Recent years have brought modern methods of brain stimulation into the mainstream of neurology and psychiatry. But their mechanism—how exactly deep brain stimulation (DBS) relieves Parkinson’s disease symptoms and how repetitive transcranial magnetic stimulation (rTMS) improves depression, for example—remains obscure. Research into this question has largely shifted from effects on isolated target areas to communication within networks of considerable breadth and complexity.
“The question of how the brain is wired together and functions as units in some sort of coherence pattern is what’s driving everything now,” says Helen Mayberg, professor of psychiatry, neurology, and radiology at Emory University and a member of the Dana Alliance for Brain Initiatives. Understanding these systems better, she and others believe, will lead to more precise and effective treatment.
Since DBS involves implanting pacemaker-like electrodes in the brain, researchers suspect it affects neural oscillations, the rhythmic firing of neuron groups that are apparently central to communication among brain regions and go awry in neurological disorders.
“Results suggest DBS overrides pathological rhythms in PD in a twofold way, by suppressing local neuron activity at the cell body, and driving their output at a very high frequency that is apparently less deleterious,” says Peter Brown, who leads the Experimental Neurology Group at Oxford.]
A 2013 study reported in Proceedings of the National Academy of Sciences suggested one way that this process might work. Here, UCSF researchers found that rapid (gamma) rhythms in the motor cortex are "hypersynchronized” with slow (beta) oscillations in the subthalamic nucleus (STN). With the motor cortex locked into this steady beat, rhythmic contractions of tremor overwhelm the varied repertoire of normal movements. In several people with PD, the researchers showed, DBS relaxed the rigid coupling.
An important step toward more effective DBS, Brown thinks, will be fine-tuning stimulation parameters like amplitude and timing to modulate brain activity more precisely. "We happened to have found, by accident, that high frequency stimulation is effective. But there's nothing to say that more intelligently controlled forms of stimulation won't be even more effective."
For example, a “closed-loop feedback” system might monitor electrical activity within the brain and customize stimulation to hit disturbed rhythms at points in their cycle when they are most vulnerable to disruption. Brown and colleagues showed the feasibility of this approach in a recent study, reported in Current Biology, in which they applied transcranial alternating current stimulation to the motor cortex and monitored tremor as an external sign of pathological oscillations. Timing the phase alignment of the current to maximize tremor suppression reduced the symptom by nearly 50%, significantly more than with random timing.
Closed-loop feedback using electrodes implanted in the brain has been shown to improve DBS results in a primate model of PD, but devising ways to do this in humans “with as little additional intervention as possible” represents a formidable challenge, Brown says. “One possibility might be to pick signals up with the same electrode that delivers stimulation."
In depression, clinical studies have repeatedly shown DBS to be effective for about 60% of people with severe unipolar and bipolar episodes and who failed to respond to multiple medications and electroconvulsive therapy. A large multicenter trial, now in progress, seeks FDA approval for this application.
Helen Mayberg, who has been involved in much of this pioneering research, says her lab is now looking beyond proving that DBS works to making it work better.
Patient selection may partly explain why some people recover using DBS and some don't, "but maybe there's also more to implantation than we thought... we need to get the target exactly correct," she says. "We originally targeted an area [the subgenual cingulate gyrus] based on imaging, but subsequently learned that a network of regions is involved, and that we had happened to get into a key node, a set of fibers that impacts this collection of brain areas as a unit."
Much of her research now focuses on locating this key node more precisely. "Our original targeting was necessarily approximate; we soon realized that white matter was as important as grey matter—if you're only in the grey matter, DBS doesn't work. Now we've determined that if you don't get a combination of specific white matter tracts, you won't get the patient all the way well."
In her depression research, Mayberg sees analogies to findings in PD. "They've gone from looking at cell firing to network oscillations. We're doing the same thing." Dysrhythmic oscillations could disrupt the "fluid choreography [through which] emotion colors thinking and actions," she says.
While there's no obvious oscillatory abnormality in depression, there is good reason to infer a more nuanced disturbance that is addressed by DBS. "What happens in the O.R. [when electrodes are implanted] happens immediately,” she says. “It's like throwing a switch. Patients aren't well, but they have the capacity to do things they couldn't do three seconds before.
“What mechanisms in the brain can change that fast? Differences in timing of communication among brain regions. And the language that speaks to that is brain oscillation."
Mayberg seeks to elucidate this process "by listening to the brain in the OR, simultaneously looking at oscillations in the target area and at EEG readings in frontal regions. How does cross talk between area 25 and frontal areas change?"
Going deep with TMS
Research into rTMS, which uses fluctuating magnetic fields outside the skull to create currents within the brain, is in a similar period of exploration and refinement.
Alvaro Pascual-Leone, professor of neurology at Harvard, thinks that rTMS, like DBS, probably alleviates depression by altering interactions between brain networks—including circuitry connecting the prefrontal cortex and subgenual cingulate. "What we're doing fits very well with results Helen has found in targeting the subgenual and following the impact on the PFC. Whether by deep or transcranial stimulation, through a cortical or subcortical entry, we're getting into the same circuit."
As in DBS, outcome may hinge on precise localization. "In studies we've done [reported in Biological Psychiatry in 2012], we've found that 80% of antidepressant variability can be explained by how well the TMS target site in the PFC matched the optimal location for anticorrelation with the subgenual area," Pascual-Leone says.
Because functional connectivity patterns differ among individuals, a customized intervention approach “should be worth exploring,” he said. “Can we identify the optimal location on an individual level with sufficient precision to guide TMS?" In a 2012 Neuroimage paper, Pascual-Leone and colleagues showed this was indeed possible, and his lab is now conducting an NIH-supported study to determine whether it will improve efficacy.
The electrical field generated by conventional TMS only reaches 2 cm into the cortex. Using a different magnet has been shown (on models of the head) to generate fields nearly three times as deep. The “deep TMS” (dTMS) device received FDA approval earlier this year (based largely on its resemblance to existing devices).
“The main theoretical advantage in deep TMS is its capacity to modulate the activity of deeper brain areas,” says Francesco Saverio Bersani of Sapienza University of Rome, who has authored several articles reviewing clinical studies of the device. Such areas might include the mesolimbic reward circuit, which plays a role in mood disorders and addiction.
“It’s an interesting hypothesis, but it’s not proven,” says Pascual-Leone. The dTMS field includes the shallower region affected by rTMS, and these areas are connected, he points out. “If you target a cortical area and a deeper area that receives connections from it, you’re really hitting the deep area twice. It needs to be shown that really matters.”
A meta-analysis by Bersani and colleagues found that dTMS was significantly more effective than conventional TMS and comparable to ECT in reducing depression symptoms in patients who had failed to respond to medication. But the studies, Pascual-Leone points out, were not head-to-head comparisons, and were underpowered (statistically) to establish differences.
According to Bersani’s other review, dTMS trials have reported positive effects in bipolar disorder, for auditory hallucinations and negative symptoms in schizophrenia, and for autism. Recent studies suggest efficacy for substance use problems as well.
These are mostly small series and case studies. “The general results are convincing in terms of safety and tolerability and promising in terms of clinical effectiveness,” says Bersani. “But comparison with standard TMS is needed to determine whether dTMS represents a real therapeutic advance or not.”
In any case, he suggests, a new TMS permutation should have advantages in an era of personalized medicine. “Every person is biologically unique, and a wide range of therapies will be necessary to address these differences.”