Alison Motluk reports

TONY IS A STUTTERER. He trips over every second word. It takes him eleven tries before he can even say his own name. Asking the time or ordering a cup of coffee can be excruciating, and making a telephone call is worse: people hang up. From the time he learnt his first words, Tony has had trouble getting them out of his mouth, and if he could say the word "frustrated" without fumbling, he'd tell you that that's what he is.
 This spring, Tony will take part in a study in Texas that aims to eliminate his stutter altogether. At the University of Texas Health Science Center in San Antonio, a team of neuroscientists will try to interrupt the faulty circuit in his brain that they think causes his stuttering. They will start by temporarily paralysing parts of his brain with localised magnetic pulses. If that goes according to plan, they may try something more permanent--brain surgery. "We're seriously considering lesion work," says Peter Fox, who is leading the study.
 It sounds drastic: brain surgery for stuttering. But this speech disorder is severe and disabling. The volunteers have already tried every therapy on offer, says Roger Ingham, a speech pathologist at the University of California at Santa Barbara, another member of the team. "Some of the subjects have said quite explicitly," says Fox, "that they're willing to do whatever it takes."
 Stuttering is not the only unlikely-sounding disorder for which neurosurgery may soon promise a cure. Neuroscientists at the San Antonio unit are considering similar treatments for the movement disorders caused by prolonged use of antipsychotic drugs and even for post-traumatic stress disorder. Elsewhere in the US, and at centres in London and Stockholm, neurologists are investigating whether disorders ranging from migraine to Tourette syndrome might also someday be candidates for surgery.
 Such treatments are still highly experimental. It could be some years before the first operation for stuttering is actually performed. But the pendulum that swung away from the idea of surgical intervention for neurological diseases over the past few decades is now swinging back, says Mahlon DeLong, head of neurology at Emory University in Atlanta, Georgia. "And it's swinging with a great deal more understanding behind it," he says. "We are on the threshold of a very exciting era. We are now appreciating the potential to treat some of these disorders that don't adequately respond to drug treatments."
 Driving the surgical renaissance are two powerful technologies that go by the acronyms PET and TMS, and some new and cunning ways of turning neurological disorders on and off (see "Throwing the switch"). Fox calls it "systems neurosurgery". PET, or positron emission tomography, is a brain imaging technique that allows neurologists to see which brain regions are involved when a person speaks, say, or twitches. TMS, or trans- cranial magnetic stimulation, is a newer technology that uses magnetic pulses to temporarily disable parts of the brain. So while PET scans can tell you which brain regions are activated, TMS can test theories about whether switching off one of those regions will make the condition disappear.
 The idea of selectively injuring the brain to make it function better is certainly not new. In the heyday of psychosurgery, in the 1940s and 1950s, thousands of people were put under the knife for disorders like schizophrenia. Some improved dramatically; others showed no change or, worse, were left devastated.
Therein lies a good deal of popular suspicion. Mention brain surgery for anything short of a life-threatening tumour--let alone a speech impediment--and the word "lobotomy" is on everyone's lips. "Everyone sees Nicholson being punished," says Per Mindus, a neuropsychiatrist at Stockholm's Karolinska Institute, referring to the fate of Jack Nicholson's character in the movie One Flew Over the Cuckoo's Nest. But it is not only the lay public that is uneasy. Tomá`´s Paus of the Montreal Neurological Institute in Canada, like some other neuroscientists, worries that surgery for stuttering may be going too far. "It is pretty radical to remove a part of the brain," he says.
 An important landmark for the new era of systems neurosurgery came in the late 1980s. By first developing a model for the brain circuitry of people with Parkinson's disease, DeLong's group was able to explain for the first time why a particular operation, which had been performed on the brains of Parkinson's patients during the 1940s, achieves some success.
 Until that time, most brain surgery was based on trial and error and intelligent guesswork. DeLong and his team tested their theories first. They gave monkeys a substance known as MPTP, a variation of the drug Demerol, which kills off the same cells that degenerate in Parkinson's disease and produces parkinsonian symptoms. Then they inserted microelectrodes into the brains of the monkeys and read the firing patterns of their neurons, establishing how the brain's circuitry had gone awry.
 In Parkinson's disease, cells in the substantia nigra, where the neurotransmitter dopamine is produced, start to die off. The substantia nigra forms part of the basal ganglia, which is tucked deep within the brain. Normal movement depends crucially on there being the right amount of dopamine in the basal ganglia. Too little, and patients suffer muscular rigidity, tremors, balance problems and slowed movement.
 DeLong was among the first scientists to propose that, in fact, many parkinsonian symptoms are not caused directly by cell loss in the substantia nigra. What's more, he showed that in the MPTP monkeys parts of the globus pallidus, a thumbnail-sized structure on the edge of the midbrain, had become overactive. This explained why burning out a tiny part of the globus pallidus could reverse the symptoms. He did just that in some of his MPTP monkeys--and the once-rigid primates were suddenly able to move again.
 Since DeLong conducted his first human pallidotomy operation in 1992, hundreds more patients with severe Parkinson's disease have chosen to try it, and many have shown dramatic improvement. It is no longer even considered experimental.
DeLong's investigation into Parkinson's took more than a decade. But with the newer techniques, says Fox, many neurological disorders could now be studied in half that time--and in the living human subject. Some disorders, like stuttering, may be candidates for surgery.
When a stutterer stumbles over words, the brain functions in an unusual way: regions switch on that are supposed to be dormant, connections are made where there should be none. "To understand the circuitry of a disorder," says Fox, "you need a phenomenon to exist, then not exist."
 Stuttering lends itself particularly well to this approach, since it can be switched on and off very simply. When a volunteer reads out loud in chorus with a nonstutterer, the speech problem completely disappears. With PET, Fox and his colleagues have already taken snapshots of the volunteers' brains during episodes of stuttering, and then, during chorus reading, in respite. By comparing these images with each other, and with those of nonstutterers, they were able to identify where the brain goes wrong during stuttering. They reported their findings in Nature last summer.
Out loud
When anyone reads aloud, many parts of the brain light up with activity. The parts of the motor cortex that control the movements of the mouth are active, for instance, as well as the part of the visual cortex which is involved in reading. In nonstutterers, parts of the auditory cortex also light up, presumably because they are busy monitoring their own speech. Another region, called the supplementary motor area, or SMA, is also activated. It is thought to transform the intention to speak into the action of speaking. For those without a stutter, almost all this activity happens on the left side, the hemisphere of the brain that is dominant in language.
 The PET scans of Fox's stuttering volunteers reveal an entirely different pattern. The region that controls movement of the mouth lights up on the right side, not the left. There is virtually no auditory self-monitoring, and activity in the SMA is extremely intense compared with nonstutterers, and occurs in slightly different areas. Also, motor areas of the cerebellum are more than twice as active in stutterers. Finally, a spot in the right superior premotor area of the cortex, which is shut down in normal speakers, glows in stutterers; its location suggests it is "likely to be involved in movement planning", Fox says. Interestingly, when the stutterers read in chorus, without the usual machine-gun pattern, the different regions of their brains function in almost exactly the same way as those of people who never stutter (New Scientist, Science, 13 July 1996, p 15).
But it's a long way from finding the pieces that make up the puzzle to understanding how they all fit together--let alone which piece you can take away to improve the picture. Before they can contemplate surgery, the scientists have to work out where to intervene.
 In a few weeks' time, Fox and Ingham will start to study their five male volunteers. With TMS, which aims short, rapid pulses of magnetism at specific sites in the brain, they will make what Fox calls a "transient lesion", a temporary injury to a precise part of the brain, in their wide-awake volunteers. With any luck, it will make them stutter-free--if only for a few seconds. "Even transient effects would be encouraging in pointing us towards a more invasive procedure," says Fox, although he is hopeful that the effect will last for days, if not weeks.
Fox and Ingham have decided to target the spot in the right superior premotor area that is involved in planning movement. Fox says it is the most obviously unusual area in stutterers; in normal speakers it is not active at all. "It just doesn't belong there," he says.
 If disabling this region by magnetic stimulation temporarily stops the stutters as he expects, Fox and his colleagues will do PET scans to see exactly what is happening in different parts of the brain. "Do areas turn up, do they turn down?" asks Fox. "We don't know any of that." If the effect is long lasting, they will scan every few days to watch how it fades.
Future unknown
Fox and Ingham have not ruled out the possibility that TMS might become a long-term treatment in itself. TMS has already been shown to alleviate the symptoms of depression for weeks or even months ("Happiness is a magnet", New Scientist, 5 August 1995, p 24). Severe stutterers might be able to nip in for a magnetic fix a few times a year.
 Still, the long-term effects of TMS are unknown, Fox points out. Does it have a cumulative effect, becoming more and more effective over time, until the person is more or less cured? Or does a person become immune to it, so that after a year or two magnetism has no impact at all? "Both of these effects have been seen," says Fox, "so we are walking into the unknown."
 There are other uncertainties too. Roger Lemon, head of neurophysiology at the Institute of Neurology in London, agrees that it is "an intelligent way forward" to use a noninvasive tool like TMS to work out what's going wrong before taking the huge step of destroying tissue permanently. "The worry I would have," he says, "is that it assumes the magnetic stimulation is affecting exactly the same population of neurons in the cortex that will eventually be cut out."
 Fox says that he and his team will tread cautiously. They will examine how effective TMS is in attenuating the stutter, what else is affected if the premotor region is knocked out, and, if possible, how the effects wear off. "We will explore the noninvasive treatment quite thoroughly," he says. "But a surgical lesion might be the right answer."
 And if the stuttering experiment works, other disorders will be queuing up for investigation as candidates for systems neurosurgery. "A whole spectrum of disorders could benefit," says DeLong. Helen Mayberg, a colleague of Fox's at the Health Science Center, is looking closely at mood disorders and depression, and has come up with innovative ways of comparing which parts of the brain are activated in people with severe clinical depression compared with nondepressed people. Elsewhere, neurologists are exploring the possibility of using the systems approach to unravel the neurological defects that underlie panic disorder, obsessive compulsive disorder and a variety of movement disorders. "It is a new horizon," says Fox. For neurological disease, "this strategy of experimentation is coming."
Throwing the switch
Stuttering can be switched off simply by asking stutterers to read along with normal speakers. Some other symptoms of neurological disorders can be turned on and off equally easily, allowing researchers to compare brain activity in the normal and abnormal states. Want a snapshot of panic in a person who suffers from panic disorder? Sodium lactic acid, administered intravenously, does the trick. The PET scans look a lot like the PET scans of people who are expecting an electric shock. The brains of people with obsessive compulsive disorder may light up when they desperately want to remove filthy rags from their hands...and scrub. Cluster headaches can be switched on for the scanner by spraying nitroglycerin onto the tongue of a sufferer. A technique called "scenario-based recall" induces states of mind like sadness in normal people. They are taught to call sad events to mind, then forget the events, but hold onto the sadness. Under the PET scanner, this sadness looks a lot like some stages of clinical depression.
From New Scientist, 1 February 1997