Professor of Neurology and Pediatrics, Dr. David Frim, discusses concepts to improve care for children with neural injuries, primarily of the neck and brain.
[MUSIC PLAYING] DAVID M. FRIM: Well, thank you very much. And it's always a great pleasure to talk to all my pediatric friends. Hopefully, there's a lot of them. So neurosurgical injury has become so complicated that I thought in half an hour, rather than trying to give a whirlwind through a lot of specific entities and practical issues of management, I thought I'd spend 15 or 20 minutes going over some conceptual things. And then go over a few of what I call kind of the controversies. But a lot of them are sort of myths that we hold. And a lot of things are based on belief, as much as they are on data or more. And then maybe we can entertain some questions to discuss how we approach that. So here is my disclaimers. So just dividing up the next few minutes, I thought we could go over this definition of what a traumatic neural injury is and then talk a little bit about the mechanisms behind that. Spend a few slides on when we decide that a neural injury has occurred or what the markers are that we have now in 2015. And then spend the rest of the time talking a little bit about how we can use those concepts both in the laboratory as an example, and then ultimately at the bedside, or for us, the operative table side, to inform our clinical care. Then I have a few slides, which would be nice to spend a few minutes on, of current practice and the controversies and whether or not there's really data to support it or we just do it because that's what we've always done. So what's a traumatic neural injury. So that's the definition. And I'm this sort of excluding spine a little bit. The spine is somewhat different. And probably best to just think of most of what I'm going to say has to do with the brain and the head. So in the IDEA law, which was passed 2004, they actually had the definition. I guess it's the government-accepted definition of brain trauma, acquired injury to the brain. It has to be an external force that results in total or partial functional disability, psychosocial impairment, or both; adversely affects a child's educational performance; and on and on and on; results in a number of impairments. So this is the totality of injury to the brain, at least according to the government. But it's not a bad definition because it's all-inclusive in that the injuries that we receive to the brain can really stop almost any or all of our function. Here's some examples. It's the usual stuff. I have to show them. I feel obligated. Epidural hematoma, everybody sees this. It generally has a very good outcome. It has its own imaging. Interesting enough, there's never been a double-blinded, controlled study on whether we should actually remove epidural hematomas or not. We probably will never do that. But it's something that we do by belief. But it's probably a pretty good one. Subdurals, maybe a little bit more controversial just because of the way they can accumulate or not. Here's what they look like on the inside. Oftentimes, it can be very troublesome in children, more so even than the epidurals. So far, the only thing that has data to support the care of subdurals is that the sooner you bring the child to the operating room, the better is the outcome. Operative technique and all of that, nothing has been proven one way or the other. This is a thrown-against-the-wall injury. That shows another of the things we see, a skull fracture or the penetration into the brain below that, a lot of brain direct impact from an orthogonal force. And then the various repairs that had to go through, both for removing the brain that was already beyond repair and then trying to make it look like the lower right-hand corner, where there's still a little bit of a contusion. There's penetrating trauma to the brain, gunshot wound. It's a frustrating thing for us because the forces involved are great and the impact is not that great. It's difficult for us to think of what's happening to the rest of the brain while we're trying to work on a small pellet or a bullet that's now imparted all that force and injury around it. And this entity, which I'll come back to in a little while, which is a nonhemorrhagic contusion. And it's a rare thing. I only have two or three pictures of experience with this over my whole career. But this is maybe is something that is going to lead us to understand this controversy on concussion and concussive injuries. So what are the mechanisms, of which there are many, that underlie this? This is a very kind of dense slide. I'll just go through it quickly. So close your ears, if you don't want to hear about it. [LAUGHTER] Orthogonal forces cause severe traumatic brain injury. That would be sort of the baseball bat to the head, kind of thing. They don't generally cause mild traumatic brain injury because mild traumatic brain injury doesn't cause a lot of crush, instantaneous death of what's inside the head, particularly the neuronal populations. If you think of orthogonal forces as sort of blunt or penetrating, their necrosis is nonrecoverable. And all these other names go to as crush injuries, tears or schisis to the brain, evascularization. So head and brain motion though are more than the head contact, with a degree of angular acceleration and deceleration. That's what causes moderate and mild traumatic brain injury. And that's going to be an important distinction, probably in the next several months. But hopefully in the next few minutes, when I talk about it. These sort of rotational and angular forces then cause force vectors of tension that stretch neurons. And the neurons can be short or they can be very long. And in tall people, the neurons go for over a meter, from the head to the bottom of their spinal cord. And that leads to these entities of diffuse axonal injury or a generalized white matter injury and then the degeneration that goes with that. Membranes break down with all of the above issues. And the neurons that survive the direct injury then have to survive repairing themselves. And one of the hardest things that the brain has to do after injury is to self-seal its membranes. And it's very energy consuming. And the mechanisms for that are also many. So the cascade that effects the neurons are this list of things that I made. Unregulated ion fluxes-- so that leads to the metabolic stress because ultimately neurons are just electrical circuits. And whenever the wiring is disrupted or the installation disrupted, that has to be repaired. And the things that disrupt it are the reactive oxygen species; the peroxidation of the lipid bilayer of the membrane; and the direct cell death around it, with the release of various toxins. So microtubules disassemble when that happens. And that decreases the axonal transport. And then the brain can't move things around the way it would like to. And that leads to what we call axonopathy. So the axons don't function. So now, all of a sudden, the chip's working, but the wiring doesn't. And the brain responds, also with the body, with inflammation. And that, though it's a reactive process, it can add injury to the mechanical problems and then lead to a secondary cell demise. And then beyond all that, even if the neuron survives for a few days, there's apoptotic pathways, that are understood to a greater or lesser extent. That can be chronic and are usually irreversible. So that's what's going on over the course of time. And when we're treating these patients, at some level we have to be thinking about that. The time course is variable. We don't know if symptom relief actually approximates the end of the process, or the beginning of the process, or the middle. All these things are really not understood, as well as why the same injury to one person does nothing and to another person can put him in the hospital for a month. This is a quick slide. Don't pay much attention to it. It just shows that there's a lot of things, maybe 1/10 of the things that go on in the brain when you get hit by a baseball bat. And the time course is over minutes to days. And one of our biggest problems right now is understanding all of these things. Now, I've got some cartoons. I love cartoons. They're easy for me to think of. This is a hole in the membrane. And you can see all the little circles and triangles of the electrolytes. Now, they're free flowing. And then the brain has to fix that hole or the axon individually has to be fixed by the cell before it just kind of explodes from losing it gradients. This mechanistic slide, this comes when I was a postdoctoral fellow. It was a long time ago. But we still use it and think of it that way. That the neuronal death on this pathway has to do with the inability of the cellular energy generation to keep up with fixing the membranes. And if you think of it, much of what we do practically-- that was just figured out, I think, on an empirical basis-- is based on that. It's kind of the modern version of all these things. It combines these various mechanisms that have to do with glutamate toxicity, which is a cellular and an energetic insult, along with a variety of cascades of intracellular protein, either release or synthesis, that leads to pathways that could be cell death or cell repair. And that slide becomes one more complicated as the years go by. OK. How do we decide when there's an injury? Clinically speaking, we use very blunt tools. Did you get hit on the head? It's a good question to ask. And if somebody remembers that they got hit on the head, they probably had a head injury. And what kind of force was it? Then all these other things. Were you knocked out? I've always wondered how we ask a child were you knocked out? If they were knocked out, they probably don't know they were knocked out? And then how we then get that information from the people around them. And then various other functions, mental status changes, the physical exam. It varies from being very specific or what seems to be specific and severe . In traumatic brain injury, we have our Glasgow Coma Scale. We have other coma scales. We have specific things we test. But that specificity, if you want to think about it that way, is lost when we try to use as a biomarker across all brain trauma. And then we image. Imaging is probably our big hope here because it's much less blunt than everything else. So obviously, we use a CAT scan. It's easy. More and more people are using MRI for trauma. And then there's a variety of what I call intermediate length of injury things, like the PET scan. They can look with various ligands at exactly what's going on. And the degree of injury can correlate with a variety of these PET-specific ligands that will label different things going on in the brain. The list of biomarkers is very long. As to which ones are actual biomarkers, that's kind of different. So here's another picture. I found this because the writing was really small. So you shouldn't worry about reading it. Just realize there's a lot of letters there and not one of them is an absolute biomarker of injury. We talk a lot about tau protein build-up and amyloid plaques for chronic injury. That's useful. But we don't know whether that is a response of healing or a marker of impending cell doom. Leaking things out of the neuronal cell body, like neuron-specific enolase, probably a bad thing. But we don't really have proof of that. So this one is for traumatic brain injury in general, particularly severe or traumatic brain injury. The list for a marker's potential or fluid biomarkers, things that you can just draw by blood tests or CSF tests, is even longer for a mild traumatic brain injury and even less specific. But just as an example of the things that are being worked on that are probably going to be available to us not too far in the future, is, for instance, this paper that was in JAMA at the end of 2014, that looked at these Swedish hockey players who, I'm told, like to hit each other on the head as much as possible when they're skating around and bumping into each other. I'm not Swedish. So I can say that I think. And they chose a couple of proteins in the serum to study. Tau was one of them and also the S10 and the neuro-specific enolase. And found that if they follow the cohort of players and just marked which ones had clinical concussion, meaning they got knocked out or they were hit and they were dazed, and just kept track of that, and then drew blood from those players and then everybody else, they could correlate the rise of these protein markers the more they were concussed. So it may be that eventually we're going to be able to draw a blood test on the field and dip it in something and if it turns purple, you're out for the game. And if it turns black, you're out for the season. And that would be very helpful to everybody because right now it's very difficult to standardize that across all of the concussive injuries. So how can these concepts that I've whirlwinded through begin to inform our approaches to treatment of neural injury and specifically for this talk, neural injuries in the brain? This brings me back to this nonhemorrhagic cerebral contusion. So it's just a traumatic injury. There's actually a few of these little flare bright spots in the brain. And there's no blood. So something happened to this brain to do blood-brain barrier breakdown. So this brings me to a few slides discussing how we model these things, if we're wanting to figure out how to treat them because we can't concuss humans experimentally. I guess that was one of the topics this morning, us doing that. It's not a very good thing for the human brain. But we can make models of various transfers of energy in various animals and see what happens. And then that can inform how we might want to interrupt those cellular processes. Here's a device that we conceived of, which is an impactor for rodents. And it has this little head holder, which in many ways can be modeled as a helmet with two sides. So the brain actually can move inside the head. And the head can actually move inside the two bars of metal, which would sort be the two sides of the helmet. And it can cause fractures to bone. And more interesting to us, regarding that nonhemorrhagic contusion, it can cause injury at low joule strength to the brain, that you can see in the animal as it's waking up from injury, but no bleeding. And then we can look, if we're quick enough, to find areas of blood-brain barrier breakdown, which is what this IGG immunostain can tell us in the brain, both on the side of the injury and on the side opposite the injury. That would not be seen on a CAT scan, for instance. Because the CAT scan can't differentiate blood-brain barrier, for instance, the way the MRI can. And then we can look at other parts of the brain that are away from where the impact was. And we can say, wow, the brain stem also gets concussed to a certain point and even there's a tear in the tissue. The most interesting part about all of this is that these tears, which we've decided to call brain schisis, disappear, along with the blood-brain barrier breakdown. And if you do that same thing to another animal, and you don't MRI them, for instance, or you don't immunostain them for a couple of days, all of this disappears. And there's a whole anatomy to how the tissues can be torn within the frozen head. So this is a model where we can create this entity of nonhemorrhagic contusion. And I use it because it's rare and because it's unrecognized. And it's been sort of something that I and some of friends have been thinking about for a long time is how to do that. Because ultimately, I think, we're going to find that that's what may be behind the issue of concussion is that there's some injury. There is blood-brain breakdown. And there's an energetic loss while the brain is trying to repair itself. And then decide how this, what's probably end of a model of mild traumatic brain injury, is recreated and how we can study it. One thing that comes out of this is that I would propose that if you had an MRI scanner, where you could scan an athlete that was knocked out within the first few minutes after being knocked out, you could probably see changes. And that might be a way to mark an injury to the brain. OK, skipping on to then treatment. So what are the general concepts that we can take away from the first 17 minutes of this? Pathophysiological mechanisms of the disease generally determine how we derive therapy. So that's a general approach to most things. In this particular issue, there are those list of four or five things that I wrote down as mechanisms. And so we can look at those and decide how we can as sort of, what can I call it, molecular brain surgeons-- I don't know-- try and repair all that. So unregulated ion fluxes could be repaired by some sort of metabolic augmentation, if we had a way to make the brain more energetic, if we had a way to do membrane reconstitution, if we had a way to keep the lipids from being periodixated with some sort of oxidation. So all of these things, if they could be delivered in a timely fashion to the cells, would help them get better. Disruptive membranes, the axonopathies as far as moving things around the individual neurons and then around the brain. Is there a way to acutely repair the membranes? Some molecule that would fill the holes? And then other molecules that would make the microtubule movement up and down the axonal-- oops-- up and down the axonal reach get better? And then the inflammatory damage. What sort of anti-inflammation is going to be able to get into the brain and not injure it and keep the inflammatory cells from adding to the injury. And then all these other later things have to do it the apoptotic cascades, and the glial scarring, and toxic proteins, the deposition. If you believe that's what the tau and the enolase is. How do we make those things not happen and go away? So here's some of the things that are either being thought about or are in trial and some that have failed. So one is, of course, the antioxidants. We've heard about that. Another is the neutotrophic agents. Though oftentimes one thinks of those as having to be given before an injury. Proteolysis inhibitors are for historical interest only because I think that has become a dead end, that's no one's traveling any more. And then there's a number of synthetic surfactants that are in clinical trial at various stages now. So a slide about each. Antioxidants then in traumatic brain injury models. This, I thought, was a great example of kind of a thought process gone wrong. The megadose methylprednisolone for spinal cord injury was very popular for a while. Though in the end, the large first study seemed to turn out to have a lot of methodological problems. In this situation though, in the experimental model, that dosage of methylprednisolone was given a couple of hours before a lesion. Interestingly enough, it made the brain worse. So this is an avenue of treatment that hasn't yet matured enough. And then with regard to steroids for trauma, there's not yet been a human head injury in a child that has shown efficacy. And the one that was maybe a little bit was in focal contusions to the brain. That was done in Glasgow. How about the neurotrophic agents? So lots and lots of reports in animal models of amelioration of injury by nerve growth factor, [INAUDIBLE] growth factor, and brain-derived neurotrophic factor. And many of these have been in human trials. But none of the human trials has yet shown efficacy. We don't really know why because they seem to work so well in the animal models. But I think that ultimately this is going to probably be able to be put together. And then the membranes. Some more cartoons, but this is a good one. If you make a hole, you have to find something to fill it. And there's a number of companies that make these synthetic surfactants that are called poloxamers. And they can plug the hole. And so the membrane doesn't leak. And the energetics that would have been expended on repairing the membrane now will no longer need to be used in that direction. They can do other things, like repair the axonal transport. And it works in an animal model because you can do these lesions in volume and you can use the poloxamer molecules to make the injury smaller. And these substances are in clinical trial. OK, so perfect. I got eight more minutes. Let's talk about what we believe, what we can support, and what we do. So here, these slides, these are old slides. You don't have to pay much attention. How do we treat suspected head injury? We'll use hyperosmotic agents, mannitol, hypertonic saline, all those things. We hyperventilate. We argue about what numbers to hyperventilate to. But I think everybody does that. We measure a cerebral perfusion pressure. And we either raise the blood pressure, or lower the blood pressure, or try to lower the intracranial pressure. We do all sorts of things. But that requires arterial monitoring and intracranial pressure monitoring and a variety of the things that we can't do, like give free water. And we almost always give anticonvulsants, which have some side effects. And it's a little bit unclear if they change the outcome. In the ICU, we do similar things. We watch the perfusion pressure, sometimes for days. And in general, and I even believe this too, that cerebral perfusion is probably more important than intracranial pressure. Though that remains a controversy with some people. And then we've developed this concept of euvolemic hyperosmolarity as the goal. To keep as little extracellular water as possible in the brain and most of it out of the brain. And hypervolemia is better than hypovolemia. And then we closely manage the PCO2. And then we figure out ways to take things out of the brain because less things means lower pressure. A lot of it has to the pressure. A lot of it has to do with perfusion. And this is what I call the known clinical management factors affecting neuronal survival in the brain, of which there are not many. But these are neurosurgical issues. We relieve direct compression, sometimes to miraculous results. The kid has a blown pupil. They're not moving half their body. They have a big epidural hematoma. They happen to be in the right place at the right time for a helicopter to take them to us or to another institution with a pediatric trauma service. And they go home, post-op, day two, normal. It's miraculous. The cerebral perfusion pressure. We think perfusion in the brain is also very important. And I would agree. It's very important. And when this was first thought up or bumped into by accident, this also was miraculous. Patients after trauma, who with their heads up and with no idea of what a perfusion pressure, are just having mild increase of blood pressure because of volume repletion, would be laid flat and then they would be hypertensed. And all a sudden, they were awake. And then the issue of the hypoxia after injury and how that affects the energy metabolism. But what are the controversies? So mannitol and hypertonic saline, so we all use it. Is there evidence? Interesting enough, in the last few years, most of the evidence is comparing the two without asking the question, do we really need to do that? Is that the best thing that's making that sodium 155 or above? Is using a lot of mannitol so that the osmolarity of the serum goes over 310? I couldn't find a single trial of not doing that. So we all do it. And it's probably the right thing to do because it seems to work. But the interesting thing is that there's not a convincing statistical difference shown in any of these comparative trials between mannitol and hypertonic saline either as far as I could tell. Both seem to be equally efficacious. Though there's a lot of comment about which costs more. I would say that mannitol has this added benefit of making the red blood cells more pliable. And also it's a weak antioxidant. But I don't think that that's a measurable benefit right now. How about the glucocorticoid steroids? They don't really seem to have an effect on brain edema in a trauma setting. Though oftentimes they will in a tumor setting, which could be also miraculous. Some people are using methylprednisolone in spinal cord trauma. And to be honest with everybody, I use it sometimes too. But I think most people who have looked at that initial study think that it's been well discredited because of methodological issues. The people in Glasgow, they did some dexamethasone trials as they closed head injury. And they determined that it didn't really do much. And then, as I mentioned before, methylprednisolone is maybe good, but maybe bad, for experimental head injury. And beyond all this is just the plain observation that one milligram of dexamethasone saturates all of our cortisol receptors for over a day. So whatever it's doing at 10 milligrams, or three grams of methylprednisolone, has nothing to do with our usual use of cortisol. It's some chemical nature of the thing. Hypothermia, which I haven't mentioned because I would just go to the bottom thought. I couldn't think about pathophysiological basis for cooling the brain to enhance its repair. Interestingly enough, there was a big trial that the NIH funded at I think $12 million. It's supposed to be less. The trial was only supposed to run I think four years or so. But it ended up running almost twice that long and they couldn't recruit enough patients. But they stopped it because they would have had to recruit so many more patients than they initially planned to show the reality of what was a very, very small, but nonsignificant, benefit, that they just quit. So I know that a lot of people, even in our institution, do hypothermia. I'm not sure that it's something that if it's going to cause a problem to some other part of the body that is going to be very easy to defend. And then the anticonvulsants. I should probably let Dr. Gorman talk about this. But we use the anticonvulsants. We generally think of them as benign, which in some situations they are not. But again, we don't know if they alter the course of the recovery from the injury. And then there's all these things about moderate traumatic brain injury that may or may not work, dark, quiet room; brain rest. That's empirically used. But I can't think of a pathophysiological reason why the brain would want to not work, other than to make itself not work by being sleepy. But if you're totally wide aware in a dark room, one could argue that maybe that makes you a little more anxious, not very good for your brain. And then the body rest. And then promoting sleep. All of which we do. In fact, I even prescribe it for my patients. And say, well, go home and rest and all that. And usually the kids feel better after a while. But there may be a better way. This thing about when return to activity when symptoms resolve. So I think that's a totally open question. And I think that we need to remember that when we're clearing kids to go back to activity. That just because the headache went away, we don't really know when those repair mechanisms, that can spread for a long time, because of the way the axons have to be repaired, the way the establishment of those brain network sometimes has to be redone. Second impact syndrome, very interesting, very unusual, very few reports of it actually happening. Does it really exist? But we all talk about it. I talk about with my patients. So you're at risk. For the next hit, it's going to be worse. But we have no way yet of quantifying that and whether the concussions are cumulative. So these are things that we all talk about. And there's protocols, even in our institution, to say we should go over this with the patients and their families. But the data behind it is incomplete. And I think that while we're saying this, we probably have to think back to what's behind this pathophysiologically. OK, I think I get two minutes of questions. Thanks very much. [APPLAUSE]
