So we're going to get started with Dr. Fagagy, who is joining us from South Africa. Thank you. Thank you very much to the conference organizers for the very kind invitation. My talk is on individualized care for TBI. My primary disclosure, I don't have any commercial disclosures, my primary disclosure is that I'm not a big fan of guidelines and hope to say in the course of the next 15 minutes a few things that are somewhat controversial and hope to say them quickly so that we have time for questions and Frank is going to do a much better job interrogating the guidelines a little bit later. So that idea of personalized medicine for me is probably best exemplified with antibiotics and with oncology. So antibiotics, I started many decades ago with one antibiotic. We never approach a complex patient with an infection or sepsis with one antibiotic. So we go through a fairly rigorous process and we do it fairly routinely to find out what exactly, which precise antibiotic is right for that particular person. And oncology, particularly in the last five or ten years, has really exploded in terms of personalized meds and there are two aspects of that. One is to find a drug, a chemotherapeutic agent that's targeted not against the whole group of tumors, histological group of tumors, but against a particular biological pathway in a particular patient. And that also has implications for various protocols. So if you look at medulloblastomas, we know that they are broken up, even though histologically they may all look like medulloblastomas, they're broken up into four different groups. And each of those groups have a very different prognosis. And arguably, we need to be exposing them to very, very different treatment protocols, such as chemotherapies. And I think it's this kind of thinking that somehow we need to bring to bear in TBI care and we just haven't got it right. And all our guidelines are still focused on what I consider as average patients. So this is what oncology looks like, you know. So at the top over here, these are all medulloblastomas, but they're different groups. So when tumors are completely, completely different groups and they've got a great prognosis, almost 100% of these patients will survive, except if we expose them to toxic treatments. And we may be able to skip radiotherapy in them. It's completely different from group 3 and group 4 medulloblastomas, but they're still the same tumor, theoretically. And the oncology group, certainly in neuro-oncology, has this approach to histology. So there's a big battle between the biologists and the pathologists, the pathologists digging in their heels, still thinking that pathology rules, and biologists says that the age or the era of pathology-based clumping everybody together in one group is over and it doesn't do patients any good. And I think our community needs to go through the same process for TBI care. So we don't have too much time, but I'd like to make these three statements, or one statement to three components. I think we can extend it to many different aspects of TBI care, but I don't believe there's a standard ICP threshold. I don't believe there's a standard CPP threshold. I don't believe there's a single indication for decompressive craniectomy, even though the guidelines that we churn out every year or every few years would lead us to believe that these things are true. So this is a paper that we've just discussed so many times over the years, and it's become a controversial paper, and I've waxed and waned about how I think about it. Currently, I think this is one of the best things that could have happened to a neurosurgical community, however you read it, however you interpret the outcomes. But it's a shot across the bars to say that TBI care must be more than ICP monitoring, and our approach to ICP monitoring must be more than thinking about it just as a binomial above or below 20 millimeters of mercury, and I'll try to justify that. So this is the guideline, and Frank is going to talk about it. But I think the idea of moving the threshold up from 20 to 22 millimeters is meaningless. So this is how and why we want to treat intracranial pressure, because we know patients herniate and patients become ischemic because of high intracranial pressure. And you see that as intracranial pressure goes up, it decreases all the indicators of perfusional metabolism of the brain, and that's the reason that we treat intracranial pressure. And for that reason, it would seem fairly straightforward. But in the Chesnut trial, the one alternative to ICP monitoring was not ICP monitoring and just examining them. And that's also a little bit of a problem, because when you lighten patient sedation or when you do a vigorous examination to determine their GCS, all hell can break loose in terms of their intracranial pressure. So that's not good either. But this also happens. So intracranial pressure has complicated relationships with the physiology of the brain, depending on the cause of intracranial pressure. And there's not just one cause of intracranial pressure, and there's not just one underlying metabolic activity of the brain. So therefore, the relationship between intracranial pressure, and not just brain oxygenation, but indeed any measure of perfusional metabolism, is much more complicated than we would think. And we see this as well. There are complicated reasons whether these are spreading depolarizations or subclinical seizures. These patients don't need a decompressive craniectomy. They need an addressing of the underlying problem. So we can do a decompressive craniectomy, of course, and we know that it works. And Professor Hutchinson is going to tell you that it does work. But it works to decrease intracranial pressure. And we're all very, very satisfied that that's exactly what it does. But the reason that it doesn't manifest in better outcomes in many of our patients is that the indications can't just be intracranial pressure alone. And this is primarily why. And if you took home one slide, it's probably this. It's intracranial pressure against brain oxygenation in this case, but you can put anything on the vertical axis. You can put any measure of perfusion or metabolism, and you will not get a linear relationship with intracranial pressure because such a linear relationship just does not exist. And it doesn't exist, of course, because it's pathophysiologically diverse. It is not one thing. And even if you did treat intracranial pressure, and you normalized intracranial pressure, and you had your normal perfusion pressure, still in our experience, about a third of our patients still have not only low brain oxygenation, but very, very low brain oxygenation below 10 millimeters of mercury. So not only is intracranial pressure complex, but even if intracranial pressure is sorted, it doesn't mean that everything in the patient is sorted. And so that's why I'll just skip that. And so that's why I put this here for David. So this is a great work that they've been doing with the BOOST trial. There's a suggestion, of course, that if you use brain oxygenation, patients may end up being treated better than if they were being treated with ICP alone. A little bit ambivalent, as much as I use brain oxygenation in my practice routinely, because brain oxygenation, just like ICP or just like cerebral perfusion pressure, again, is a monitor. And it's never going to be the monitor that benefits patients. It's going to depend on what you do about that monitoring data. Similar with cerebral perfusion pressure, you can delude yourself in thinking that there's a perfusion pressure that you can use for all patients. But the reality is, when you take autoregulation into consideration, this is impaired autoregulation. So I'll show you four slides, two with blood pressure going down, two with blood pressure going up, with completely different relationships with intracranial pressure. So if you just focus on the top two, this is impaired autoregulation. As mean arterial pressure goes down, intracranial pressure goes down. It's a passive relationship. But this is intact autoregulation. As blood pressure goes down, the normal autoregulatory response is vasodilation. So you vasodilate, you increase cerebral blood volume, and you increase intracranial pressure. This is the same patient on two different days. So if you had the same perfusion pressure threshold or strategy, not only across two different patients, but even in the one patient, you'd be doing them a disservice because the impact that you'd be having on intracranial pressure would be completely different. And so this is increasing blood pressure, intact autoregulation. The normal response is vasoconstriction, so you're reducing cerebral blood volume, and therefore intracranial pressure comes down. That's what you'd predict. But again, if impaired autoregulation exists, then the relationship with intracranial pressure is completely different. So it makes absolutely no sense, knowing that you have at least a third of your patients with impaired autoregulation at some point, to have a protocol that says target the perfusion pressure of 60 or 70 millimeters of mercury. Same is true for ventilation, and I'll just go through these very quickly. I don't want to show you all, but we know that CO2 has an incredibly strong and robust and usually retained effect on cerebral blood vessels in TBI. So that's why hyperventilation became popular. So these are CO2 challenge tests with intracranial pressure and brain oxygenation and cerebral blood flow using hemidexin. The only message that I want to get across to you is that you can hyperventilate patients or you can hypoventilate patients, and in both of those, you can have completely different effects on perfusion of the brain. The effect on intracranial pressure is generally quite robust. So if you hyperventilate patients, you reduce ICP. If you hypoventilate them, you raise CO2, you're going to increase intracranial pressure. But the effect on brain perfusion, which you're much more interested in, depends on the prevailing ICP. So if you look at them on average and you hyperventilate patients, this is what you'd get in cerebral blood flow and brain oxygenation. But again, it's average, and we don't treat average patients. If you look at individual patients, so this is a hyperventilation test. Brain oxygenation and cerebral blood flow comes down because his veins are constricting. A different patient, you hyperventilate them, but despite the fact that your veins are constricting, and this is because of the complicated relationship with intracranial pressure, the brain perfusion actually goes up. And so the problem, again, is if you treat this patient like an average patient with no knowledge of what you're doing to perfusion of the brain, you're going to do that patient a disservice. So how does one make sense of all this? And I want to try to end very quickly, just in case you have any questions. My approach is to use as many monitors as we have available, but with a provisor that we know what we're doing. And the problem is it's a complex problem, but the answer is not just to throw a whole bunch of monitors at these patients, because the monitors aren't going to benefit the patient. Judicious application of the monitors and of the data and trying to use the monitoring data to figure out what is pathophysiologically going on in an individual patient so that we can target the treatment according to the underlying pathophysiology, for me, makes a lot of biological plausible sense. I think the chances that we're going to have grade class one data to support this in the future, I think, is really, really small, because this is very, very difficult to study. And the various studies that have been done, the randomized controlled trials in TBI over the years, have served to prove to us, mostly, how weak our methodology of randomized controlled trials are. So microdialysis is one opportunity. It's a lot of work, though. It's expensive. You've really got to spend time by the bedside thinking about what's happening to lactate, pyruvate, and glycerol, et cetera. So it's not business as usual. So it's not just a simple matter that we thought it was of intracranial pressure 25 or 15. It's much, much more complex. And how one puts that into a protocol, and even with greater difficulty, how you put that into a guideline that is robust, I think, is a tremendous challenge. So the take-home message for me is I think we need to overhaul our thinking, and therefore also our recommendations about TBI care. I think modern-day decision-making depends on better information and better application of that information. And I think we need to forget average. An era of precision medicine looms. So for me, it's the very same way of thinking about penicillin for infection. You can't use penicillin for infection. Same way that in 2018 and 2019, we're not going to have one treatment for medulloblastoma. We need to move on with the times, and I just think in the TBI care community, we haven't moved on terribly much in terms of a philosophical approach for TBI care, but I think the time is now. Thank you. And greetings from Cape Town. We have time for questions. So I'll start. Walk me through your thought process to the patient without an intracranial mass lesion who doesn't need an emergent neurosurgical intervention but is admitted to your ICU in coma. What is your first thought process when you think to yourself, I want to personalize the care? How do you structure that? So again, I'll start with monitoring, and for better or worse, a threshold for monitoring is patients in coma. So it might not be true that those are the best patients to target, and it might be the patients with moderate TBI that have the potential to deteriorate that we should be targeting. But for better or worse, that's our threshold, and that remains my threshold. My routine is brain oxygenation and microdialysis to add to ICP monitoring, and then to adjust my thinking about whether the intracranial pressure is a problem or not based on how robust the rest of the monitoring is. If I want to treat ICP, then hopefully those are the monitors, and this also includes continuous EEG. Hopefully my thinking is also modified by what's happening with those other monitors. So if ICP is 25, the scan looks okay, and the brain oxygenation looks okay, and the microdialysis looks okay, I'm not going to take them for a decompressor craniectomy. But by the selection criteria for our randomized control trials, and no disrespect to the great trial that Peter Hutchinson did, we're still selecting patients of GCS and ICP. And I think that's crazy. So two patients, one who is a young patient with a GCS of 12, but is deteriorated to 4 with high ICP, is a very different patient from an older man who had a GCS of 4 at the scene. And those are completely, completely different patients, but for our randomized control trials they're the same. Right. Tony, I have a question for you. I come from a developing part of the world, Detroit, and I would like to know what your vision is for precision treatment of TBI patients in the developing world. How do you think this will come about? Is it going to be evolutionary? Is it going to be revolutionary? Is it always going to just be too expensive for developing countries? Oh, man, that's a tough question. I'm much better at posing questions than answering them. I think the principles are generalizable, even if you don't have monitoring. I think the developing world circumstance is very difficult because it's massively diverse. And so there could be – so we had a consensus meeting about decompressive craniotomy that Peter very kindly organized in Cambridge, and we had a developing world session. And one of the things that came out in that is that the circumstances are just so different. So on one end there's an argument, well, why don't you decompress everyone? Because you don't have money for ICP monitors, and it's going to reduce the number of days that patients are ventilated, and in the ICU you know that because that's robust data. So that would make sense as a good cost-effective strategy. On the other hand, the exact opposite argument can be made that if you're going to do that for all patients, you're going to increase your rate of severe disability, and in a context where people don't have the money to be looking after disabled patients. So I think it's enormously complicated, and I think you'd have to organize a symposium for a week just on that one topic. And that's my way of getting out of the question. Thanks. Questions for me? Don't be shy. We're here to hear your comments and questions. Okay. Well, thanks, Tony. Thank you.

individualized care

traumatic brain injury

personalized medicine

multiple monitors

treatment decisions