 Good afternoon, everybody. It's my pleasure to introduce Nico Schiff, who is, in addition to being an amazingly pioneering neurologist and neuroscientist, is a family friend, and we were just debating between the two of us and with Nico's mother, Louise, who's sitting right here whether Nico and I met at zero, one or two. We don't know, but one of those. And it's just been so much fun for Nico and I to come into neuroscience together, and I have been blown away for over 10 years by Nico's really imaginative and revolutionary studies on the consciousness within that does not see the light of day or see the light of movement. And what he's going to tell us about today are some really innovative studies assessing who's at home in both minimally conscious and persistent vegetative states, people with altered states of consciousness. I should give a little bit of background. This is sponsored by the McLean Center for Clinical Medical Ethics and the Grossman Institute for Neuroscience, Partuitative Biology and Human Behavior. That's a mouthful. This is our third lecture in our series on ethical issues in neuroethics. The series was organized by myself, Mark Siegler, Dan Somasi, and John Mansell. And Nico has achieved great heights. He is the Gerald B. Katz Professor of Neurology and Neuroscience at Cornell. He's also a professor of neuroscience and he's also a professor of public health, which I think is very interesting. And I hope that we talk a little bit about the public health implications of this work. His research bridges the pathophysiology of impaired consciousness, the mechanisms for arousal regulation, and most recently, what can we do about it? Well, it used to be that we couldn't do anything, but Nico's work has shown us that we can use deep brain electrical stimulation to bring patients back into a different state of consciousness. So please join me in giving a warm welcome to Professor Shepp. Thank you, Peggy. It's a pleasure to come back to Chicago. I grew up here, first time ever spoken here, since I was in high school as a lab student for those lab grads in the audience. So because this is an ethics talk, and I generally give my usual science stump speech, I've tried to frame the talk around what I think are the most interesting things that, as ethicists and people thinking about the changing landscape of how policy, medical practice, and the law, which is really all the work that my close colleague Joe Finns, who spoke with the neuroscience meeting on this just on Monday, really has focused on professionally, I want to sort of give you the backstory. I want to give you the backstory as best I can, and I'll try to translate this in language that's not overly expecting you to have neuroscience, neuroanatomy, and neurophysiology at your fingertips. So I'm going to frame the talk around two related but distinct problems, because really there are two main problems, and the talk is going to focus on these problems. I'm going to make a little contact with the brain stimulation work, but what's changing as a result of clinical and basic neuroscience is our understanding of how widely damaged brain networks can either gradually or abruptly become capable of supporting conscious behaviors of varying levels of complexity. So I think if there's a theme to the entire talk, and what I'd really like to convey or teach people here, is that the scope of this work, the scope of this problem space is much broader, much richer in contextual detail than you will pick up from any popular account of it, and there really are two problems. One is actually a bigger problem than the other, and it's the less well-recognized problem, and the one that I think has captured most of the public imagination is a very important problem, and it has a higher urgency as a problem, but it's a smaller problem, and that's the identification of the retention of high-level integrative cerebral activity in patients who have no behavioral evidence of this capacity. That can mean that they have no behavior, or it can mean that their behaviors are very, very limited, yet we can make measurements to show that they have higher-level behaviors that are not in our view. So that's the problem. I think most people have become aware of this field as a result of neuroimaging and pictures of patients who were described as not being able to do very much or anything at the bedside, showing pictures of brain activity that demonstrates evidence that they're able to do things like imagine walking around their home or playing tennis or doing other types of activity. And then as I go through the talk, and I'm going to talk about both of these in some detail, the goal is for me to expose and address what I would call, following on Aristotle, the difficult metal of novel measurements, because as we advance biological understanding of the recovering brain to the measurements we're making, it's clear that there are a lot of ambiguities and interpretation, and as some of these results make certain points unambiguous, they also lead often to ambiguities that are very challenging right now to deal with professionally, either as clinicians or as scientists. So let's just start by some definitions. There's an evolution in the last 10 years or so from the old terminology, which was pretty much restricted to coma, persistent vegetative state, which was a term that had a time component to it. Being the vegetative state for 30 days and then after 30 days you were in the persistent vegetative state. And really there weren't any other terms except for death. Severe disability functionally approximating persistent vegetative state more or less was treated as if it was persistent vegetative state. So there was no category for patients who clearly didn't fulfill vegetative criteria. Vegetative patients only can open their eyes and close them. They don't show any response to their surrounding environment. I'll go through the details of all this in a minute. But patients who showed a little bit of response were more or less lumped into the category of persistent vegetative state, even though it was understood that they lived somewhere on the lower boundary of severe disability. Over the last 10, 15 years what's happened is the category of severe disability close to vegetative state has become parcelated into different components. I'll go over those in a moment. And the best way to really organize this is to put it on a two-dimensional grid. So if we have cognitive function here at the top, if we follow the x-axis from no cognitive function behind this interrupted blue line to normal cognitive function here at the extreme of the x-axis and then look at motor function from no motor function below the interrupted orange line, we see that there are a range of conditions where at the bedside, things are indistinguishable. A person could be eyes closed or have eyes open but not move their eyes to command or track. And we wouldn't necessarily know by behavior alone whether they were comatose. If their eyes were closed, we would call them comatose. If they had periods of eye opening and eye closure but didn't have any response to the environment, we would call them vegetative state. In the next category over, minimally conscious state, we identify by having people be behaviorally within this box between the interrupted yellow and the interrupted orange line. They have inconsistent motor function. For patients who only can track an object or localize to an auditory input, we now call them MCS minus but that gives them the first evidence that there's some awareness and registration of sensory input and consciousness. And then if patients can follow a command but cannot use their command following system to consistently communicate accurately and operationally what this means is answer six out of six questions like am I pointing to the ceiling with a yes no response or am I pointing to the floor and be accurate. That's the boundary condition that's been set sort of arbitrarily to get out of minimally conscious state. And once you get there, then you pop up above the yellow line but below the orange line, below the yellow line, there are equivalences at least at the level of behavioral assessment. There are not necessarily equivalences at the level of assessing brain function and that's where the story gets more interesting. So as I said, there are really two problems. So the first problem scientifically is to understand how patients move in general. Anybody in coma has an ambiguous trajectory. They could end up with full recovery. They could settle out somewhere in between. What is the biology of this process? That's been the focus of my lab's work over the last 20 years and that's really what I think frames all of the issues that we have in front of us. The related problem is when we see people who look like they might have some responsiveness, we have the comparison that there are patients who have very limited motor channels who if they can control them well enough to establish a communication system can demonstrate normal consciousness and those are patients in the locked-in state. A locked-in state is not a disorder of consciousness. There's a related condition, complete locked-in state, which is problematic. So most complete locked-in state patients have been identified in amyotrophic lateral sclerosis from patients who initially could use a brain-computer interface but then as the disease progressed and the motor-bader rhythm that was used typically to trigger the brain-computer interface no longer is present sufficiently to drive the technology, they slip into a state where they don't have any response. Now, being, you know, economical in our thinking and following Morgan's Canon or Occam's Razor, it would be silly to think that probably that they've just lost consciousness now just because the motor system that was our one toggle to a switch to connect them to the world is no longer working. But this has been remained sort of an ambiguous question. There are some studies now using near-infrared techniques that suggest that that capacity is, of course, still there. We just don't know and in that particular disease process, other parts of the brain undergo degenerative changes so there could be aspects of alteration of consciousness that make it hard to control the brain-computer interface as well. But it gives you a sense of the ambiguities that may arise and I'll pick up on that at the end of the talk. That's our second point. And of course, when we are looking at patients, we're not looking at people who have ALS or people who have focal brainstem strokes and have fully intact cerebral function outside of the motor system. We're looking at people with mixed multifocal injuries. So they have both often motor dysfunction plus central nervous system damage. So then they have the combination of issues of state control and perhaps very, very limited motor channels out. And that makes it even more complicated because then we have people who are legitimately motorically impaired, income of vegetative state, minimally conscious state, who because they do not show us through their motor function start to recover and may end up over here or anywhere in between. So what we need are measurements that help us understand the graduated recovery process, what we measure, what we can correlate with people who can move that show us the transitions in the minimally conscious state patients who have behaviors and in the patients just over what's called the confusional state who can speak and interact, typically can communicate accurately by the minimally conscious criteria but are not oriented to their personal history or why they've been injured and operationally cannot be tested on standard psychometric tests. And then there's a range of patients who can be tested on standard psychometric tests but have impaired cognitive function before recovering to a level where they're within a normative range. And that sort of sets the full boundary conditions from unconsciousness and coma to normal functional recovery. And this is the game that we're trying to play but it's not just restricted to structural brain injuries evolving slowly over time. I think if you think about it, it's related to anybody who has an acute or transient process that's impaired consciousness in the hospital who has an underlying stroke, has other potentially reversible causes of brain dysfunction and has a limited cognitive function that in principle may be reversible and has to be considered from a mechanistic point of view. I can tell you that that is not the standard that's applied now because there isn't a frame for most people to become educated and to think about restoration of brain function in the neuro ICU and the surgical ICU and the medical ICU but that's where this is going. And all these issues are going to trace back not just there but also to how you treat people who have early stages of dementia, early aging processes who get acute disturbances and alter their brain function. So it's one kind of problem space to understand. So but in this problem space we're mostly looking at patients who are young who've had traumatic brain injury and there's some things that I think are not well known but are very important to understand sort of just demographically. And one is that even though you can be in the minimally conscious state for a month, it's been known for quite some time that if you follow patients prospectively, at least the patients who are minimally conscious and have had a traumatic brain injury, often most of those patients do well. In fact, 10 percent of them can have no disability at one year even though that recovery process is slow. That's not true for patients who have non-traumatic brain injury and it's not true for patients who are remaining in vegetative state at one month but in general if you've had a traumatic brain injury or recovery process over the first year is pretty good if you're in minimally conscious state. On average most of those people actually do much better. If you've remained in minimally conscious state for a full year after a severe structural brain injury, you typically don't recover to better than severe disability outcomes but most patients will recover the ability to communicate, interact with their environment. Many of them can go home with assistance and retain executive function. And then of course there are the other things that come up in the news that draw people's attention. Those are the very, very late recoveries that are reported of comatose or vegetative patients but invariably are in patients who have met the criteria for minimally conscious state early on in their convalescent period. Nonetheless, this man Terry Wallace who's a national figure and this man Don Herbert who's another nationally reported patient both made very late recoveries from minimally conscious state. Wallace remained in a minimally conscious state for 19 years and started to speak for the first time in over a three-day period, recovered full fluent language. We had the chance to study him twice in New York and have written a few papers and have more coming on the changes in his brain function but he actually did make this transition and then once he became conscious started to gain additional functions including motor functions and cognitive functions and they had correlates in his central nervous system both structural and functional correlates. Don Herbert was a fireman in New York and he was injured in a house fire. He was where the floor collapsed on him. He had a mixed traumatic and anoxic brain injury and his story is very disturbing and not atypical. He started to get better. Within the first few months he was ambulatory. He was speaking a little bit. He was clearly minimally conscious and then he just slipped back into a state where he was basically MCS minus. He was labeled as vegetative but if you examined him carefully there were signs that he was in minimally conscious state and he stayed that way for nine years in a nursing facility until one day a guy on rounds came around and was looking at the patients. He said, well gee, you know, why don't we try something? So he put him on a cocktail of L-dopa, Selexa, and Stratera. So noradinergic, serotonergic. We have take inhibitor and a dopaminergic agonist. And one month later Herbert woke up and Herbert really woke up. He woke up. You can read there's a book about him. I actually, we tried to study him and it's a long story I won't get into now but we didn't actually bring him to Cornell even though his family was ready to do it. And he spent four hours speaking the very first day and it was all videoed. And his recall, his insight, he was heartbroken that he'd left his family for 10 years. And it was, I mean it was eerie. So the Harvard capacity was there. So that's really the question we're asking. How can this happen? So one thing we do know is that once you get to minimally conscious state it's going to be about the brain and the potential for late recovery may be indefinite. Now these are very obscure late cases but the rules of the game are going to figure that out and I think the rules are going to become more apparent. The population story is becoming a little bit more obvious now too in a study that was published in 2012. 400 patients, all of whom had a very severe brain injury. The lowest level of Glasgow Coma scale scores at their onset of Coma. What they found was virtually 70% regain consciousness and of those patients, if you track them out 21% actually regained independent function and most of those were employable. The patients were employable by standard the disability rating scale criteria. But also if they followed this cohort at one, two and five year time points and they continued to improve at each point. So this is the general story. The very severely injured brain as long as somebody is healthy and doesn't have another intracurrent process slowly starts to get better. So how does that work? Well, that's what we're trying to understand and we spent a lot of time looking in the minimally conscious state because minimally conscious state can show some very significant transitions which can allow us to gain a little bit of insight and so I want to share a bit of that with you right now. So here's an example of a patient who had a very severe traumatic brain injury in a car accident about three and a half years before the images that you're going to see now. This is a man with multifocal diffuse external injury more left-sided damage than right and he's being tested using a standard psychometric assessment for minimally conscious state. He's going to be given two objects, a comb to sort of demonstrate how he brushes his hair and a spoon to see if they can show how to use a spoon. If he could use both objects, demonstrate their use effectively and switch them around then that would be another criteria for being out of minimally conscious state. So here's what you see. So he holds the comb, he takes it. This is actually his dominant arm. It has a resting tremor like a Parkinson's tremor. He can use that arm spontaneously. It was his dominant arm, but he doesn't use it when the therapist hands him the objects. He's looking down, his posture is stooped over and so this is sort of the second to highest level on the scale of motor function within the range of minimally conscious state that's tested. Now, in this particular subject, it was discovered empirically that they paradoxically improved with the use of a sleeping pill Ambien, which has got the generic name Zolpidem. And if you... Very few patients have this response, less than 5%, but the ones who have it, it's very dramatic and it was very reliable and we were able to study this patient over a week doing test retest and what you're going to see now is the same tests, but a half an hour after receiving Zolpidem on the same day that the first film was shown. And before I do that, just point out, you can immediately see that his posture has changed and I've blotted out his eyes for identification, identification, but he's looking as she comes. So for the neurologist in the audience, he's using his dominant arm, right? Using his dominant arm and the latency of the movement, the accuracy of it, and even, you know, when you think he's kind of losing the task set, he's just deciding on his pantomime for the use of the spoon. So, rather remarkable. He also speaks, writes, it can verse fluently, but he's in a confusional state. He's only oriented to his name, the people in his life, he's not oriented to the year or what's happened to him, and he comes in and out like the smile of a Chesser cat. And for the ethicists in the audience, a fine point. His family was forced to only give him the medication in the mornings on Sundays when they visited because if they left him on the medications too much of the time, he would be awake at night and there was no infrastructure or support in the facility he was in to make sure he was safe. So, this is, and this is, and so when he came to us, we wanted to be able to give him multiple doses. We hired a one-to-one nurse with him every night he was in our study to make sure he didn't climb out of bed. So, how do we try to understand and get into this? So, we use a variety of techniques. We use neuroimaging techniques, MRI methods. We also use positron emission tomography, but we spend a lot of time looking at EEG because EEG is a very flexible tool and allows us to look 24-7 at brain state and also introduce various types of perturbations and compare them over time in a way that the other imaging studies just aren't as flexible to do. But to look at EEG data, if you're not familiar with it, it's a little unfamiliar. So, I'm just going to teach you in 20 seconds how to read an EEG spectrum. It's really not that bad. So, this is a classic figure from a 1954 textbook from Wilder-Pentfield and Herbert Jasper on the different brain rhythms associated with wake and sleep. Now, this is a typical channel over the front part of the brain of EEG. You see that the electrical activity of the brain is very fast and it's kind of low amplitude. They call it excited. We would call this sort of a beta rhythm. Relaxed here is actually a channel taken from the back of the brain with a sort of a spindle-like activity at 10 hertz. This is a kind of a little alpha rhythm, usually elicited when the eyes are closed over the back of the head. As the day wears on, you start to become drowsy. There's a slowing of the rhythm and an irregularity of the rhythm so that they become slower than 10 hertz, it's about 7 hertz, and that's a theta rhythm and that's associated with drowsiness. Sleep itself has its own characteristic rhythms. The second stage of sleep has very fast rhythms that are co-mixed with slower rhythms and things like this, which are called K-complexes. Deep sleep is typically filled with very slow rhythms that are irregular and here's a trace from an injured patient showing that in coma, very slow rhythms, kind of similar but different in comparison to deep sleep are typical. These are just the time traces of the electrical activity. If you put an electrode on the head and you just let it run out. The way we approach this quantitatively for a analysis and spectral analysis, you don't need to work through the equations. Think of it this way. This is a statistic on the time-varying voltage here that tells you where the dominant power is in terms of sine waves at particular frequencies. If you have a signal that's almost entirely one frequency like this 10 hertz alphabet even though the amplitude of it varies, it's going to peak at a single frequency on this summary statistic called the power spectrum and have a little bit of width because there's some waxing and waning of the power and maybe some slowing of the frequency around this 10 hertz vocal frequency. If you have a clearly slower and somewhat more mixed set of frequencies you'll have a broader peak. This is the 7 hertz of the theta here and for these slower rhythms you'll have even slower parts of the power spectrum having the dominant power selected. For other rhythms you may have two kinds of rhythms. For this rhythm, although you can't see it, there's an alpha rhythm in it and you can sort of see that it peaks up occasionally but you won't see it by eye but you will see it if you do the power spectrum. So you'll see the alpha peak it's not as prominent but it's there and then there's a peak for these higher frequencies. This would be a normal wakeful spectrum. For sleep, if you have these two 3 hertz fast rhythms show up here and then the slow pieces would also have a component of the power spectrum. So that's how you read a power spectrum and now I'm going to show you some power spectra. That's the quick glance of them but it's I think transparent. So what do we see when we look at the patient I just showed you? So here is what their resting traces of their brain ways look like in the state where they're holding the spoon with the left hand and not moving. What you see is there's some very slow stuff. These sort of almost look like sine waves running through the record and they're pretty diffuse. They're almost over all of the channels of EEG. And then when they get the drug you see a big change. They're faster kind of like that first tracing I showed you in the model slide but what we don't see are those slow rhythms. Now when we convert those two pieces of EEG actually multiple copies of them about 100 swatches of equal length combined in average into power spectra over each channel of the EEG we see the overall story summarized. We see that in the pre-drug state there's a very strong low frequency oscillation at 7 hertz that's present over all the channels more dominantly in the front. The patient is given the medication within the first hour you see that those rhythms are almost replaced entirely on these channels by high frequency activity into roughly the 20 hertz range. And in fact if we take the patient I just showed you and compare this patient's profile within channels and across the head to two others open and responsive patients who were in the paper in eLife if anyone's interested to read it who had completely different types of brain injury. We see that these findings are generalizable that they're almost generic and this is the kind of inference that we've been drawing that there are consistencies at the cellular and circuit level for how dysfunctional brain activity looks after multifocal injury and what happens when you restore it I'm simplifying and glossing over quite a bit but let me do one more round of doing the same thing and pack it a little bit. Here's a look sort of overview of behavioral testing of this patient that we just saw through multiple rounds of being given the drug zolpidum. We saw the motor scale the motor scale is a six point scale the first images you saw were a five on that scale the second images were a six you can see he lived between a five and a six on that component of the scale but this scale the communication scale is a yes no system and he didn't have a yes no system at baseline and it took a little bit longer for him to have an accurate communication system so for example in the first dose that we gave him in this study he showed yes no responses but he didn't meet the bar of a six out of six point to the ceiling point down it took him the second dose but after the second dose and the next two days each time we gave him the medication we we saw him reach we saw him reach that level where he was off the off of this metric entirely it doesn't mean that he couldn't speak and he couldn't tell you his name and other things it meant that operationally when you asked him six questions in a row he didn't get all six right so you see these are they're sort of fine gradations but they're the kinds of things that allow us then over time to compare these hours of behavior to those hours of behavior and this is the kind of thing that you see so here's a here's an example of what's going on in the first four that basically across this dose where he starts at a communication level of zero and starts to communicate right away after the dose but then picks up actually almost everything picks up within the first half hour here this is one channel of his EEG you see the low frequency rhythm the drug comes on board and he gets this burst of high frequency activity which you see even in normal subjects who take the drug to go to sleep this is what would happen in your own EEG except that it would go back to the baseline and your baseline would be normal within 20 minutes what happens in this patient is after this burst of activity he then changes state completely from this low frequency rhythm into a rhythm where two types of beta appear they stabilize within the first hour and then they carry on during the time that he's behaviorally most re-engaged so this is the kind of information that gives us a sense of this looks very much like what happens when people come up out of anesthesia except that something like this happens within about five minutes in most instances not over three hours and it's a window onto this process the other measures we have this is taken from the second patient in the study who had very similar findings in the EEG is metabolism and in this patient we found that cerebral metabolic rate went up a factor of two fold effectively across the entire brain on and off the drug so you can have a very injured brain with cells that aren't dead functioning at a very low rate marked by dysfunctional electrical activity but yet harboring a capacity to reintegrate virtually every brain network to restore motor function speech function and other other levels of capacity and this is what we really want to understand in greater detail so in thinking about how how this kind of phenomena might arise through lots of different things which I can't go into we've been led to a proposal that there is a common circuit mechanism underlying recovery of brain function after severe brain injury that involves the circuit connections between the central thalamus the frontal cortex components of the striatum the basal ganglia particularly the output neurons of the caudic and the medium spine neurons the inhibitory connections of the sternum and their inhibitory influences on the central thalamus and the general models is something like this if you have enough multifocal brain injury to cause coma the first principal component the first problem you run into is that your brain now has lost a lot of cortical-cortical and thalamocortical connections and it's much quieter and if you have a much quieter brain and because of certain properties of this cell type and because of the connectivity properties of the central thalamus the central thalamus will stop firing at the same rate these cells will likely shut down almost entirely and the whole frontal systems will disproportionately shut down compared to the rest of the brain and we've been testing various aspects of this hypothesis and finding verifications of some of the important detailed predictions of it it's an inference drawn from basic known physiologic properties and one assumption which is that very severe multifocal injuries cause neuronal death and disconnection across large territories in most patients and we think that's probably not much of an assumption and what it also does which is nice is it helps us organize several fairly disparate observations about things that help patients recover like introducing dopamine agents the zilpidin effect which probably works with the central phalamus which is one of the most powerful points including the stritum in the frontal cortex and organizes the approach that we've taken in minimally conscious state to trying to restore function which is using electrical stimulation of the central thalamus and this works through the glutamatergic afference coming out of the central thalamus so I don't have time to kind of tell this story because it's a separate lecture Globus Paldus and Terna to give you a couple pieces of data that give it face validity if you look at the fusexonal injury segregated by outcomes from vegetative state minimally conscious state to moderate disability loss of metabolism gets worse in these areas of the brain as you move down the level of functional recovery conversely in some studies where drug effects have been shown to bring people out of minimally conscious state recovery of function in the same regions of the brain is associated with drug response to some of the drugs I had on the other slide this is to amantidine, there's another study mesulpidum now this is a study of mesulpidum that was done by yet another group, there are a couple studies now showing that the frontal part of the brain can be restored and the thalamus with the use of mesulpidum similar to the patients I showed you and finally because I'm going over this quickly and I'm sorry to rush through this part there are some other predictions that should be consistent with this idea that have nothing to do with brain injury and really have to do with the general circuit mechanisms that are laid out in terms of the physiologic properties of the neurons and that is that if you have other contexts in which there are natural reductions in global background synaptic activity you should see similar stuff so if you look carefully at human sleep studies that were done at the NIH for regional blood flow measurements it actually turns out although the authors didn't explain it they identified it that maximal increases in blood flow in the brain actually occurred in the striatum in the transitions from slow way sleep to REM or from sleep to wake and that's consistent with the MSN compartment being shut down in these more inactive states and then having a state transition going forward as gullimatergic affrontation reemerges something similar happens in the first 20 minutes after sleep in the period called sleep inertia I'll show you that in a moment and general anesthesia has stages that in intact normal brains that are quieted by an or an injectable anesthetic show very strong similarity to this so briefly in the first 20 minutes of awakening blood flow in your brain hasn't normalized and this is the period of time when you're stumbling over your words bumping into things have psychomotor retardation on standard psychometric tests and when that resolves, blood flow has been restored in the thalamus, medial frontal regions and the rest of the brain and this is a general finding in healthy volunteers and then in anesthesia there's a lawful progression of changes that happen as subjects go under to general anesthesia and probably most related to what I showed you in the video is this phenomenon called paradoxical excitation which happens after a period of general quieting and relaxation and sedation a period of hyperactivity occurs which we proposed more or less mirrors the release of activity through this anti or forebrain meso circuit there's much more to say about this this wasn't the focus of this but I want to give you an understanding that there is a way in to think about how you would study this, make measurements in pet and EEG and try to falsify the hypothesis driven set of predictions so now I'm going to turn and then end of my talk to the problem of retention of high level capacity in patients who don't show it at the bedside because this is increasingly important and it's going to be a small percentage of the patients but there's never going to be I think an area of more urgency to treat in medicine than what we've discovered here as a field and Joe Finns just wrote a book on the rights of patients in this condition and others and everybody who was faced with this problem with these patients and I'm going to try to make the case to you through what this looks like, these patients are underserved and this is a big deal and now I think what's different from say 10 years ago is 10 years ago we had some examples now we're getting to the population data slowly, it's not prospective true demographics, epidemiology data it's just enough to say it's just big enough that we got to really rethink it however there are difficulties in interpretation and you know we're not going to solve it very quickly so the first question the first issue is what really should we call somebody who shows us that they can do something like imagine playing tennis or walking around their house with an EEG study or an fMRI but who at the bedside shows no behavior or very little behavior so there have been proposals to call patients minimally conscious with a star there have been proposals we've actually in the past considered calling them non-behavioral MCS proposals to call them functionally locked in and there are problems with all of those and they can be summarized basically in one sentence they draw a conclusion about the state of consciousness that you can't demonstrate so what we really need are operational definitions so as of Monday I published this editorial to offer what I think is an operational definition in the label Cognitive Motor Dissociation and Cognitive Motor Dissociation would be you have an exam that's consistent with coma, vegetative state or MCS- at the bedside so it would include people who could track or show auditory localization but that fMRI or EEG some other measure shows that you can follow a command so we've dissociated it by the measurement and your bedside exam but what remains completely ambiguous is where are you on this x-axis you could be somewhere down here you could be fully conscious what I'll show you now and I'm anticipating the conclusion so the last part of the talk is the weight of evidence is saying you're probably over here to the right these people have structural brain injuries in the central nervous system so maybe you're over here to the right most of the time maybe you're always over here perhaps that's not what's going on perhaps you come in and out perhaps when you're sick you're down here you're in some range and you have two problems one is that you don't have a communication system the other is you don't have a brain that has stable integrative function that's the nuance on this game the next point to make with this slide is there is a proposal to call vegetative patients unresponsive weakfulness syndrome the problem with that is it's not a mechanistic label and that it really encompasses all of this so these are the arcana of nomenclature but to get onto the data so two groups have looked at cohorts of about 50 patients and found about four or five who show cognitive motor dissociation this is a paper we published last year looked at 44 patients we had studied at Cornell patients had FDG PET in most instances and fMRI studies to look for command following we found four examples of patients with command following I'll point to this patient here is going to come up again and I couldn't find the slide using the other technique to show so remember you'll see this man's command following I mentioned that I showed it to you but I don't have the slide these four patients were the only ones in the group who showed fMRI responses and when compared against the entire group of patients who we had studied using continuous EEG measurements 24-7 3-5 day stays and over 2200 hours of recording we found that all of these patients showed very well preserved sleep and wake architecture and they showed well preserved cerebral metabolic activity so that if you were following commands you had sleep spindles you had a mildly abnormal EEG background and background and you had well preserved cerebral metabolic rate so these studies say you look more like normal subjects you don't look like the MCS population or the VS population from the point of view of how much energy your brain is using or how well it's organized electrically you look more like a healthy volunteer you may not be a healthy volunteer and you may not have a normal state of consciousness until you can initiate a communication system and tell us we don't know but that ambiguity can be weighted by these data similarly Steven Laurie's group did a very large study comparing PET and fMRI and they found that which we all know is a very insensitive test finding people who do high level motor imagery tasks is sparse in patients who at the bedside show signs of consciousness and minimally conscious state but when you do find people who do it FDG PET is always preserved in their study as well but it's also preserved in patients in minimally conscious state so these are again data saying that the patients who have all of this are subset over to the right and of course there's going to be a continuum in a spectrum in this population as well but just to give the clinicians the opportunity to really get this I've done this a few times and I find it kind of fun we're going to do a Rorschach test this is a clinical Rorschach test this is the last part of the talk I'm going to show you two of our patients I'm going to take you through the data but I'm going to stop and ask some questions before we get to the end two patients one is a 22 year old woman with an idiopathic etiology of occlusion of the basilar artery this is a stent in her basilar artery you see she had an extensive injury to the brainstem those of you wondering where there's a connection it's lateral to this image this is the medulla that's the spinal cord there's the quadrogyminal plate and that's what used to be the ponds and the midbrain in this cut so laterally there's a little bit left on each side of both but not much this lesion extended into the interlaminar nuclei in both thalamine and knocked out the lateral deniculate body on the right very extensive injury these images were taken at Cornell almost two years after her injury and they show widely preserved intact gray white with very little atrophy across her entire cerebrum nonetheless her exam at the bedside even in our hands for the first day was vegetative state she had come to us with the observation that occasionally she could track with one eye down and we we established for the first time at 18 months after her injury that she could actually follow the command to look down and to intermittently use that channel to communicate that's patient one patient two is a man who was in a T bone motor vehicle car went into the middle of an intersection it was hit, spun around and ended up with diffuse axonal injury suffering very bad radial diffuse axonal injury had dilatation of the ventricular system widespread atrophy damage and this is actually a good cut of his brainstem a very thin brainstem in most of the section so very severe traumatic brain injury remained in vegetative state for three months and then his family noted that he had visual fixation at six months and when he came to us we saw that he had head movement to command three years after injury so here are your two pieces of data and honestly I can tell you that the family had noticed this, they had worked with the therapist there were some more details I'll share with you later but most of the doctors and people who had seen the patient had told them you know get real and he came with the label vegetative state the other woman came with the label vegetative state with the caveat that a very skilled neurologist had made this observation of downward eye movement and was concerned because of the degree of preservation so both these people came with an understanding that there was a question of ambiguity and I'm giving you you know our first look with the experience of our specialty center but now we have the behavioral and the clinical data let's do a tally just quickly how many people think that we could establish evidence of command following or communication and either of these patients using an fMRI a couple people actually a few and then how many think that if we could show evidence of either command following or communication that we could a systematic communication system could be developed to aid the patient most people and then how many people feel that their answers to this question is more than just the impression of an ink blot okay not too many so you're intrigued to know what the ground truth is I'm sure so we'll move on so we set out and this is the first paper to look at cohort of patients with the technique of functional magnetic resonance imaging to see whether command following was uniformly present in patients who could or couldn't follow commands if it was present could they be turned into a communication system using a strategy for using their command following to answer questions and we looked at six patients two of them were studied longitudinally and what we found in a nutshell was all manners of dissociation we found people who could communicate at the bedside who couldn't generate these signals people who could not show command following or communication at the bedside who could and we found people who could show command following signals but then couldn't turn them into a communication system and vice versa so let's look at the two patients that I framed for you so here here are group data for our healthy volunteers this is one of our controls and it shows the canonical activation in the supplementary motor area or in our case imagine yourself swimming because the woman who is now a publicly identified figure her name is Maggie Worthen I'm showing you here and I'll show you her picture at the end of the talk was a swimmer and she's the subject of Joe Finns' book that's just come out and she was able to do this task perfectly in fact these data which are taken from one of the areas outside of the canonical region of interest which is always present when you do this task it's just that other groups tend to mask the other activations in the brain so this is the area so you can see that it's not just this area that activates it's other areas and sometimes the other areas are actually better this is as good if not better than any of our normal subjects in terms of a reporting region of interest so she could do this command following task very well I don't have the slide from the other subject because I couldn't pull it off my dongle and get it into the talk and I didn't realize it wasn't in here until I got to it but I showed you his command following responses in the other paper a few slides back when I said remember I'm showing you one so they both showed command following now interestingly that patient could not do what I'm about to show you she could do the next test was communication so we set up a communication paradigm that was a little less complicated than others in that we only had the one response imagine yourself swimming and we used it in a multiple choice framework so that the time structure of when it arose would be the reporting information as to when the signal was being generated in response to the question so we showed healthy volunteers playing cards a confederate would show them the card we then went into the MRI machine and the person who analyzed the data knew nothing about the card and we found that 100% of the time healthy volunteers unsurprisingly like in all of these tests could signal accurately the card using the supplementary motor area as the reporting region of interest and tell us that their card in this case was the ace of diamonds so when we went to the woman named Maggie who had the brainstem injury and asked her to report to us her card which was the ace of spades we got this result which was different for those who noticed that there are two areas that seem active here note that this is the reporting area and it's deactivated here not activated this is only one response but in Maggie we actually got two responses one for club and one for ace that were in this region of interest and then one unambiguous response for the face but it was wrong it was Jack and we struggled with this because she could do the command following so well that signal was great and we weren't sure what was going on but we did know that she had these injuries to the central part of the thalamus which in patients who don't have brainstem injuries and have these injuries alone tends to slow their responses so they can delay for 20 or 30 seconds in the motor response so we thought maybe there's a delay and maybe what's going on and this is why these are going to be difficult measurements right this is the difficult measure middle of this because now we've got structural brain injury we don't just have somebody who's locked in with a brainstem injury we have somebody who has a brainstem injury and has thalamic injuries or stridal injuries or multifocal injuries right so what happens well we looked at this and we realized that in our healthy volunteers there was this very smooth blood oxygenated oxygen level dependent signal that kind of went up and then just went down but in all of our study patients and this was reported by our colleagues in Cambridge University to there were sort of these biphasic responses and we were always serially ordering our presentations so that club always followed Spade and Jack always followed Ace so it wasn't impossible that that was what was going on and it took a very clever graduate student John Barton who did his PhD with me to come up with a strategy which is to use multivariate pattern classification which is a technique and algorithm that can model the full whole head MRI signal and he took the command following data that were so unambiguous from this patient as the model and then he used it as that to filter the communication responses and when you did that what you found was that there was only one response and it was the one that was at the delay the others were at chance and it was at just the same criteria in a classifier performance as the one that was unambiguous in the other run and so we had pretty good evidence that she was one off in two separate runs and recognize that these three experiments do the command do each of these take 40 minutes of time for this person in a scanner so this is like a lot so we think that she was delayed and she was trying to communicate but the other patient who had the command following response did not communicate we tried and we tried this trick and it didn't work it wasn't in the data so by a show of hands what do you think the clinical outcome was in these two subjects you get to choose one so I'll read the choices and then we can vote choice one is neither patient establish reliable communication using augmentative communication systems choice two is patient A the brainstem stroke but not patient B the diffuse external injury patient establish reliable communication using augmentative communication systems choice three patient B but not patient A so the DAI patient but not the brainstem stroke patient establish reliable communication using augmentative communication or choice four both established reliable communication so who wants to vote for neither of them establish communication one person who wants to vote for patient A but not patient B majority who wants to vote for patient B but not patient A no takers and who wants to vote for few okay so this is how it works every time I've shown this I went to the Boston Neurological Society and the answer is three the answer is three okay the answer is three because these tools you know without a so what next step don't take us to what we might be able to do I'm sure we could have done something for the woman with brainstem stroke with other technologies that we saw evolving this week at the neuroscience meeting but we couldn't do it with standard technologies like putting a mouse on the patient's head which is what our medical student who went on to be a neurology resident did for patient B and he turned it into a very effective brain computer interface system that's him a few years later from our first studies and this is the email he sent us when we invited him to come back because we wanted to see how he was doing he said hello going sounds really good because I can show you what I can do now which included writing a book showing interest in going to culinary school kicking us out of the room repeatedly so he could have an hour to email his friends during the study and interestingly over about three years he was being able to use a thumb so he could use a joystick so the point here is the following only when patients can get to the point where they can initiate communication can we probe their quality of consciousness and their subjective experience but we can all agree that our expectation was that this guy's likelihood of harboring that quality of experience was much less than the woman based on those images which probably means that her quality of experience might exceed his but not that he didn't have one but we don't know how to deal with that yet and to date no study has really shown that you have a patient who is meeting VS or MCS criteria and has these communication or command following findings that they get to consistent BCI now there will be patients like that but they'll be closer to normal and the question is how do we deal with everybody and that's really the next step and finally this is the last thing this is to show you how far we got with the woman unfortunately she died this summer and we were hoping that she would be the first participant in a new study that we're hoping to be able to do for patients like this but we had hooked her up with the Speak Your Mind Foundation which is a group in Boston that's the nonprofit associated with the brain people and this is Dan Bacher I didn't put the website here it's called Speak Your Mind if anybody wants to look it up on the web he's an engineer at Brown and he works in both the BrainGate and the Speak Your Mind Foundation he created a $14 solution with some very very expensive high-end training as a software engineer to use a CCD camera that meticulously captures the eye of this patient and this is after a few training sessions over about 7 weeks her ability now to harness the downward eye movement to rapid command so she's going to get up to 90 this is the end of a 30 minutes a 3 minute session but we in the early days could struggle to get a few reliable clear command following eye movements down so she could learn and at times she could use a technology to answer questions but she never got to the point where that was reliable enough to establish initiation and a full range BCI so her whole story you can read about in Joe's book that's her with her mother Nancy and it's a pretty remarkable story and with the ethicists and the audience I know there are many who will understand now I think is how unfair the treatment of this mother was as she tried to protect her daughter to give her a chance because pretty much at every step along the way and her and every other patient's family that we've seen they're all treated the same way which is that you know the initiation team is right outside the door and you know let's just go that's just you know you just can't make that a blanket approach and what it was like on the inside and what kinds of issues that she had to struggle with and clearly we didn't solve the problem for her daughter but it was a real problem to be framed so it's complicated I don't suggest there are any easy answers here but that's the frame of all this work there's a lot of good available evidence supports that patient subjects who show imaging or electrophysiology evidence of motor imagery and communication probably in a separate category and they're likely to be a little closer to locked in but as you work through the data and try to understand how our measurements relate to mechanisms underlying recovery of consciousness there are going to be problems in interpretation it's going to really rely on the biology and this is going to be a more complicated area to fill out than say cardiology was just you know for the clinicians in the audience there are some very positive points one is that both FDG PET and standard time domain EEG measures which are normalized for general population use are probably very useful in this population because they're well validated and they're widely available and particularly the clinical EEG which is inexpensive so we've argued that anybody who's in minimally conscious state or vegetative state when they leave the hospital should at some regular interval reassess behaviorally and have an EEG and if it normalizes then we've got to have a path to doing more for those patients to figure out where they're at and when we find people with the evidence I've shown you today we need a path to start clinical trials to see what kind of things they can do with some of the more modern BCI technologies that are available so I'd just like to thank the people in my lab Joe Finns my colleague who probably got if Techie wasn't going to invite me here then this is why I would have been here my other colleagues at the imaging center our mentor Fred Plum who really started us all in this direction and a group of international collaborators who are working together to try to meet some of the goals that I've suggested to you you need to do that and our funding agencies and that's it thank you we got through three patients in our first clinical study one of whom was the unambiguous responder we published the paper in nature in 2007 the other two patients both showed physiologic effects but had no movement on the primary outcome measure the coma recovery scale revised one of the patients interestingly recovered executive motor control within the first stimulation of the thalamus on one side held it for the first couple days but had a very disorganized thalamic electrophysiology pattern on both sides because of large brainstem deaffirmentation and when we turned it off for 60 days and brought it back on that patient had lost that effect and so he did not show a change in the trial the other patient of the first three was 20 years out had a very severe DAI was half the behavioral level of the first person and only made it into our trial because they could show command following in an arm that was otherwise dystonic that person showed no effect at the level of behavior on DBS but we continued to track that person and we reported last year the Society for Neuroscience meeting will have a paper out on that subject that they actually normalized the electrophysiology of sleep which is interesting because the electrodes were off during the sleep period they were only on a 12 hour duty cycle so there were clearly effects of afferent stimulation in that patient across the corticophilamic system it's a complicated story in terms of his clinical exam but there were signs that he was a little bit more responsive but he was a trial failure so those were the three that we got through for minimally conscious state we have been stymied by a lack of interest in funding a phase 2 study so we can get criteria established I mean clearly the patient who responded and got better is not the only minimally conscious patient in the world who could be helped in the way he was helped. Right now unfortunately he is because there's not a lot of enthusiasm to go after solving that problem that's the downside the positive news we've made huge progress in understanding the circuit mechanisms underlying recovery and MCS. I showed you a little bit of that with the Zolbinum there's a paper on positron emission tomography showing the change in the central thalamus GPI and frontal cortex that we predicted between EEG and PET criteria I think we could come up with selection criteria for patients who might harbor reserve to do the DBS we've had big advances in the pre-clinical work in that area so we've had we have two papers coming out hopefully soon one using optogenetics and rodents from a group at Stanford the other primate study characterizing the actual target in the intact primate brain we have a very good idea of where that is and how that works and which cells we're contributing the vocular variance and then finally we've just been funded by the brain initiative to do a first in man study of central phlamic deep brain stimulation in a higher level population of traumatic brain injury patients so what we learned from that study will undoubtedly move our knowledge back into the minimally conscious state as well so we're hopeful that this is going to become a very viable area of therapeutic engagement there's a first of all thanks for a very very stimulating lecture I I'm reminded of a famous philosophical essay called what's it like to be a bat and there's the question that I'm asking is what's it like to be in the minimally conscious state you at least have the advantage over the problem of reading the mind of the bat is that you have a few people who have awakened from that and I wonder whether it's totally dysphoric to be unable to communicate and yet aware of what's going on or is it a state that's characterized by constant urge to communicate and hopefulness or is it so primitive that there's none of that whatsoever because that would be helpful information in assessing and talking to families early on about what the state is that we're preserving if we're trying to do that so I'll go back to my undergraduate philosophy degree and give you my answer well to start with I think you know it's an operational question that once we can answer it we probably not talk to somebody in minimally conscious state right so that once we establish a communication system with someone who's clinically vegetative or MCS by behavioral criteria and they're sharing their subjective conscious awareness with us they're probably going to be physiologically speaking like locked in we may find they have memory problems there's more under the hood here we have some examples of people who have communication systems who show us that they have evidence of anterior grade amnesia for example so they're there they're answering autobiographical questions but they're not answering forward questions in time well so there's everything in between I think from just a philosophical point of view I would take I would take the question this way which is that any capacity to be dysphoric has to in principle be containing the capacity to be euphoric you know you can't you can't have extreme pleasure without pain and vice versa so you have the potential either have the potentiality to experience mental states or not and well hopefully we'll get to a point where we have some correlational data that guides us more finally with this and the kind of correlational data that we might get is the following if we get patients who can use implantable brain computer interfaces of the type that are that were shown at the meeting this week in Chicago neuroscience which are meeting criterion levels that no one's ever seen before being able to type at rates that really make them viable devices in this sort of setting then we're probably going to start to see people who slip physiologically and you know into conditions that are at least from that point of view more similar to MCS than they are to their state when they're communicating then we will have with in-person opportunities to query them last week when we couldn't get you to use your BCI do you remember anything did somebody come to visit you did you hear a song playing you know we'll be in that instance once we're given a window onto subjectivity start to accrue the kind of evidence that would give you some insight into that question so I don't think it's a non-starter but I think it's at the technological edge of where we're at there are a few pieces of data like that but you can't generalize them and it's hard to know exactly where they were at so they represented a minimally conscious state or they represented somebody who couldn't move and wasn't recognized to have a higher level of function in other words the question you really want to guide families is something that's contingent and lawful about parameters that consistently associate with minimally conscious state and aspects of subjective awareness we don't have that yet not even close but what I'm suggesting is in principle that's not opaque indefinitely I'm saying you well I guess what I'm saying is that we shouldn't assume that you could only experience negative emotions that's what I'm trying to say is that if you have a brain that's capable of truly experiencing dysphoria we I don't think there's any neuroscientific explanation foundation to believe that in principle you couldn't experience other affective states in other words I don't know of a model that suggests that a human brain in principle could be restricted to experiencing only one affective state not that you couldn't be dysphoric all the time or be trapped in dysphoria but that the wet where per se was conditioned to be experiencing one affective state and particularly that if we found evidence that that was typically more the case than not to attribute it to the class of patients I mean actually let me be much much more specific because there is a way to condition this expectation which is the data from locked in patients which does exist in the aggregate intabulational form from multiple groups where just those kinds of Likert scale questions are asked about quality of life how they feel about their condition where they're at and the surprise in that data is that they all feel the same way that the average person in this room does like things are okay the one difference in that cohort is their willingness to sign a DNR order they have the lowest willingness to sign a DNR order less than 5% will sign a DNR because they're afraid that people will just let them go and there have been patients in locked in state who have been resuscitated come out and still won't sign their DNR order so I guess you know I guess part of what you're hearing and what I'm saying is not to slip into the assumption that from our point of view this would be a horribly dysphoric we never trade our life for that that doesn't give us insight into what it's like to be them just like we don't know what it's like to be a bat that I think is a more complete answer yes I think they're very related mechanistically in fact this is kind of this is come up there's a guy at Sony Brook Max Bank and I have talked about this he's written a big book on catatonia you know what's probably going on I didn't make the connection and I could have I just I dropped in the talk this paradoxical excitation phenomena that you see in anesthesia if you look at what happens with zolpidum as an agonist it's activating beta rhythms in the frontal cortex and the striatum it may be directly inhibiting palvel inhibition and the net net effect is bulk excitation over the forebrain to unlock a system that's in this case been locked up because of loss of glutamatergic transmission in many catatonic patients there's some dopamine abnormality some alteration in function from schizophrenia or in some cases from D2 blockade or other phenomena a similar mechanism would be thought to underlie the response to catatonia well you know I guess most people are not excited about inducing generalized seizures and structurally injured brains because you may get runaway excitation and many of these patients do have underlying at least vocal epilepsy and things like that here and there so on average that's one of the reasons why not but the other reason is a much more subtle reason and I can direct you to a reference which you may know or may not know you know the paper by Burris and Chaco 1999 so it's in the journal of I think neuropsychiatry and it's a great reference because it actually it's one case but I think the general point is the same ECT will up-regulate gene expression the first couple times it's used and then chronically down-regulated so these guys found actually a catatonic patient with structural injuries after stroke and the basal ganglion and the thalamus following a subarachnoid hemorrhage who had remained akinetic mute for 18 months and then he had a generalized seizure and started speaking for the first time just like the fireman in other cases like this in literature and the wife noticed that she brought the patient to doctors and happened again so they decided to do ECT and it worked the first two times and then it stopped working so yes and no probably