Google+

BBC Future

Mosaic@Future

Reading the minds of the ‘dead’

(Thinkstock)

(Thinkstock)

Thousands of patients remain trapped in a vegetative state between life and death. Three scientists are working to free them, as Roger Highfield reports.

“Imagine you wake up, locked inside a box,” says Adrian Owen. "It's a perfect fit, down to every last one of your fingers and toes. It's a strange box because you can listen to absolutely everything going on around you, yet your voice cannot be heard. In fact, the box fits so tightly around your face and lips that you can't speak, or make a noise. At first, this feels like a game. Then reality sets in. You see and hear your family lamenting your fate. You're too cold. Then too hot. You're always thirsty. The visits of your friends and family dwindle. Your partner moves on. And there's nothing you can do about it."

Owen and I are talking on Skype. I’m sitting in London, UK, and he’s in another London three-and-a-half thousand miles away at the University of Western Ontario, Canada. Owen’s reddish hair and close-cropped beard loom large on my screen as he becomes animated describing the torment of those with no voice: his patients.

People in a “vegetative state” are awake yet unaware. Their eyes can open and sometimes wander. They can smile, grasp another’s hand, cry, groan or grunt. But they are indifferent to a hand clap, unable to see or to understand speech. Their motions are not purposeful but reflexive. They appear to have shed their memories, emotions and intentions, those qualities that make each one of us an individual. Their minds remain firmly shut. Still, when their eyelids flutter open, you are always left wondering if there’s a glimmer of consciousness.

A decade ago, the answer would have been a bleak and emphatic no. Not any longer. Using brain scanners, Owen has found that some may be trapped inside their bodies yet able to think and feel to varying extents. The number of patients with disorders of consciousness has soared in recent decades, ironically, because doctors have steadily got better at saving patients with catastrophic injuries.

Today, trapped, damaged and diminished minds inhabit clinics and nursing homes worldwide – in Europe alone the number of new coma cases is estimated to be around 230,000 annually, of which some 30,000 will languish in a persistent vegetative state. They are some of the most tragic and expensive artefacts of modern intensive care.

Owen knows this only too well. In 1997, a close friend set off on her usual cycle to work. Anne [name changed] had a weak spot on a blood vessel in her head, known as a brain aneurysm. Five minutes into her trip, the aneurysm burst and she crashed into a tree. She never regained consciousness.

(Alamy)

Images of people in comas in films such as Hable con Ella (Talk to Her) are a long way from the reality (Alamy)

The tragedy left Owen numb, yet Anne’s accident would shape the rest of his life. He began to wonder if there was a way to determine which of these patients were in an unconscious coma, which were conscious and which were somewhere in between?

That year, he had moved to the Medical Research Council’s Cognition and Brain Sciences Unit in Cambridge, where researchers used various scanning techniques. One, positron emission tomography (PET), highlights different metabolic processes in the brain, such as oxygen and sugar use. Another, known as functional magnetic resonance imaging (fMRI), can reveal active centres in the brain by detecting the tiny surges in blood flow that take place as a mind whirrs. Owen wondered whether he could use these technologies to reach out to patients, like his friend, stuck between sensibility and oblivion.

Conscious decision

Half a century ago, if your heart stopped beating you could be pronounced dead even though you may have been entirely conscious as the doctor sent you to the morgue. This, in all likelihood, could explain notorious accounts through history of those who ‘came back from the dead’. As recently as 2011, a council in the Malatya province of central Turkey announced it had built a morgue with a warning system and refrigerator doors that could be opened from the inside.

The problem is that the scientific definition of “death” remains as unresolved as the definition of “consciousness”. Being alive is no longer linked to having a beating heart, explains Owen. If I have an artificial heart, am I dead? If you are on a life-support machine, are you dead? Is a failure to sustain independent life a reasonable definition of death? No, otherwise we would all be “dead” in the nine months before birth.

The issue becomes murkier when we consider those trapped in the twilight worlds between normal life and death – from those who slip in and out of awareness, who are trapped in a ‘minimally conscious state’, to those who are severely impaired in a vegetative state or a coma. These patients first appeared in the wake of the development of the artificial respirator during the 1950s in Denmark, an invention that redefined the end of life in terms of the idea of brain death and created the specialty of intensive care, in which unresponsive and comatose patients who seemed unable to wake up again were written off as “vegetables” or “jellyfish”. As is always the case when treating patients, definitions are critical: understanding the chances of recovery, the benefits of treatments and so on all depend on a precise diagnosis.

In the 1960s, neurologist Fred Plum in New York and neurosurgeon Bryan Jennett in Glasgow carried out pioneering work to understand and categorise disorders of consciousness. Plum coined the term “locked-in syndrome”, in which a patient is aware and awake but cannot move or talk. With Plum, Jennett devised the Glasgow Coma Scale to rate the depth of coma, and Jennett followed up with the Glasgow Outcome Scale to weigh up the extent of recovery, from death to mild disability. Together they adopted the term “persistent vegetative state” for patients who, they wrote, “have periods of wakefulness when their eyes are open and move; their responsiveness is limited to primitive postural and reflex movements of the limbs, and they never speak.”

In 2002, Jennett was among a group of neurologists who chose the phrase “minimally conscious” to describe those who are sometimes awake and partly aware, who show erratic signs of consciousness so that at one time they might be able to follow a simple instruction and another they might not. Even today, however, we’re still arguing over who is conscious and who isn’t.  

Revealing scan

Kate Bainbridge, a 26-year-old schoolteacher, lapsed into a coma three days after she came down with a flu-like illness. Her brain became inflamed, along with the primitive region atop the spinal cord, the brain stem, which rules the sleep cycle. A few weeks after her infection had cleared, Kate awoke from the coma but was diagnosed as being in a vegetative state. Luckily, the intensive care doctor responsible for her, David Menon, was also a Principal Investigator at the newly opened Wolfson Brain Imaging Centre in Cambridge, where one Adrian Owen then worked.

In 1997, four months after she had been diagnosed as vegetative, Kate became the first patient in a vegetative state to be studied by the Cambridge group. The results, published in 1998, were unexpected and extraordinary. Not only did Kate react to faces: but her brain responses were indistinguishable from those of healthy volunteers. Her scans revealed a splash of red, marking brain activity at the back of her brain, in a part called the fusiform gyrus, which helps recognise faces. Kate became the first such patient in whom sophisticated brain imaging (in this case PET) revealed “covert cognition”. Of course, whether that response was a reflex or a signal of consciousness was, at the time, a matter of debate.

The results were of huge significance for science but also for Kate and her parents. “The existence of preserved cognitive processing removed the nihilism that pervaded the management of such patients in general, and supported a decision to continue to treat Kate aggressively,” recalls Menon.

Kate eventually surfaced from her ordeal, six months after the initial diagnosis. “They said I could not feel pain,” she says. “They were so wrong.” Sometimes she’d cry out, but the nurses thought it was just a reflex. She felt abandoned and helpless. Hospital staff had no idea how much she suffered in their care. Kate found physiotherapy scary: nurses never explained what they were doing to her. She was terrified when they removed mucus from her lungs. “I can’t tell you how frightening it was, especially suction through the mouth,” she has written. At one point, her pain and despair became so much that she tried to snuff out her life by holding her breath. “I could not stop my nose from breathing, so it did not work. My body did not seem to want to die.”

Kate says her recovery was not so much like turning a light on but a gradual awakening. It took her five months before she could smile. By then she had lost her job, her sense of smell and taste, and much of what might have been a normal future. Now back with her parents, Kate is still very disabled and needs a wheelchair. Twelve years after her illness, she started to talk again and, though still angry about the way she was treated when she was at her most vulnerable, she remains grateful to those who helped her mind to escape.

She sent Owen a note.

Dear Adrian,Please use my case to show people how important the scans are. I want more people to know about them. I am a big fan of them now. I was unresponsive and looked hopeless, but the scan showed people I was in there. It was like magic, it found me.

In a forested campus south of Liege, Steven Laureys studies vegetative patients in research that dates back decades. Working there as part of the Cyclotron Research Centre in the 1990s, he was surprised when PET brain scans revealed that the patients could respond to a mention of their own name: meaningful sounds produced a change in blood flow within the auditory primary cortices. Meanwhile, on the other side of the Atlantic, Nicholas Schiff was finding that within catastrophically injured brains lay partially working regions, clusters of remnant neural activity. What did it all mean?

Anyone for tennis?

At that time, doctors thought they already knew the answers: no patient in a persistent vegetative state was conscious. Never mind that staring at images made the brain light up, they carped: you can do that in a sedated monkey. Based on previous experience, a brain starved of oxygen as a result of a heart attack or a stroke was unlikely to recover if it didn’t in the first few months. These patients had suffered a fate that many people regarded as worse than death itself: they were functionally brainless. Undead. Doctors, with the best intentions, thought it was perfectly acceptable to end the life of a vegetative patient by starvation and the withdrawal of water. This was the age of what Laureys calls “therapeutic nihilism”.

What Owen, Laureys and Schiff were proposing was a rethink of some of the patients who were considered vegetative. A few of them could even be classed as being fully conscious and locked-in. The establishment was doggedly opposed. “You cannot imagine the environment in the late 1990s,” says Schiff. “The hostility we encountered went well beyond simple scepticism”. Looking back, Laureys pauses and smiles thinly: “Medical doctors do not like to be told they are wrong.”

Then came 2006. Owen and Laureys were trying to find a reliable way to communicate with patients in a vegetative state, including Gillian [name changed]. In July 2005, this 23-year-old had been crossing a road, chatting on her mobile phone. She was struck by two cars.

Five months later, a strange stroke of serendipity allowed Gillian to unlock her box. The key arose from a systematic study Owen started with Laureys in 2005. They had asked healthy volunteers to imagine doing different things, from singing songs or conjuring up the face of their mother. Then Owen had another idea. “I just had a hunch,” he says. “I asked a healthy control to imagine playing tennis. Then I asked her to imagine walking through the rooms of her house.” Imagining tennis activates part of the cortex, called the supplementary motor area, involved in the mental simulation of movements. But imagining walking around the house activates the parahippocampal gyrus in the core of the brain, the posterior parietal lobe, and the lateral premotor cortex. The two patterns of activity were as distinct as a ‘yes’ and a ‘no’. So, if people were asked to imagine tennis for ‘yes’ and walking around the house for ‘no’, they could answer questions via fMRI.

Gazing into Gillian’s ‘vegetative’ brain with the brain scanner, he asked her to imagine the same things – and saw strikingly similar activation patterns to the healthy volunteers. It was an electric moment. Owen could read her mind.

(Science Photo Library)

Functional magnetic resonance imaging, or fMRI, measures activity in the brain by detecting changes in blood oxygenation and flow (Science Photo Library)

Gillian’s case, published in the journal Science in 2006, made front-page headlines around the world. The result provoked wonder and, of course, disbelief. “Broadly speaking, I received two types of email from my peers,” says Owen. “They either said ‘This is amazing – well done!’ or ‘How could you possibly say this woman is conscious?’”

As the old saw goes, extraordinary claims require extraordinary evidence. The sceptics countered that it was wrong to make these “radical inferences” when there could be a more straightforward interpretation. Daniel Greenberg, a psychologist at the University of California, Los Angeles, suggested that “the brain activity was unconsciously triggered by the last word of the instructions, which always referred to the item to be imagined.”

Put to the test

Parashkev Nachev, a neurologist now at University College London, says he objected to Owen’s 2006 paper not on grounds of implausibility or a flawed statistical analysis but because of “errors of inference”. Although a conscious brain, when imagining tennis, triggers a certain pattern of activation, it does not necessarily mean that the same pattern of activation signifies consciousness. The same brain area can be activated in many circumstances, Nachev says, with or without any conscious correlate. Moreover, he argues that Gillian was not really offered a true choice to think about playing tennis. Just as a lack of response could be because of an inability to respond or a decision not to cooperate, a direct response to a simple instruction could be a conscious decision or a reflex.

What is needed is less philosophising and more data, says Owen. A follow-up study published in 2010 by Owen, Laureys and colleagues tested 54 patients with a clinical diagnosis of being in a vegetative state or a minimally conscious state; five responded in the same way as Gillian. Four of them were supposedly in a vegetative state at admission. Owen, Schiff and Laureys have explored alternative explanations of what they observed and, for example, acknowledge that the brain areas they study when they interrogate patients can be activated in other ways. But the 2010 paper ruled out such automatic behaviours as an explanation, they say: the activations persist too long to signify anything other than intent. Owen is grateful to his critics. They spurred him on, for instance to develop a method for asking patients questions that only they would know how to answer. “You cannot communicate unconsciously – it is just not possible,” he says. “We have won that argument”. 

Since Owen’s 2006 Science paper, studies in Belgium, the UK, the US and Canada suggest that a significant proportion of patients who were classified as vegetative in recent years have been misdiagnosed – Owen estimates perhaps as many as 20%. Schiff, who weighs up the extent of misdiagnosis a different way, goes further. Based on recent studies, he says around 40% of patients thought to be vegetative are, when examined more closely, partly aware. Among this group of supposedly vegetative patients are those who are revealed by scanners to be able to communicate and should be diagnosed as locked-in, if they are fully conscious, or minimally conscious, if their abilities wax and wane.

In 2009, Laurey’s team asked one of the original group of 54 patients that he and Owen had studied – patient 23 – a series of yes-or-no questions. It was the usual drill: imagine playing tennis for yes, navigating the house for no. The Liege patient, who had been in a vegetative state for five years, was able to answer five of six questions about his earlier life – and all of those were correct. Had he been on holiday to a certain place prior to his injury? Was such-and-such his father’s name? It was an exciting moment, said Laureys. “We were stunned,” adds Owen, who helped independently score the tests. “By showing us that he was conscious and aware, patient 23 moved himself from the ‘do not resuscitate’ category to the ‘not allowed to die’ category. Did we save his life? No. He saved his own life.”

Nachev has not changed his view since he first criticised Owen’s work and has spelt out the basis of his unease in a more detailed paper published in 2010. “For every relative of a living PVS [persistent vegetative state] patient given (probably false) hope, another is burdened with the guilt of having acquiesced in the withdrawal of treatment from someone who – he has been led to believe – may have been more alive than it seemed,” he says. “There are moral costs to false positives as well as to false negatives.” 

“I find the whole media circus surrounding the issue rather distasteful,” he told me. “The relatives of these patients are distressed enough as it is.”

Laureys, Owen and Schiff spend a great deal of time with the families and understand these sensitivities only too well. Owen counters that, from his years of experience dealing with the families, they are grateful that doctors and scientists take an interest and are doing everything that they can. “These patients have been shortchanged over the years,” he insists.

Owen is adamant that doctors have a moral duty to provide a correct diagnosis, even if the results do cause guilt, unease or distress. “We must give every patient the best chance of an accurate diagnosis, so we can give them the appropriate care that goes along with that diagnosis.”

Mind limitations

The art of mind reading is constantly being refined. Owen and Lorina Naci have come up with a more reliable way to communicate with patients by getting them to focus their attention while in the scanner. First, a yes/no question is asked, and then a recording is played of the word “yes” repeated several times interspersed with distracting, random numbers, and a similar recording with “no”. The participant has to count how many of the correct answer they hear and ignore the incorrect answer. This mental effort (selective auditory attention) shows up distinctively when Naci and Owen examine the brain scans, so they can decode the responses correctly based on activity changes within the attention network of the brain. In follow-up studies using this method, Scott Routley showed he knew his own name, as distinct from another, and that he was in a hospital rather than elsewhere, indicating he possessed a higher level of self-awareness.

Yet there are many issues left to resolve. After the initial diagnosis, relatively little effort is made to systematically explore brain function in these patients, says Schiff. There are also minimally conscious patients who may not be able to imagine tennis and so on, when a few exceptional vegetative patients can. Other limitations are caused by the use of medication during trials or the huge diversity of the patients that are usually collapsed into groups (to spare doctors from carrying out the same procedures on the same patient again and again). And when it comes to younger patients, there is a limit to the number of PET scans they can have in a given period because a radioactive tracer has to be injected into the body.

Then there is the fact that huge, multimillion-dollar imaging machines – confining and magnetic – are unsuitable for patients whose bodies are affected by spasticity or have been rebuilt with screws, plates, pins and other metal. But more convenient alternatives are in development. Laureys is studying pupil dilation, which is linked with thought (the wider the pupil, the higher a patient’s emotional arousal, while more subtle dilations have been linked to mental functions such as decision making). Another method implants fine electrodes in the hand of a patient to measure “sub-threshold” muscle activity triggered when they are asked to move. 

Perhaps the most promising alternative is electroencephalography (EEG), which detects crackles of electrical activity in the brain through electrodes attached to the scalp. This is cheap, relatively portable and fast (with milliseconds of lag, compared with 8 seconds for fMRI), meaning that a research team can ask up to 200 questions in 30 minutes. This method can also cope with patients who twitch and move, or who have been reconstructed with implants. “This is a vulnerable patient population, and moving them is never easy,” says Owen, whose team have equipped a jeep. “We pack our gear in our ‘EEJeep’ and visit them instead.”

(Science Photo Library)

One of the most promising methods for measuring patients' brain activity is electroencephalography, or EEG (Science Photo Library)

Schiff’s team is sceptical whether one particular EEG methodology used with the detector really works. “One has to be careful of the dead salmon effect,” admits Laureys, referring to an apparently frivolous study of a deceased fish that made a serious point about the limitations of fMRI. The methodology struggled to distinguish real brain activity from background ‘noise’, suggesting that the dead Atlantic salmon that had been put in the scanner was actually thinking. “We don’t want to get excited about dead fish,” says Laureys, “but, on other hand, we do not want to be so conservative and demanding of statistics that we miss things.”

Illuminating the dark

Today it is normal to think of the transition between life and death as a question of “how the brain is” rather than “how the heart is.” A patient in a persistent vegetative state still has a functioning brain stem and can breathe unaided. They may possess some degree of consciousness and have a slim chance of recovering. By comparison, a PET scan of a brain-dead person reveals a black void within the skull, a barren neural landscape with no chance of sparking back into activity again: their body cannot survive without artificial help.

Schiff believes that a combination of devices, drugs and cell therapies, laying the foundations for a new generation of diagnostics and treatments, will illuminate the penumbra between conscious and unconscious. “We’re not quite there yet,” he stresses. Much of the work to date has demonstrated the value of brain scans on populations of patients but, ultimately, they need dependable methods that will work on a patient-by-patient basis. “We are going to have to do some amazing small-scale studies to show what is possible in one or two subjects before everyone gets simple things done that can help them today,” Schiff says. Eventually, he believes there will be a “cultural shift”. Laureys thinks we may need to start with the language used to describe these patients – he wants to replace the loaded term “vegetative” with the neutral “unresponsive wakefulness”.

Despite the scepticism, the difficulties in dealing with such diverse groups of patients, and the challenges of standardising diagnosis, the research is moving forward. It is already making a difference, enabling a few patients to tell their doctors whether they need pain relief. 

Back on Skype, Owen smiles, considering whether to tell me what he is planning next. Owen’s partner, Jessica Grahn, also a neuroscientist, became pregnant at the start of 2013. What happens when consciousness winks on in the developing brain?

He emails me a video of their unborn child, a montage of fMRI slices through their baby’s head, as it twists and turns in Jessica’s womb. “My colleagues have been doing fMRI on my wife’s tummy every week for a few weeks now to see if we can activate the fetus’s brain,” he writes. “It is AMAZING.”  

Scott Routley died in September 2013 with his family by his side.
Adrian Owen’s friend Anne remains in a vegetative state.
Adrian Owen and Jessica Grahn’s baby boy, Jackson, was born on 9 October 2013.

This is an edited version of an article originally published by Mosaic, and is reproduced under a Creative Commons licence. For more Mosaic articles click here. If you would like to comment on this, or anything else you have seen on Future, head over to our Facebook or Google+ page, or message us on Twitter.