A recent New York Times article discusses the interesting phenomenon that some people in a persistent vegetative state (PVS) or minimally conscious state will have a temporary improvement in consciousness after being treated with the sleep-aid drug zolpidem (brand name Ambien). The journalist, Jeneen Interlandi, does a fair job, although the scientific discussion is shrouded in personal emotional stories   – the kind of tales that make for good storytelling but I think may give the reader the wrong overall impression of the science.

I have written here often about coma, and mentioned the zolpidem issue before, but an update is in order. Zolpidem is a sedative given to help with sleep onset (there is also an extended release version that can help with sleep maintenance). In 1999 it was first observed that a person in a coma appeared to respond to zolpidem with increased consciousness.  Since that time there have been a few further case reports and small case series, and even a small placebo-controlled trial. The overall result of these cases is that a small percentage of comatose patients appear to have a paradoxical response to the sedative in that it increases their brain activity. In one study only one patient in 15 responded. In another study of three children, none responded.

Is such a response plausible? Zolpidem is a GABA agonist, which means it activates the neurotransmitter GABA (gamma-aminobutyric acid). GABA is the primary inhibitory neurotransmitter in the brain, which means it decreases or inhibits neuronal firing. For this reason it is commonly use to treat seizures or abnormal nerve pain, by decreasing the activity of hyperactive or hyperexcitable neurons. How, then, could it increase brain activity?

Well – if you inhibit a part of the brain that is itself inhibitory, then the net result will be activation (two negatives make a positive). It is therefore possible that in some patients in a coma one part of the brain that is still functioning is having a net inhibitory effect on other parts of the brain that might be contributing to consciousness. The introduction of zolpidem then disinhibits brain function resulting in a net increase in brain activity.

The question is – do the minority of coma patients who respond to zolpidem respond because of their brain anatomy or because of the nature of their injuries or a combination of both? Some people who use zolpidem for sleep have paradoxical parasomnias – they may sleepwalk or engage in bizarre activity during sleep. This could be evidence of paradoxical brain activation, and may be the result of their brain anatomy or chemistry. So perhaps the coma patients who respond to zolpiden are the same people who would sleepwalk if given zolpidem as a sedative to help them sleep. This is currently an unanswered question.

Or – perhaps those coma patients who respond have no relationship to pre-morbid sleep response. Perhaps it has everything to do with the pattern of their brain injuries. You can imagine different parts of the injured brain working to various degrees, with the net result being a somewhat random assortment of brain activity, with a great deal of variation from coma patient to coma patient. There appears to be more variability in those who are comatose following trauma than diffuse lack of oxygen (anoxic-ischemic injury).  This makes sense. A lack of oxygen would affect the whole brain. Trauma would have much more variable effects, perhaps destroying some parts of the brain and leaving others uninjured. Patients in a coma post trauma generally have a better prognosis than those in a coma from lack of oxygen, probably because in the latter case the damage is simply too widespread.

In any case – perhaps those who respond to zolpidem with increased consciousness just happened to have a pattern of partial brain function that includes an inhibitory network that is dragging down overall consciousness, and this inhibitory effect is temporarily inhibited by the zolpiden.

A third possibility is that there is simply some part of the normal brain, either in most people or just in those who naturally respond to zolpidem with activation, that the zolpidem is acting upon in order to produce the paradoxical alerting effect. If this part of the brain is sufficiently undamaged then the coma patient will respond to zolpidem. If it is damaged beyond a certain point, then there will be no response.

As the NYT article discusses, researchers are planning studies to tease out the above questions. What is going on in the brains of those who respond to zolpidem and how is it different from those who do not respond?

The answers to these questions could lead to new and effective treatments for some comatose patients. The idea is essentially to coax as much activity out of the injured brain as possible by stimulating those parts of the brain that are injured but still have some function, or by inhibiting those parts of the brain that have a net inhibitory effect on consciousness. In this way consciousness can be dialed up a notch or two, which can have a clinically significant effect on some patients in a coma.

Using drugs like zolpidem is only one approach. Researchers are also looking into implantable electrodes that can stimulate key brain regions, such as those that are normally activating and necessary for the maintenance of the waking state.

Parallel to these efforts is research simply looking into how to tell which patients in a coma have the potential to recover somewhat or to benefit from such interventions, and which ones are irretrievably gone. These efforts are currently using fMRI and digital EEG to assess brain activity and correlate the activity to careful clinical exam. They have not yet, however, connected any differences to a meaningful difference in outcome.

That is a point that needs to be emphasized. So far we are talking about minimal improvement in overall function. Even the dramatic stories of patients “waking up” generally involve patients being able to look at someone and follow simply commands, like wiggling a finger. Perhaps they can even speak a little. But patients, after several months or more of being comatose, generally do not have significant neurological recovery to the point that they can lead semi-normal lives. They always remain (perhaps with exceedingly rare exceptions) neurologically devastated.

The ultimate question, therefore, is  – what are the limits of the approaches that I described above, coaxing some extra activity out of the remaining brain tissue? Even if we can achieve this goal optimally, we are still working within the limits of the damaged brain. The most dramatic recovery in comatose patients appears to be in those whose brains healed to some degree – new neurons were recruited and able to make new meaningful connections that restored lost function. There are severe limits to this as well, but it can lead to clinically significant improvement also, after months and in rare cases even after years of coma. (All with the came caveat as above – they remain neurologically devastated.)

Perhaps in order to make the kind of recovery that families hope for – a return to the ability to participate meaningfully in life – we need to do more. We will likely need to find technologies that actually repair or replace damaged brain tissue. There are two approaches that I can think of that have the potential to do this. The first is neuronal stem cells – injecting stem cells into the brain that will then turn into mature neurons and be recruited to make meaningful connections. This approach can increase the raw brain power of the patient, not just “overclock” the remaining brain tissue.

The second approach is using brain-machine interface technology to essentially implant computer chips in the brains of injured patients that will take over for and enhance brain function.

These two technologies do not currently exist, and it is difficult to predict how long they will take to develop. In my opinion using drugs like zolpidem and electrical stimulation to enhance brain function in comatose patients will produce, at best, modest improvements in neurological function. This may bring some patients across functional thresholds that they and their families will find meaningful, but it seems the potential for significant neurological recovery is inherently limited. In order to achieve what we really want – a return to near-normal function – we will need new technology like stem cells or computer enhancements.