi Gallimberti, a prominent Italian addiction physician who runs the private Villa Maria Clinic in Padua. Father and son presented themselves there, and the father handed Gallimberti the article. “My son is a cocaine addict,” he said. “Can you help him?”

Cortex1 Dorsolateral prefrontal cortex2 Ventromedial prefrontal cortex MidbrainCaudate nucleusNucleus accumbensVentral tegmental area Magnetic field Coil Positioning frame Seconds 0 4 8 12 16 18 24 ... 200 50 pulses “Cold” (executive control) circuit In one form of transcranial magnetic stimulation, pulses are delivered many times per second, on and off, for a few minutes. This “intermittent theta burst” stimulation of the dorsolateral pre-frontal cortex may propagate to the midbrain (arrows, left) and strengthen the “cold” (right, dark pink) circuit that overrides drug-seeking impulses. “Hot” (craving and reward) circuitContinuous theta burst stimulation applied to the ventromedial prefrontal cortex is thought to inhibit the neurons of the “hot” (light pink) circuit that connects to the midbrain’s nucleus accumbens and ventral tegmental area. It is abnormally active when people addicted to cocaine are exposed to cues such as white powder. Magnetic medicineElectric pulses in a coil held near the scalp induce a changing magnetic field that creates electric currents in the cortex. Changing the frequency and pattern of magnetic pulses delivered to the cortex can either increase or decrease neuronal firing. Multiple stimulation strategies are being used to battle cocaine addiction.
V. Altounian/Science

In the years since Rossi visited that Italian clinic, TMS has attracted interest among a small cadre of researchers and physicians who are using it to target cocaine addiction, even as skeptics raise their eyebrows. After an initial study of 32 people hooked on the drug yielded encouraging results, Gallimberti began offering the treatment in his clinic; he and his colleagues have now treated more than 300 addicted people.

More rigorous tests of TMS’s ability to blunt the drug’s grip are now underway. Last year, a group led by neurobiologist Colleen Hanlon at the Medical University of South Carolina (MUSC) in Charleston launched the first randomized, double-blind trial of the method as a therapy for cocaine addiction. In May of this year, scientists at the National Institute of Psychiatry in Mexico City launched another. And in Baltimore, NIDA investigators are completing a pilot study in cocaine users to prepare for launching a large, controlled trial next year.

TMS isn’t new to medicine; it is already an approved therapy for depression. But in applying the treatment to drug addiction, investigators are moving into new territory. Although using TMS in depression has provided leads, no one knows for certain how best to apply it to cocaine-addicted brains—and the answers may vary among patients. “There is so much that we don’t know about TMS that I won’t be surprised if the current therapeutic trials don’t work,” says Michael Fox, a physician and brain network imaging expert at Harvard Medical School in Boston, who uses TMS to treat depressed patients. Fox nonetheless believes that “there’s a huge amount of promise” in deploying TMS to fight addiction.

Hanlon and others agree. “Basic science research in drug abuse in the last 3 to 5 years has told us, ‘Look, there is a potential for a treatment here,’” she says. And for the roughly 1 million addicted to cocaine in the United States and for another 13 million users worldwide, the need for a therapy is acute. Of the major addictions—nicotine, alcohol, and, most prominently of late, heroin and synthetic opioids—cocaine alone lacks a treatment approved by the U.S. Food and Drug Administration. It is a wicked habit to shake without help: The 1-year relapse rate in people trying to quit hovers at about 80%.

“Imagine that this does become the first approved neurobiological treatment for cocaine addiction. It’s a big deal, a game-changing therapy,” Bonci says. (In a NIDA-approved use of personal time, Bonci, a neurologist, began consulting on cases at the Padua clinic in January 2016. He and Gallimberti entered into a financial partnership that month and will open a clinic for people addicted to cocaine in Milan.)

Anthony Barker, a medical physicist at the University of Sheffield in the United Kingdom, introduced TMS in 1985 as a lab tool for probing human neurophysiology. By holding an electromagnetic coil near the scalp, above the motor cortex, and running a brief, strong, pulse of current through it, Barker and his colleagues elicited involuntary hand and leg movements. They surmised that the rapid, time-varying magnetic field generated by the coil’s current induced currents in the cortex, the brain’s outer layer, prompting neurons to fire.

At first, researchers used TMS to study how the motor cortex controls muscles and later to examine how the visual cortex works in blind people. In the 1990s, researchers began to experiment with using repetitive TMS (rTMS), which delivers sustained, closely spaced pulses, to treat several diseases. They had learned that low-frequency stimulation, at one pulse per second (1 hertz), made neurons less excitable, whereas high-frequency pulses, at 5 to 20 hertz, made the cells more prone to fire. Depressed people, they found, responded to high-frequency rTMS, presumably because it boosted the activity of sluggish neurons. By contrast, low-frequency TMS seemed to tamp down the auditory hallucinations that can plague schizophrenics.

Researchers also tried rTMS in people with obsessive-compulsive disorder, posttraumatic stress disorder, stroke, Parkinson’s disease, epilepsy, tinnitus, and chronic pain. Since 2009, they have been testing it against other addictions, primarily to nicotine and alcohol. More recently, prompted by the opioid epidemic, several groups including Hanlon’s have begun evaluating TMS as a tool to decrease pain among those at high risk for that addiction.

In 2008, the United States approved rTMS for attacking treatment-resistant depression, its only approved clinical use to date. Regulators in Europe, Canada, and elsewhere also have approved it for that purpose. But skeptics say that the supporting data are weak and that companies that make rTMS devices—more than 700 machines are in use in the United States alone—are hyping the benefits. “I have colleagues who I respect and who I think are honest who swear by it—who think it works. I don’t,” says Walter Brown, a psychiatrist at Brown University who studies the placebo effect, which he thinks may explain most of rTMS’s success with depression.

He is also skeptical that TMS can battle drug abuse. “I have no doubt that some people with cocaine addiction treated with TMS will get better. But it won’t be the effect of the treatment, in my opinion.”

Hot and cold

One morning this May, in a windowless room on NIDA’s Baltimore campus, Donald Baker, an unemployed 45-year-old with a weathered, unsmiling face and an open-necked shirt that reveals a fresh sunburn, is completing a computer questionnaire. “I am thinking of ways to get cocaine,” the survey prompts. The scaled answers run from 1 to 7, with 7 being “strongly agree.” Baker clicks 4. “If there was cocaine right in front of me it would be hard not to use it.” Another 4. “I would not be able to control how much I used.” That gets a 5.

Baker is the second subject enrolled in a pilot study that postdoctoral student Vaughn Steele is runnning in the lab of Elliot Stein, a neuroscientist at NIDA. The pilot marks the launch of its in-house effort to evaluate rTMS as a treatment for addiction. Steele is aiming for a real-world therapy. “It’s designed for somebody to come in off the street and say, ‘I need some kind of treatment; what can you give me?’ And then we can do TMS that day,” he says.

Baker has just returned from a 2-day, midtrial break. The phase I trial is meant to establish the treatment’s safety and tolerability in cocaine users, not determine whether it dulls the desire for the drug. Indeed, this morning, Baker tells Steele that he used while he was away—which a test of his urine confirms. Still, Baker reports a small victory over his demon. During his break, he earned $150 trimming bushes for his uncle—and today $120 is still in his pocket. “I smoked $30 worth and I asked myself what I was doing. I’ve never done that in 25 years. When I smoke, I can’t stop. But I did on Sunday.”

Baker has lost a marriage, a house, and too many jobs to count during his quarter-century of drug use. He has no car, no regular work, and no contact with his three adult children. Although he understands that this study isn’t designed to show efficacy, he still hopes the treatment will help him get clean—not least because he has a new love whom he wants to marry.

In a room down the hall, Baker settles into a big cushioned armchair. He puts his head in a chin rest. On a tight cap, researchers have marked the cortical structures to target. High on the left side of the skull, not far behind the hairline, is the left dorsolateral prefrontal cortex (DLPFC), a key part of the brain’s “cold” circuit, which overrides impulses and is often called the executive control network. Because the DLPFC is known to connect extensively to deeper brain areas that imaging studies have shown to be underactive in people addicted to cocaine, the trial aims to stimulate the region. The researchers are using a form of rTMS called intermittent theta burst stimulation, in which pulses are delivered in bursts, like machine-gun fire. (Theta refers to the frequency used.) The regimen consists of 3-minute sessions, three times a day, for 10 days.

It’s 3 minutes of pain to get rid of a lifetime of misery.

Donald Baker, trial participant

rTMS is loud—the electromagnetic force behind each pulse generates an audible clicking sound—but has a strong track record of safety. That doesn’t mean it’s pleasant. The pulses cause the scalp muscle under the coil to contract; for about one-third of subjects, that tips into pain. “If someone snaps a rubber band at you, you want it to stop,” Stein says. “Your brain doesn’t feel pain, but your scalp and skull do.” Beyond that immediate discomfort, transient headaches are the only commonly reported side effect, occurring in about half of people. Rarely, subjects have had seizures during treatment.

If the 10 pilot subjects tolerate the treatment well and no safety issues arise—seizures are the primary concern—Steele’s group will launch a double-blind phase II trial with 60 cocaine users. Subjects will be randomly assigned to receive either the actual treatment or a sham therapy. To keep both participants and investigators unaware of the treatment given, the coil looks identical on both sides but one side replicates the sound and feel of rTMS without inducing magnetic fields. (The researchers receive a randomized code telling them which side to use.) As well as gauging whether the treatment helps the cocaine users become and stay clean, the team plans to use functional MRI to probe whether rTMS strengthens the activity of the cold, executive control circuit.

Today, Baker, now halfway through his 10-day treatment, is the center of attention. Betty Jo Salmeron, a NIDA physician administering the stimulation, holds the heavy, figure 8–shaped coil close to his scalp. Her eyes are glued to a display that tracks the coil’s position in relation to a red dot that marks the target, the DLPFC.

“You all set to go?” Steele asks Baker. “Yup.”

“It’s 10:02,” Salmeron says. “Start.”

The room goes silent but for the intermittent bursts of loud clicks. Baker focuses on a screen displaying a train of craving-prompting images of the drug and of wads of cash that would let him buy it. Meanwhile, the coil delivers 600 pulses in 50-hertz bursts—2 seconds on, 10 seconds off—in the course of about 3 minutes. One minute passes; then 2.

Baker’s eyes draw into a tighter and tighter squint. He begins blinking nonstop. His mouth becomes a straight, pursed line. Then it is over. “It’s 3 minutes of pain to try to get rid of a lifetime of misery.” he says.

Defining the parameters for effective rTMS—how many pulses to deliver, at what frequency, for how long, and to which part of the brain—remains a fundamental challenge for researchers seeking to deploy it against cocaine and other addictions. Most efforts target the cold, impulse-controlling circuit of which the DLPFC is a key part. But Hanlon, who trained under NIDA’s Stein before setting up her own lab at MUSC, has chosen a different target: a “hot” brain circuit that transmits rewardand craving-oriented impulses. It is typically active in non–drug users when they see normal rewards such as photos of delicious foods. But in cocaine users, the circuit is pathologically overactive when they are exposed to cues such as the sight of cocaine paraphernalia or the drug itself. Hanlon hopes to put the brakes on it by using continuous, rather than intermittent, theta burst stimulation.

In a study reported in the 1 September issue of Drug and Alcohol Dependence, her group found that among 25 people dependent on cocaine, this form of rTMS significantly decreased neural activity at a key way station in that circuit. The treatment also seemed to dampen activity in functionally related regions that drug cues activate, and that are overactive in cocaine abusers.

Through a sister who has struggled with heroin dependence, Hanlon has experience with addiction. She acknowledges that her study is preliminary; her team delivered the theta burst stimulation in six sessions in a single day, and weeks of treatment may be needed to alter brain circuits. But with NIDA support, in August 2016 she began enrolling cocaine-addicted people in a longer, double-blind trial that will test the treatment’s ability to dampen activity in their hot circuit. “We are optimistic,” she says, that we will “change cocaine use.”

Erasing the memory of cocaine's "pleasure"

Gallimberti is already more than optimistic: He’s convinced. He had treated people with cocaine dependance for 20 years when Rossi and his father appeared in his Padua office in 2013. Gallimberti’s attempts to help users with psychotherapy, medications, and residential treatment had been difficult and frustrating. “Cocaine addiction is a disease of brain circuits,” he says. “But not one of the approaches we were using corrected those circuits.”

Gallimberti tried to help Rossi by prescribing an antidepressant, an antianxiety drug, and psychotherapy. Nothing worked, and Rossi quit attending the Padua clinic. His cocaine use escalated. But within 4 months of learning of the Nature article, Gallimberti had bought a TMS device, trained himself and his staff to use it, and treated two cocaine addicts. As Rossi, on the brink of fatherhood, careened toward self-destruction, the clinic asked whether he would become the third. 

Rossi vividly recalls the broiling hot August day of his first treatment with rTMS. “I remember walking out of the clinic. It was beautiful. I looked at the street, the cobblestones. It was magnificent. I felt as if I had never taken drugs in my life.”

But Rossi’s fiancée, also a medical doctor, viewed rTMS as quackery. She peppered Gallimberti and his staff with questions. Afterward, she made it clear to Rossi that she didn’t believe that first treatment had helped him, and he dropped out of the program. He relapsed again. But 2 days after the birth of his daughter in late September 2013, he returned to Padua. There, he entered a 6-month course of rTMS treatments, staying at the clinic for the first week of daily sessions. He says he has not used cocaine since.

Gallimberti and his colleague, psychologist Alberto Terraneo, soon began enrolling patients in a nonblinded treatment study funded by several Italian organizations and by the U.S. National Institutes of Health. As they did with Rossi, the clinicians treated a group of 16 cocaine users with high-frequency rTMS stimulation designed to strengthen the impulse-overriding cold circuit by stimulating the DLPFC. An equal number of cocaine-using subjects in a control group received medication. During the 29-day study, the treatment group reported significantly less craving and produced significantly more cocaine-free urine tests than the controls, as the researchers, including NIDA’s Bonci, reported last December in European Neuropsychopharmacology. (Members of the control group were later offered rTMS, too; their outcomes also improved significantly, the authors reported.)

Gallimberti and Terraneo did not feel the need to wait for larger, more rigorous trials, such as NIDA’s, that would take years to complete. By the time their study was published, they had treated 220 cocaine users at the clinic in Padua—and had become believers. By 31 July of this year, that number reached 342. (On average, the clinic charges €100 for a single session, but it says people who can’t pay are treated for free.) The results remain encouraging, the Padua team says, although they won’t discuss specifics until the work is published. To Terraneo, rTMS’s effect on cocaine users is simple and clear: “It’s as if they lose the memory of the pleasure.”

Four years after finally having a full course of rTMS treatment, Rossi by all accounts remains clean. Although he and his fiancée broke up in 2015, he says he is in a new, happier relationship. He sees his daughter, nearly 4 years old, on alternating weekends. They take trips to a local park, and he makes her laugh by pretending to eat insects. He is completing a residency in general medicine and preparing to start another in angiology. He says he no longer feels tempted to use cocaine “because I focus on the important things of my life.”

Baker, too, interviewed in early August, said that he remains clean. He was spending his days writing a proposal urging policymakers to do more to help recovering users transition from rehabilitation programs to independent life. Baker was still thinking about cocaine. “But I don’t act on it. I kind of consider the consequences a little bit more than I used to.” Maybe his cold circuit—bolstered by TMS—is talking.

The NIDA team hopes that its ambitious follow-up trial, likely to launch next year, will produce more such promising results. “This TMS story presents the opportunity to directly intervene on circuits. So it’s very exciting,” Stein says. “The question is: Will it play out?”

*Luca Rossi is a pseudonym.