We are electric, p.13
We Are Electric, page 13
No one could propose implanting a stop switch into a human brain for such speculative reasons. But soon a much more compelling use case presented itself.
A pacemaker for the brain
It was an otherwise quiet morning in 1982, and a patient named George had just been admitted into a psychiatry unit with a diagnosis of catatonic schizophrenia. The designation had been applied not because it fitted, but because nothing else did. The patient was unresponsive, but he still seemed alert, a combination outside all the existing frameworks available at the time. The psychiatrists were sure the patient had a neurological disorder, and the neurologists were sure the patient had a psychiatric disorder. Eventually, the chief resident ran to the office of the director of neurology, Joseph Langston.
Langston set aside his coffee, looked up from that morning’s EEG reports, and began conducting his own tests and consultations. Initially, he concluded George exhibited all of the symptoms of advanced Parkinson’s disease, a cruel neurodegenerative disorder whose iconic symptom is a trembling so violent, a person can’t so much as hold a cup of water, and which progresses after many years into a rigid freeze. But Langston knew the diagnosis couldn’t be right for two reasons. The patient was only in his early forties, which was about twenty years too young for a diagnosis of Parkinson’s. And instead of manifesting gradually over the course of years or even decades, his (apparently) end-stage symptoms had appeared literally overnight.
The mystery only deepened further when they found George’s girlfriend frozen in the same state despite being even younger – she was only thirty. Eventually, the team located five more identical cases. It took some sleuthing and luck, but Langston and the police eventually uncovered the common factor: each of the patients had recently used heroin – or at least what they thought was heroin. When Langston’s group got their hands on some samples, what they found wasn’t heroin at all: the street chemists had mistakenly synthesised a compound called MPTP. A search through the medical literature revealed several existing studies on MPTP, and their findings didn’t bode well for the couple. By destroying an area deep inside the brain called the substantia nigra, MPTP had been found to create irreversible symptoms that mirrored Parkinson’s – notably that rigid freeze.
The discovery of that relevant brain area was consequential. During the 1970s, a few neurosurgeons had been experimenting with implanted electrodes for chronic pain and epilepsy. They drilled open the skull and pushed a penetrating electrode deep into the grey matter. It was a promising solution to the big problem that had driven psychosurgery out of business: unlike the traditional approach of burning or cutting out bothersome bits of the brain, ‘electrical lesions’ were adjustable as well as reversible. If you were applying too little electricity, you could dial up more; if you were putting in too much, you could dial it back down.
As they did this, doctors began to notice two patterns among the patients exhibiting troublesome symptoms: first, the electrical stimulation alone was sometimes enough to blunt the symptoms. Second, the faster the electrical stimulation pulsed, the more the patients improved.
These patterns were intriguing, but the electrodes couldn’t be implanted for take-home care. Like Hyman’s earliest pacemaker, they were hooked up to an unwieldy external apparatus, a large power source which was connected to the electrodes that stuck out of the head by wires.38 Moreover, no one had done any large clinical trials to verify that stimulating some particular brain area worked for everyone, or if it was just bespoke to the individual patient. The only assessment of whether it worked at all was the assurance of the implanting surgeon.39 However, all the growing interest had led Medtronic to work on adapting their pacemaker to make it suitable for brain implantation. They sent their experimental devices to specialist centres, and even trademarked the term DBS (‘deep brain stimulation’). Their devices were still limited to small, one-off experiments. But that all changed when George’s revelatory case study made it to the desk of Alim-Louis Benabid at the University Hospital in Grenoble, and he connected the dots.40
Benabid was among the few psychosurgeons still using implanted electrodes to identify the correct brain area in advance of psychosurgery. He had become fascinated with the clear and obvious effects he saw in his Parkinson’s patients: symptoms would quiet in real time in the operating theatre. When he grasped the significance of George’s case study, he obtained Medtronic’s new brain pacemakers and implanted a handful of patients. The improvements were dramatic. Stopping the faulty neurological code emerging from that part of the brain settled the trembling, allowing patients to once again move their limbs as they wished. Medtronic hired Benabid to design massive trials. Now, instead of unscientific zaps accompanied by an unconvincing thumbs-up from a clinician, here was a disease with extremely well-spelled-out symptoms, a relevant brain area, and a robust response to an already approved medical device.
Medtronic had been desperate to find a way to expand its hugely successful pacemaker business. In Benabid’s work, they saw a new opportunity. And in trial after trial they found the same dramatic effects: start the current flowing through those deep electrodes, and trembling subsided instantly. People who before the surgery could not hold a cup of tea were now able to confidently brew themselves a whole pot. EU regulators approved the implant for Parkinson’s in 1998, and the Food and Drug Administration (FDA) approved the treatment in the US in 2002. One of the doctors who began implanting his patients hailed it as ‘providing a new life’. The pacemaker had moved into the brain and deep brain stimulation was born.
More than 160,000 of these ‘brain pacemakers’ have been implanted to mitigate the disabling muscle spasms of people with Parkinson’s, essential tremor, and dystonia.41 As brain surgeries go, it is actually pretty straightforward. First, drill two holes in the skull. Next, push two metal electrodes, each about the dimension of dry spaghetti, into the region of the brain responsible for the symptoms. Finally, snake the wire through the head and down the neck until it reaches, implanted under the skin near the collarbone, an object about the size of a stopwatch. That’s the pacemaker! But now it’s sending a current whose pulse and amplitude, over the following couple of weeks, a technician will adjust until the symptoms subside.
In big clinical trials with many participants, knowing which brain area to overwhelm with electricity has allowed surgeons to successfully short-circuit faulty signals in these broken regions. What other brain areas were amenable to freezing in this way, to widen the net of conditions controllable by a pacemaker? Many small trials offered clues to the next big ailment.
In 1999, researchers at the Catholic University of Leuven (KUL) in Belgium implanted DBS electrodes into an area called the internal capsule in four people with severe obsessive-compulsive disorder. Symptoms improved for three.42 More trials on other ailments followed in rapid succession – again small, investigational studies, often only with ten or fewer participants. But despite their small size, these trials generated dramatic headlines, as my colleague Andy Ridgway noted in 2015 in New Scientist .43 DBS allowed a thirteen-year-old autistic teenager to speak for the first time.44 It freed people with Tourette’s from bone-breaking physical tics. It allegedly stopped obese people from overeating and anorexic people from undereating.45 The small trials proliferated – what else would yield to a brain pacemaker? Anxiety? Tinnitus? Addiction? Paedophilia?46
Medtronic bet on depression. They weren’t alone in thinking this was promising. The idea had been germinating since 2001, when neuroscientist Helen Mayberg had the idea to investigate DBS for intractable depression (the kind that refuses to budge no matter the treatment). DBS, she told me when I met her at the International Neuroethics Society symposium in San Diego in 2018,47 ‘seemed to block abnormal brain function in Parkinson’s, so we wanted to block our own depression-specific area’. Mayberg focused on a brain region known as Brodmann’s area 25, which has been dubbed the brain’s ‘sadness centre’. Too much activity here, Mayberg and her colleagues thought, was driving symptoms like negative mood and that characteristic lack of a will to live. What would happen if you froze those neurons? Four of her first six patients saw dramatic improvement.48 Twenty more small trials recorded improvement rates as high as 60 or 70 per cent. ‘People come out of that very dangerous state, and stay well,’ Mayberg told Ridgway. ‘They just get back on the bus.’ In other trials all over the world, depression lifted similarly.
After enough such intriguing results had piled up, St. Jude Medical – a Medtronic rival – took the plunge and bankrolled a major trial. It seemed all but guaranteed to culminate in the first new commercial application of DBS since Parkinson’s. Two hundred participants at more than a dozen medical centres got the implant. The buzz was immense. And then, after six months, the trial stopped. Gossip spread among industry insiders that it had failed an FDA futility analysis, meant to make sure spendy trials get kneecapped if they are clearly wasting time and money. There were stories of terrible side effects and a suicide attempt.49 The implication – bad for the future of the technology – was that there was no difference between a placebo and an implant.50
When all the drama and recriminations settled, the story ended up being a lot weirder and less straightforward than the first reports suggested: as the journalist David Dobbs concluded in a deeply reported post-mortem analysis published in The Atlantic in 2018, it appeared to be a case of a treatment that actually seemed to work being sabotaged by a trial that didn’t. For the people in whom it worked, it was a gift, yielding immediate results that were so dramatic as to be nearly magical. ‘What did you do?’ would be the response of the awake patient, still in the operating room, the moment the stimulator was turned on. And when that happened, the results were enduring. ‘If you got better, you stayed better, given continued stimulation,’ Mayberg told several media outlets in the wake of the trial. The same pattern held true for obsessive-compulsive disorder: those patients who had responded well in a brain stimulation trial at the Catholic University of Leuven still had their obsessive-compulsive disorder under control after fifteen years. ‘It was like someone spring-cleaned out my brain and took out all the unnecessary thoughts,’ another participant told Alix Spiegel, host of the National Public Radio show Invisibilia.51
But it was just not possible to predict who would have a miracle and who wouldn’t. There were also some strange side effects.52 The deep, ancient areas of the brain targeted for electrification in depression and Parkinson’s are involved in much more than motor and mood control. They are implicated in learning, emotion, and reward – which is to say, addiction. Interfering with these had unpredictable consequences. This was the case for one Dutch man being treated for severe obsessive-compulsive disorder, anonymised by his doctors at the University of Amsterdam as Mr B. His new brain implant had been working for just a few weeks when he chanced upon a recording of Johnny Cash’s ‘Ring of Fire’. In the five decades before the twin electrodes penetrated his deep brain, he had never been especially moved by music – he was the kind of guy who claimed to like both The Beatles and the Rolling Stones. The day Johnny Cash’s voice hit his newly electrified pleasure centres, however, all that changed. From then on, no other music was allowed. Mr B bought every Johnny Cash CD and DVD he could get his hands on. But when the electrical stimulator was off, he could not for the life of him recall what was so important about Johnny Cash.53
Not all side effects were as endearing, though. People with Parkinson’s implants have reported an increase in impulse control disorders, such as excessive gambling and hypersexuality.54
This reflects a somewhat uncomfortable open secret about DBS: despite all the complicated talk about the function of precise areas of the brain, no one is sure exactly how DBS works.55 As recently as 2018, academic reports described DBS as an effective but ‘poorly understood’ treatment, even in Parkinson’s and the other motor diseases it has been approved to treat for decades.56 ‘If you think of neurons executing the neural code as playing a melody on a piano,’ says Kip Ludwig, a former director at the US National Institutes of Health (NIH), ‘then DBS is like playing that piano with a mallet.’
There’s a limit to this approach. Electrifying specific brain areas could broadly control some maladies, but it was impossible to get truly granular enough to reliably hit a target as ephemeral as depression. We needed to know exactly what was happening in the brain’s code as a response to those zaps.
For that, we needed to decipher the neural code.
Reading the neural code
By the 1970s, Francis Crick had grown bored of molecular biology, even though he’d essentially invented it. He was looking to solve his next great mystery. If cracking the blueprint for life had been exciting, how about cracking the secret of consciousness? So in 1977, he left Cambridge for the Salk Institute in California, where he turned his attention to what he considered a deeply unpromising approach to neuroscience. He demanded new ‘theories dealing directly with the processing of information’ and ways to tie behaviours and actions to the neural firings that accompanied them.
In 1994, he summarised his research in The Astonishing Hypothesis, a slim volume that would have an explosive impact on neuroscience and philosophy. ‘One may conclude,’ he wrote, ‘that to understand the various forms of consciousness we need to know their neural correlates’.57 He further argued that all those things we think or feel or see ‘are, in fact, no more than the behaviour of a vast assembly of nerve cells and their associated molecules’.58 (He did not address how this was materially different from our identity being no more than the behaviour of a vast assembly of genes.) The subtitle – The Scientific Search for the Soul – made clear the book’s ambition.
In the whole two decades before Crick’s book was published, fewer than ten peer-reviewed scientific papers referred to the term ‘neural code’. After the release of The Astonishing Hypothesis, however, neuroscientists increasingly turned their attention to trying to find the neural signatures of a vast array of behaviours and thoughts. For students of the sensorium, neural code was the new black.
Not that they knew what the term meant. Even as Crick was writing his book, the definition of the term was the subject of a contentious spat in neuroscience. Adrian’s ideas that information could be encoded in the Morse code dots passed by individual neurons still had its adherents, but now there was a fresher idea. Brain plasticity – summarised by the axiom ‘neurons that fire together, wire together’ – became ubiquitous thanks to its succinct explanation of how different neurons learn to work together as you learn different skills, from language to ballet. A real neural code couldn’t possibly focus on the firing of single neurons, representatives of the new guard wrote in 1997, but had to take into account the way vast collections of different neurons fired together in synchrony to form a coherent pattern over time and space.59
It would be one heck of a hard thing to measure. By then, we had begun to grasp the sheer size of the brain – 86 billion neurons. No tool was, or is (or possibly will ever be), capable of reading the activity of all those neurons at the same time. But as the twenty-first century approached, we had options.
The trusty EEG was still in use, a workhorse that had given us the different waves that could reveal focus and inattention, but also much more. Scientists spent decades using these readings to advance our understanding of sleep. Because EEG doesn’t require opening the skull, just some electrodes on the scalp, scientists were able to obtain a lot of data from a lot of people. The EEG evolved from its humble origins in Hans Berger’s lab into skullcaps encrusted with dozens of electrodes that could read subtle variations in the chorus of the brain’s billions of denizens. That helped drill down into the brainwaves, and the subsequent discovery of delta and gamma waves (in addition to Adrian’s alpha and beta) helped researchers identify the different stages of sleep, yielding the familiar, ever-deeper stages of I–IV and dreaming sleep. Other work tied characteristic disruptions in these waveforms to sleep disorders and neurological disorders – even helped identify the location of brain tumours. Thanks to increasingly powerful computer processing power and better signal-processing algorithms, EEG could more finely analyse the brain’s patterns. Depression was correlated with an overabundance of alpha waves in the EEG. In Parkinson’s disease, there was a paucity of beta waves. Alzheimer’s patients have been found to have a deficit of high-amplitude gamma waves. Various research papers reported a rainbow of emotions correlated to waveforms Berger couldn’t have dreamed of.60
Another tool, electrocorticography (ECoG), could get deeper into the brain, but it would be amenable to a smaller population. It looked like a mat of electrodes placed directly onto the exposed folds of the brain, a bit like a doily on a side table, to record electrical activity from the cortex. It does require opening the skull, which is why it is rare to obtain these sorts of traces. The only human volunteers for this kind of brain reading have already had their skull opened for unrelated investigational purposes. Sometimes these individuals gave researchers permission to put the mesh on their brains and read the neural correlates of certain specific thoughts, but it still didn’t let the researchers get next to any specific neurons.
For that, you needed an invasive brain-penetrating electrode. The first of these was approved to stick into human brains in the 1990s. It was called the Utah array, and it looked like a small metal square with ninety-six electrodes sticking out of it, a bit like a bed of nails for a ladybird. Nestled into brain folds, it could record the squabbling of many neurons talking to each other, or hone in on one in particular. But this reading was the most invasive of all: it required not only opening the skull (or drilling a hole in it), but also jamming the penetrating electrode through the blood–brain barrier, and running a lead out of the skull to power and listen to the array. The only subjects on whom it was considered ethical to test this device were animals – and later, people who had suffered the kind of irreversible physiological trauma that made this the last faint hope for the possibility of a research breakthrough that would help them.
