BioEd Online

Brain imaging has allowed a man who was previously considered unconscious to answer a series of yes-or-no questions. The study, published this week in the New England Journal of Medicine¹, challenges clinicians' definition of consciousness and provides an unprecedented opportunity to communicate with those who show no outward signs of awareness.

Patients are classified as unconscious, or being in a 'vegetative state', if they are unable to respond in any fashion to an extensive series of questions and requests. But if the patient is completely unable to move, an aware or communicative mind could go unrecognized by this method of assessment. "The conundrum with the vegetative state is that it's a diagnosis made on lack of evidence," says neuroscientist Adrian Owen of the Medical Research Council's Cognition and Brian Sciences Unit in Cambridge, UK.

Owen and his colleagues made headlines three years ago when they used a brain-scanning technique called functional magnetic resonance imaging (fMRI) to demonstrate that a woman in a vegetative state could respond to verbal commands². When researchers directed her to imagine playing a game of tennis or walking through her house, the fMRI scan revealed that she activated the same areas of her brain as healthy subjects who were asked to do the same.

The findings sparked a debate over whether or not the woman was actually conscious (see 'Thoughts of woman in 'waking coma' revealed'). Some argued that what Owen and his collaborators saw was nothing more than an automatic activation of those brain regions on hearing certain words.

One such critic was Lionel Naccache, a neuroscientist at the Pitié-Salpêtrière hospital in Paris, who urged caution before concluding that the woman was indeed conscious.

This latest evidence of communication with a vegetative patient, however, has Naccache convinced. "This is clear-cut evidence of consciousness," he says.

Piercing the fog

Owen and his collaborators repeated their fMRI experiments on 54 patients who had previously been classified as either vegetative or 'minimally conscious' (a condition in which a patient may inconsistently respond to commands, but cannot communicate interactively). They found that five of these 54 patients responded to commands to imagine playing tennis or navigating through a familiar house. Four of those five patients had been classified as vegetative, but when clinicians repeated their assessment following the fMRI study, they found evidence that two of those four should instead have been classified as minimally conscious.

One of the responsive patients, a 22-year-old man who had been diagnosed as vegetative for five years after a traffic accident left him with a traumatic brain injury, was selected for further study.

Because it is difficult — if not impossible — to determine whether someone is thinking 'yes' or 'no', the researchers instead asked the patient to imagine playing tennis when the answer to a question was yes, and to imagine walking through a house when the answer was no. Visualizing these two activities stimulates different parts of the brain that are easily distinguished using fMRI.

They asked the patient a series of simple yes-or-no questions pertaining to his personal history, such as "Is your father's name Alexander?". He answered five out of six questions correctly. No brain activity was observed in response to the sixth question.

The results suggest that fMRI could be useful in diagnosing unresponsive patients, says Owen. "There are things that are just not going to manifest themselves in outward behaviour," says Owen. "This method can tell us which patients are aware, and it can tell us what they are capable of."

For Naccache, it is the patient's ability to respond using the code suggested by Owen and his team that indicates he is truly conscious. "When you are conscious, you have the ability to use an arbitrary code to communicate with somebody," he says.

Parashkev Nachev, a neuroscientist at University College London who was not affiliated with the work cautions that it is important not to over-interpret the results. The patient only answered a series of very basic questions, he notes, and the results in his view do not necessarily suggest that the patient is fully conscious or has the potential for recovery. "There is no doubt that it does merit further research," he says, "but I could not see using it as a clinical tool at this stage".

As a next step, Owen and his colleagues intend to ask the patient a series of questions with unverifiable answers. For example, the technique could be used to ask whether they are experiencing any pain — a question that frequently troubles family members and hospital staff. But should a vegetative patient be asked whether they want to live or die?

"I think there's an enormous problem with that," says Owen. "Just because a patient is able to respond with 'yes' or 'no' doesn't tell you if they have the necessary level of competence to answer difficult, ethically challenging questions about their destiny."