Profile: Helen Mayberg

Turning Off Depression

by David Dobbs

from Scientific American Mind, August/September 2006 
__________________________________________________________________

Given her background, curiosity, and energy, Helen Mayberg seems destined from girlhood to do what she's doing now, despite that what she's doing now was then inconceivable. Her father, having flirted with studying neurosurgery, practiced family medicine in Los Angeles County. Her uncle, enamored of machines that see through bodies and theories that jump across disciplines, used X-rays and other nuclear medicine tools to research biochemistry. Now Mayberg, a professor of psychiatry and neurology at Emory University, peers into brains to see mood networks — and, lately, in one of the most startling experimental depression treatments in decades, rework them with electrodes. By combining her father’s bedside dedication with her uncle’s technical curiosity, she is changing neuroscience.

Like most researchers, Helen Mayberg started her career hoping to transform her discipline. She naturally expected to do so in the usual way – by slowly accruing results that eventually amount to something significant. She has done this. But last year she also created a big peak all at once, as it were, when she and a couple collaborators described how they cured 8 of 12 spectacularly depressed patients — people virtually catatonic with depression despite years of talk therapy, drugs, and even shock therapy — by inserting pacemaker-like electrodes into an area of deep cortex known as Area 25.

Starting in the early 1990s, Mayberg had identified Area 25 as a key conduit of neural traffic between the “thinking” frontal cortex and the older limbic areas that give rise to emotion. She found that Area 25 runs hot in depressed and sad people — “like a gate left open,” as she put it, allowing negative, depressive emotions to overwhelm brain and mood. Inserting the electrodes into Area 25 closed this gate and alleviated the depression of most of the trial’s patients. The study won her instant renown for bringing cutting-edge neuroscience to the threshold of therapeutic practicality.

"Mayberg is beginning to do for depression what we did 25 years ago for cancer,” says Thomas Insel, director of the National Institute of Mental Health. “She’s showing us mechanisms we can manipulate so that we can approach it in new ways. It’s early yet. But we can safely say that Mayberg’s work shows us whole new avenues into understanding and treating depression."

There’s a certain irony here. Mayberg thinks she is probably the only board-certified neurologist whose main title is professor of psychiatry. This is “sort of strange,” she notes, for she originally rejected psychiatry as too nebulous. “I didn’t like the toolkit.” Which is how it came that a neurologist (“someone who’s all about the wiring diagram,” as she puts it) created one of the most significant findings in years about psychiatry’s most common and elusive problem.

“It’s funny sometimes,” she says, “how things work out.”

Eat dinner with Helen Mayberg, as I happily did, and you are treated not just to a good meal (for she appreciates good food as much as good ideas) but an infectious intellectual excitement. Lively of manner, with big eyes and a ready smile, Mayberg has a knack for stretching a meal while making the time pass quickly. At 50 she combines the enthusiasm of a freshly inspired grad student with the literate veteran’s appreciation of history.

“I was always a tinkerer,” she told me over dinner one evening. “Summers I used to spend hours in my uncle’s lab at Berkeley. He did early work mapping out thyroxin dynamics in the brain. We’d talk mapping, which I’ve always found fascinating, and he’d give me little lab tasks to do. I loved the lab — the logic of it, the gadgets and Geiger counters. Measuring things to solve puzzles.

She entered UCLA medical school figuring she’d be a psychiatrist. Yet when she did her psychiatry rotations in medical school in the late 1970s, she found few gadgets and little quantitative measurement. “There were no CT scans available then,” she recalls, “much less PET imaging or fMRIs. And most psychiatrists didn’t fully accept the biology underlying psychiatric disorders.” The profession viewed schizophrenia, for instance — which now has well-recognized genetic and neural underpinnings —as a reaction to maternal neglect or abuse.

Then she did a senior-year clerkship with neurologist Norman Geschwind at Harvard’s Beth Israel Hospital in 1980. Geschwind had spent over four decades pushing the notion that the brain works not as a single unit but as a functionally organized system coordinating operations rising from different areas. Dysfunction rose from breakdowns in the coordination of neural between areas.

Geschwind’s vision, buttressed by his research and brilliant readings of earlier cases from neurological literature, led the move from the monolithic view of the brain (i.e., that it functions as a whole, with little regional specialization), which dominated the early 20th century, to today’s network models, which emphasize coordinated action of semi-specialized regions. When Mayberg studied with Geschwind in 1980, this network view was being confirmed by an explosion of discovery about how hormones and neurotransmitters carried messages to and between various brain areas. Mayberg, watching Geschwind apply these models on Beth Israel’s neurology wards, found here a far more appealing model of mental function than psychiatry offered.

After graduating she took up a neurology residency at Columbia in New York City, where she investigated depression in stroke patients. She hoped to localize the neural networks involved. But the stroke patients’ lesions varied so much in location and severity that she couldn’t find consistent patterns.

Still, the project honed her interest, and when she finished residency and moved to a post-doctoral program at Johns Hopkins, she began studying depression in Parkinson’s patients. Parkinson’s offered more promise, for by definition it rises from damage to a movement-crucial deep-brain area called the globus pallidus.

Hopkins led the world in neurotransmitter research at the time, breaking ground almost monthly on dopamine and serotonin function, so Mayberg naturally started by trying to find anomalies in the patients’ neurochemistry. But focusing on chemicals suited her little better than psychiatry did.

“With psychiatry,” she explains, “the resolution was the whole brain. That was too low-resolution for me. I discovered that the chemistry” —that is, neurotransmitter action at the cellular level — “was too fine a resolution. I wanted to see how the parts worked together.

So Mayberg, applying her uncle’s old discipline of nuclear medicine, developed a new project. She and some collaborators took 60 Parkinson’s patients, some depressed and some not, and used PET scans to look for differences in activity in the frontal and paralimbic regions – that is, the “thinking” frontal cortex around the forehead and the older, more interior cortex areas surrounding the limbic centers for emotion, memory, and learning. They found that the depressed patients showed far less activity in both these areas of cortex. Over the next few years, in the early and mid-1990s, she did similar studies comparing depressed and non-depressed Huntington’s, epilepsy, stroke, and Alzheimer’s patients. The depressed patients in every study found this same reduced frontal and paralimbic activity.

She also found something else: In addition to depressed frontal areas, depressed people had one particular older area of cortex, a region called Area 25 just over the roof of the mouth, that was especially busy. Another researcher working separately — Wayne Drevets, then of Washington University and now at the National Institute of Mental Health — also noticed this hyperactivity. Area 25 proved to have strong connections to both the limbic region’s emotional and memory centers and to the thinking cortex. Exactly how Area 25 normally modulated traffic between these areas wasn’t clear. Yet in the depressed it was more active. Perhaps it was working overtime as it tried to temper a depressive loop set up between emotional and thinking centers. Or perhaps Area 25 itself caused the problem by kicking into overdrive and letting depressive loops take over. In any case, says Mayberg, “we were seeing this Area 25 thing is important.” In the combination of its excitement and the reduced frontal activity lay a pattern suggesting something fundamental about depression.

In 1997, she wrote a long, theoretical review paper describing the findings supporting this pattern. In the way of these things, few took notice.

“Quite frankly,” she told me, “no one was particularly interested. I was asking them to look at a lot of brain regions and think of depression in a new way. People weren’t ready for it. So I got put in a box.” Most of her studies to that point had been on people suffering some other neurological problem, such as Parkinson’s, epilepsy, or stroke, and the view then of depression in such patients was that the depression — branded “secondary depression” rather than ordinary “primary depression” — was an inevitable and essentially unimportant side-effect of the main condition.

“So they’d say, ‘Oh, you do that neurological depression stuff,’” Mayberg recalls. “‘Very nice.’ And I’m saying, “No, no, no! This is about all depression.’ But it just seemed to annoy people.”

Annoyance changed to attention at the century’s turn, however, as she tested this notion with increasingly revealing studies. She asked healthy subjects to think sad things, then scanned them when the tears were flowing. They showed depressed frontal activity and a hyperactive Area 25, but as the sadness passed, the frontal area revived and Area 25 calmed. She scanned depressed patients undergoing Prozac and placebo-drug treatments. In both groups, those that recovered showed a rise in frontal activity and a calming in Area 25. It seemed that no matter what the cause, depression depressed frontal activity and either caused or rose from hyperactivity in Area 25 – and that curing the depression reversed these effects

Then, in early 2004, she published a study that drew wide notice and — and threw her for a loop. She scanned two groups of depressed patients undergoing treatment — one with Prozac, one with cognitive behavioral therapy, or CBT. The Prozac patients showed the same pattern as other studies had found – depressed frontal activity that increased in those that got better, and a hyperactive Area 25 that calmed. The CBT patients, however, displayed a new and confounding dynamic: when CBT treatment worked, Area 25 slowed down, as expected, but the frontal areas showed less activity — not more, as had been the case in every other patient group.

“Oh man,” says Mayberg. “I was stumped. For a while I had to just set it aside.” Why did the CBT patients’ frontal activity go from high to low as they got better, rather than vice-versa? She finally realized that the successful CBT patients were almost by definition going to show this pattern. In CBT, patients learn to recognize and change thought patterns that help depress them. An active frontal area, then, was virtually required to make CBT work. The patients who responded to CBT did so either because they were busier thinkers by nature (and therefore more amenable to CBT) or were, when scanned at the beginning of the study, in an earlier stage of depression in which their frontal areas could still rise to the task. The CBT responders entered the study already trying to think their way out of their depression. The scans showing these initial high levels of frontal activity, then, “were pictures,” as Mayberg put it, “of the tug-of-war between the depression and their attempt to self-correct.” When their attempt succeeded, the frontal areas could relax, and the scans showed the reduced activity.

This anomalous CBT result held ripe suggestions about what sort of patients might best respond to CBT versus drug therapy. It also highlighted the key finding uniting all the various studies. For the CBT responders too showed an initially hyperactive Area 25 that calmed as therapy worked and mood improved. Area 25 seemed overly busy in all depressions — and calmed by any successful therapy.

Mayberg now possessed strong, replicated evidence that Area 25 played a key role in depression. This insight fit well with what others had discovered about the dynamics of fear, anxiety, stress, and mood. Researchers like Joseph LeDoux of New York University, and Bruce McEwen, a neuroendocrinologist at Rockefeller University, had shown that mood disorders often develop because extreme or continuous stress, whether from a difficult environment or from the worries our memories and imagination so readily generate, kick fear and anxiety centers into long-term overdrive. The survival systems that have long served us well — a heightened neural and hormonal response to acute threat – turn corrosive when thought and memory trigger them continuously. The evidence for this dynamic was robust. But the key junction boxes and switches in the circuit remained elusive. Maybe, Mayberg started to think, Area 25 was such a switch -- and tweaking it could get the circuit out of alarm mode and back to normal

At about this time, Mayberg took a professorship at the University of Toronto, where she met fellow faculty members Sidney Kennedy, a psychiatrist, and Andres Lozano, a neurosurgeon. Kennedy liked to explore neurological models of depression, and Lozano had gained notoriety modulating another neural network gone awry — Parkinson’s, as it happens. Surgeons had discovered in the 1980s that removing the globus pallidus could help severe Parkinson’s. The globus pallidus serves a role in Parkinson’s much like the role Area 25 serves in mood: it’s a key gateway in neural circuits for movement, and its hyperactivity (as removing it had confirmed) somehow threw the neurology of movement off balance, causing the tremors and rigidity that afflict Parkinson’s patients. In the 1990s, Lozano had become one of several neurosurgeons who treated this problem not by removing the globus pallidus but by inserting just next to it a tiny, low-voltage electrode — a technique called deep-brain stimulation, or DBS — that regulated its activity, returning it to normal. In most patients, this restored movement to near normal.

Might inserting such electrodes alongside Area 25 calm it down? Mayberg, Lozano, and Kennedy decided to try it. And so it came that beginning in 2004, Lozano, with Mayberg sitting in to talk with the patients and observe, implanted DBS electrodes in Area 25 in a dozen severely depressed patients. Lozano drilled a pair of nickel-sized holes in the top of the skull, slid a pair of electrodes and slender leads to Area 25, attached the leads to a small pacemaker sewn in under the collarbone, and turned it on. The pacemaker sends a continuous 4-volt current to Area 25.

The results were stunning. Some patients felt profound relief as soon as Lozano turned on the electrodes, and two-thirds returned to essentially normal mood and function within months. They saw better, thought better, felt better. They talked of walking amid flowers; of “the noise” stopping; of a horrid weight lifting. Side effects were almost negligible

“We still don’t really understand why calming Area 25 has such an effect,” says Mayberg. “That comes next. But it’s clear that it causes depression when it’s hyperactive and that calming it can bring relief.” Indeed the results shattered doubts. Mayberg et alia had shown that in the emerging circuit-board model of mood, one could identify and modulate key switches and conduits. They emphatically confirmed the network model of the brain and, no less, a long history of thought and metaphor. Reason and passion, thought and emotion, were indeed linked in a circle rather than a hierarchy. Neither stood as the other’s slave. Rather they engaged in a conversation that, to be healthy, must be both rich and balanced.


The DBS trial brought Mayberg a brain geek’s version of fame. The renown she doesn’t mind; the affirmation she likes. “It’s nice, after years of writing papers people didn’t finish reading, to have people pay attention. And as a scientist, this is what you really hope for: To feel like you’ve gripped the wheel of a really big ship and changed its direction, even a little bit.

Yet Helen Mayberg hardly thinks she’s found The Way. She doesn’t fancy that she’s solved the Big Questions of mood and mental health. She hopes to find new ways — new tools, new working models — to track and treat the complex network that links thought and mood, cortex and limbic region, and sends us spiraling into depression when it malfunctions.

Most immediately, this means detailing how Area 25 plays so crucial a role.

“I may spend the next ten years trying go figure out what we did,” she told me one evening. “We really did this mostly by eye. I want to figure out how to better work this area. I’d like to better define the neural network — the actual wiring, if you will. I’d like to map the neurochemistry more finely. I want the genetic layout. What will all that tell us about the nature of depression? Can we find more reliable differences among different types of depression? Why do some people respond to drugs and some to CBT?

Many people would flinch at so many questions. Mayberg lights up. “You know what cracks me up?” she says. “When people ask, ‘So where are you going to look next?’ I tell ‘em, ‘What do you mean, “Where I am going to look next?” I’m going to look more closely here.’"