Hello, everyone. We're here for cranial tumor section to pop the top or not with Dr. LaRue. If you've been to a prior APP course, you've probably seen him there. He always gets rave reviews. We're happy to have him back. He received his medical degree and doctorate neuroscience at Cape Town, South Africa. He later trained in Washington, Seattle. He has been in clinical practice for more than 25 years, and he has a long-steaming record of translational research and brain resuscitation, recovery, and is an internationally known expert in subarachnoid hemorrhage and traumatic brain injury, as well as neurocritical care. And he's edited several textbooks and given more than 350 invited lectures in 15 countries and published more than 250 manuscripts and peer-reviewed journals. So without further ado, here is Dr. LaRue. Well, thank you for that introduction. It's a delight to be back. I've always said I prefer working with PAs and nurse practitioners than I do with residents because I often would experience no more and always continue and eager to learn. The topic at hand today is decompressive craniectomy. This was the title given to me by the AANS, to pop the top or not. And there's evidence and there's clinical experience and the ethical challenges in deciding about this. It's important to recognize that traumatic brain injury is common. Probably 2.5 million people a year in the United States, about 50,000 of those die. There's more hospitalizations with TBI in the United States than there are new cases of breast cancer or HIV. Important also to recognize that this is a global issue. So about 70 million people a year have a traumatic brain injury worldwide and it accounts trauma does for about 9 or 10% of deaths. And they're becoming more of an issue as we see changes not only in demography, but also in how people conduct their lives. Craniotomy is not that commonly performed in traumatic brain injury in part because 80% of TBI is mild. It's about 3 to 5% of people actually undergo a craniotomy. It has to be recognized that there are different surgeries in traumatic brain injury, including that brain monitoring, CSF diversion, repair, or skull fractures. Obviously what drives much of what we do is evacuation of hematomas. Extradural, subdural, intracerebral contusions, and they are well-described guidelines that exist for that. What we're going to focus on is decompressive craniotomy, and that can be divided up into primary, secondary, or rescue. So primary is performed at the time you evacuate a hematoma. We do not have great evidence for that, and there's an ongoing trial rescue ICP for acute subdural. Much of what is done is really based on the judgment, and it's a subjective judgment by the surgeon. Is the brain swollen or not? What we do have evidence for is secondary decompressive craniotomy. That is somebody in which there's a change in condition over time, and now I'm selected to perform a decompressive craniotomy. The two primary clinical trials are DECRA and rescue ICP, and they have two real primary differences. DECRA was for an increase in cranial pressure done early, and actually the trial only recruited patients up to 72 hours. Rescue ICP was for patients who had raised in a cranial pressure that was refractory to treatment and recruited patients up to 10 days. Rescue decompressive craniotomy might be something you see in the military setting where patients with a blast injury and craniofacial fracture may have a decompressive craniotomy just to allow them to be transported to sites away from the front line. Blast injuries are very different pathology, and so rescue decompressive craniotomy might apply in the military setting, but less so in a civilian environment. There are questions though, right? The big question is, should we be performing it, or do we end up saving a lot of lives and leaving patients severely disabled? That then begs the question, who should we perform it on and when? What's the ideal technique, and how do we decide what defines outcome? The premise behind this is all about intracranial pressure, so it's important to sort of understand that and the Monroe Kelly hypothesis. The primary insult leads to brain swelling as does a secondary insult, and that leads to an increase in intracranial pressure which cycles through this reduction in blood flow, energy dysfunction, further brain swelling, and it's like a vice grip is occurring on the brain. The premise is then just to release the vice, expand the suitcase, give more space to the brain. Well, if we want to talk about intracranial pressure, we have to understand the signs and symptoms, and they vary depending on how high intracranial pressure is. So early on, if a patient has headache, they may have nausea and vomiting, they may become irritable, and it's only as you see a progressive increase if you observe a decline in consciousness. Now that's all very well said, but certainly in high-income countries, the United States and Europe, the vast majority of patients with severe traumatic brain injury, defined by a Glasgow Combs scale less than eight, are ventilated and sedated, so we don't have a good clinical exam. So what predicts the likelihood of raising intracranial pressure? There's a general consensus that reliable predictors based on CT include effacement and compression of the basal systems, a midline shift more than five millimeters, and a very large mass lesion defined as more than 25 cubic centimeters. There are some other minor criteria which might be predictive of the likelihood of developing raised intracranial pressure, perhaps are less reliable. This is perhaps the most predictive. Here is a patient, slide A, you can see the perivasal systems are wide open, and here in B, they are obliterated. So this patient in B is likely to have raised intracranial pressure if it is their measure. This is not necessarily 100% sensitive and specific, and so relying entirely on CT may not be sufficient in many patients. Well, does it have an impact on outcome? It does, but it has an impact on mortality, not functional outcome, and that impact is greater the greater the intracranial pressure. So as you can see, the odds ratio of death when ICP is between 20 and 40 is about 3.5, it's close to 7 when it's more than 40, but if refractory treatment, that odds ratio is more than 100. It is important to recognize that all of these studies that look at outcome include patients who have been treated. So we don't really have a natural history study because nobody's going to do a study where you just ignore the ICP and see what happens. But no question, elevated ICP is associated with mortality. The higher mortality, the worse the outcome. Well, our beliefs got challenged by this trial, which was published in 2012, so nearly eight years ago. Randy Chestnut, a very smart neurosurgeon, did this study in Bolivia and Ecuador with NIH funding. It could not be done in the United States because people were not prepared to do the study. You have to have clinical acropoies to randomize the patients, and the title is not actually what the trial was. It was not a trial of endocrine pressure monitoring. It was not a trial that compared different methods of treatment. It was a trial that compared ICP treatment using different ways to identify ICP. And not surprisingly, the outcomes were very similar. And this created a furor in the literature and a great debate because it questioned everything that was the premise and principle of trauma care, which was management of ICP. And to the investigator's credit, here's Randy Chestnut, principal investigator Tom Black, who was the first president of the Neurocritical Care Society, convened a meeting of so-called experts to decide what does this mean. And here you can see this crude methodology we used to sort of identify what does the BESTRIPS trial mean. And these were some of the conclusions that that group made. The primary impact is research-oriented. It should not change our management, number one. Number two, it was not a trial of ICP monitoring. It was really a trial of two different strategies or protocols for how you manage patients, and it had very little external validity. As I said, this trial was performed and believed to be performed in Bolivia or in Ecuador, where the vast majority of patients with severe traumatic brain injury arrive in a taxi and do not go to rehabilitation facilities following their treatment. And hence, those results may well apply in those countries, but may not necessarily apply in an environment such as the United States or Europe or other countries that it might be classified as high-income versus middle or low-income countries. But this is what was recognized as the number one research priority. What defines intracranial hypertension? And that might seem crazy, that we're treating something that we actually don't know what it really means. But as to recognize that the definition of raised intracranial pressure was an ICP greater than 20, was based on the evolution of clinical experience to the 1960s and 1970s, so pre-CT, and it really was based on the concept of the break point at which you would see a shift in order regulation. And at that point in time, the vast majority of patients with traumatic brain injury were run dry. The idea was don't give them fluid. So, we had these real issues with perfusion pressure, which today we don't have. Now, as you know, the recent guidelines that were published a couple of years ago had this ICP number of 22. And that was sort of based on one or two small clinical series. And that has been questioned as well. And it's questioned in part because a variety of other technologies, whether you use microdialysis or PET or brain oxygen or MRI, demonstrate that there is a shift in brain pathophysiology long before you see an increase in intracranial pressure. This has led to the concept of using more than an ICP monitor to manage patients. So, the concept of multimodality monitoring, the different ways of defining it, the different tools you can use. Ongoing now is BOOST-3, which is a very large phase three trial that's comparing outcome in patients who have an ICP monitor or an ICP and brain oxygen-based monitor. The phase two trial, which was not powered to define outcome, did show a shift in more favorable outcome when you use more monitoring tools. There is great variation in care. So, this is from CENTER-TBI, which is a European-based study. And you'll see here in the center that the number of patients actually get ICP monitoring is fairly infrequent. And between centers and between countries, there's great variation. That variation is the very basis for CENTER-TBI. It's using the variation to helpfully tease out which treatments work better. Having said that, using an ICP monitor and treating ICP seems to make a difference when you look at a variety of other methods of analyzing this data. This is from Sherman Stein, who did a meta-analysis that included 120,000 or more patients. And he defined this as high-intensity aggressive therapy and low-intensity therapy. The difference between these two was you used an ICP monitor to guide your management. In these patients, you saw a significant improvement both in mortality and an increase in the number of favorable outcomes, which was not time-dependent, so not simply based on the historical differences that occurred with time. A variety of other studies have evaluated this, because what really guides what we do are our guidelines, which have evolved over the years. A lot of what we base our care on in the guidelines is the management of endocrine oppression, cerebral perfusion pressure, and the avoidance of secondary cerebral insults. Our targets, in large part, are population-based and not necessarily individualized. But there have been several studies that have used guideline adherence to look at how it affects outcome. These being single-centered prospective observational studies, statewide administrative databases, national quality assurance programs, you have the American College of Surgeons, and then a pre-post-guideline implementation program. And all demonstrate that when you use an ICP monitor and base your care on that and manage them, patients do better. Now, many of these define outcome based on mortality alone. It's important to recognize that guideline adherence has a ceiling effect. You hit about more than 80 percent guideline adherence, you start to actually see an increase in mortality. And this emphasizes the importance of one size does not fit all patients. In other words, you need to individualize your care to most optimized patient outcome. So when you look at endocrine oppression, and this in large part was driven by conversations and research since best trips, is the number itself may not necessarily be the target. Or if it is, that's a simplification of a complex process. But that ICP should be looked at as an indicator of something else is happening, and trying to identify and understand that pathophysiologic process is important. And there are a variety of issues that that can be a manifestation of, and those are treated in a different way. So as you start to treat ICP, it might be that we don't have a specific numeric threshold, but rather use that number as a guide. And you can look at this patient, who is an ICP of 22, and three different approaches apply despite the same number. Here the patient who's got normal pupils, a normal waveform, who if you stimulate him is a rapid recovery in ICP, normal systems, I might not do anything. On the other hand, this patient here with an ICP of 22, same ICP, but pupillary abnormality, an abnormal waveform, stimulation leads to a sustained increase in ICP, a lot of midline shift, a very different approach. So that's important to recognize as we discuss decompressive craniotomy, the role for individualizing our care. This has been taken further with multimodality monitoring, where you can define the optimal perfusion pressure in this case, and the optimal ICP for an individual patient. Not surprisingly, as in many aspects of biology, there's a U-shaped curve. You want to try and keep the patient in what's optimal for them, and what's been found with ICP, if you use an index called PRX, which is a measurement that correlates ICP, it's a moving correlation between ICP and blood pressure. If you use PRX, so individualized ICP versus a dose of ICP more than 20 or 25, your relationship with outcome is far more robust using individualized targets than population-based targets. We have seen evolution of care, and this is perhaps the most recent algorithm for how to treat raised intracranial pressure, came from CBIC, which was a Seattle International Brain Injury Consensus Conference. What is utilized here are these tiered effects. These tiers are something that you see in BOOST-3, and the tiers were used in BOOST-2, which is the trial that looked at multimodality monitoring. The idea behind this is you don't have to do all these treatments, you do some. They really are suggestions to do, and you might do one or two and find out the patients refracted the treatment, you move to the next tier. Decompressive craniectomy pops up as a third tier of treatment. Using non-ICP-based management is also being codified through what's called ICE, which is the imaging and clinical examination. This was used in the BEST-TRIPS trial, and this is all about examination and CT, and you decide how to base your care. The end result is actually the same. You're trying to really manage ICP based on different signals for that. CBIC has also suggested or recommended treatments that you should not use when you're monitoring ICP, and only monitoring ICP. This differs a little bit if you're using other modalities. There's no role for a continuous infusion of mannitol. There's no role for scheduled infusion of hyperosmolar therapy. There's no role for LASIX, lumbar CSF drainage, steroids, routine hypothermia, routine hyperventilation of trying to get the CPP more than 90. No role for that at all. What about decompressive craniectomy? Well, it's used in stroke. None of us argue about the role of a decompressive craniectomy for posterior fossil stroke. These patients can be very sick. You do this, and it's like Phoenixerism. They get better. There was a study that was, it's GASCUS, which is the German-Austrian space occupying cerebral infarct trial that did demonstrate efficacy, but it was really dependent on how sick the patient was. So, again, emphasizing the importance of patient selection. Decompressive hemicraniectomy is described for stroke, specifically malignant middle cerebral artery stroke. This probably represents about 10% of stroke patients. Without aggressive treatment, mortality is greater than 80%. These patients have an NIH stroke score of more than 25 on average. It might vary a little bit, but it's a very high NIH stroke score. Remember, that range is between 0 and 42. It was analyzed, and this data was published in 2007. This pooled three studies in Europe, HAMLET, DESTINY, and DECIMAL, and what they demonstrated, you can see here, a significant survival and favorable outcome benefit. Now, important to recognize that these were three ongoing randomized trials. The investigators agreed to pool their data before they finished their trials. So, this pooled analysis included, if I recall, 93 patients, which is a very small sample size. Hence, one has to interpret that data somewhat cautiously, right, because a lot of stroke outcome data is defined by a modified Rankin score between 0 and 3 or between 4 and 6. So, between 0 and 3, it's favorable. 4 to 6, it's unfavorable. What the pooled analysis demonstrated is that if you shift that modified Rankin to 0 to 4, you see this improvement in outcome. It also then raised the question in that physicians, patients, and family members and investigators define outcome differently. So, for example, if I had a patient who was dependent on many of their physical activities, i.e., perhaps bedridden, maybe able to walk, but couldn't do anything else, at least dependent on others, I define that as an unfavorable outcome. Some patients and family members, however, define that as favorable if that individual can communicate. So there are differences of opinion, and this is important as we decide about decompressive craniectomy. Individually with patients, and as we interpret outcome data. There have been a variety of meta-analyses, and even meta-analyses of the meta-analyses, to look at the role of decompressive hemicraniectomy in acute ischemic stroke. It is important to recognize that this is different from traumatic brain injury. You've got an isolated vessel therapy that's injured versus the whole brain. But there have been several trials. Decimal, Hamlet, Destiny 1 and 2, Head First, which was in Canada, a large Chinese style, and a single center trial in Latvia. There are some other trials from the Philippines and Turkey that are smaller, and I don't think the Turkish one has been published. But this meta-analysis here included 338 patients. So, slightly bigger sample size. What it did demonstrate is lower mortality and better favorable outcome. So, they are recommendations that exist from the Neurocritical Care Society, and this was done in combination with the German Society for Neurointensive Care, that decompressive hemicraniectomy is recommended with a strong recommendation, high quality of evidence, for malignant middle cerebral artery infarction. That recommendation may not be as relevant to the older patient, defined as greater than 60, where their patient and family input becomes more important. Ideally, the decompressive craniectomy should be performed within 20 to 48 hours, before herniation occurs. Going beyond 48 to 96 hours, it does not seem to be a benefit. But this remains an area of contention. When to do it? How do you predict who's likely to get it? And there certainly are certain criteria. For example, an MRI, a DWI volume of greater than about 140 mils, suggests you should do it independent of the clinical evaluation. What about a traumatic brain injury? It has a long history, with periods of enlightenment and dark ages. It goes back to the Neolithic period, 10,000 B.C., is evidence for trepanation. Perhaps the first time this was described as a procedure was in 1518 in Italy. The first time it was used to treat a cranial hypertension was in 1901. But then it kind of slipped out of favor. Cushing used it to treat brain tumors, as brain fungus, as they call it, to allow the brain to expand out. But there was a period, as ICP monitoring started, that there was one or so paper a year published. Now it's almost one a week. The mortality was very high. And so it sort of fell in disrepute, and it started coming back into favor again in the early 1970s. Schellberg described a bifrontal craniotomy. Ransohoff, a hemicranioctomy. And then Paul Cooper and Ransohoff later described it as a very clear second-tier procedure. And even then, the outcome was not very bad. One has to recognize, at this point in time, we did not have very efficient ICP monitoring. We used ventricular drains, something like a Richmond or Charlottesville bolt, which was really tough to insert. Didn't have caminos at that point in time. We really didn't have good CT scanning. In fact, we didn't have CT scanning in 1971. The quality in 1979 was horrible. Took you 45 minutes to get a head CT versus 45 seconds now. And so it fell in disrepute. It was sort of a dark age, and then it resurfaced as these observational studies sort of repopularized it, defined it, and that then led to the randomized trials. So if you follow this data through the 70s and 80s, you'll see that the mortality is greater than 50%. Now you get into the 1990s and early 2000, the mortality is 20%. But more importantly, you started to see a shift in outcome where more people were getting a favorable outcome. So this then drove randomized trials. The first one that was published was DECRA. DECRA came from the Australian New Zealand Intensive Care Society. They are a great clinical trials research group. They do a large number of randomized trials in critical care and do them well. And they elected to do early decompressive craniotomy in patients with severe diffuse brain injury. So these patients have cerebral swelling. They all got a bifrontal craniotomy. The data showed that the patients at six months with decompressive craniotomy did worse. And this, again, just like BestTrips did, led to an immense amount of dyspepsia in the trauma field. This was the response, one of almost universal criticism. And most people said, you cannot make any conclusions about this trial, and it should not change clinical practice. And there were several reasons for that. First was 4% of screened patients were enrolled, and they enrolled 155 patients. Again, a small sample size. Most of the patients came from Australia, New Zealand, the Middle East, Asia. At the time where I was working, we were the only site in the United States in DECRA. And in five years, we could not recruit a single patient because we didn't find patients eligible. There were differences. 27% of the patients that had decompressive craniotomy had bilateral fixed and dilated pupils versus 12% of the medical. A third of patients had open cisterns, which one might argue means they didn't have raised cranial pressure. And this was the key issue with DECRA. The decision to do or randomize patients was that if you had an ICP greater than 20 for 15 minutes in a one-hour period, you were now randomized and assigned barbiturate coma or decompressive craniotomy. And not surprisingly, when you look at the pool data, the high ICP was actually very short-lived and not very high, came down with standard care, came down even more with decompressive craniotomy. These patients did not have intractable intracranial hypertension. A second issue was the technique. So these are the frontal lobes on either side. This will be the superior sagittal sinus. This is a cruciate incision. And those cruciate incisions don't really allow for the brain swelling to swell out of the brain, which is the idea behind a decompressive craniotomy. And we know that technique has a significant role in outcome. So what did DECRA really show? It showed us unequivocally that before the development of intractable intracranial hypertension, it doesn't have a role or rather decompressive craniotomy doesn't have a role because now you're exposing patients to potential surgical morbidity. One has to recognize that medical management also has morbidity. Barbiturates are not a simple thing. They're associated with several complications. And so before decompressive craniotomy, you really should escalate medical therapy or other forms of management before the procedure. In addition, given the methodological limitations, it does not rule out the possibility that some patients will benefit from this if you have maximized medical therapy. Another criticism of DECRA was the outcome was looked at at six months. And we know that people get better with time. And recently the group published their 12-month outcome and there the outcome data show identical or very similar outcome. In other words, there's no difference between the management strategies. This meant we relied a lot on what the next large randomized trials were. This was RESQ-ICP. Principal investigator was Peter Hutchinson who's a neurosurgeon at Cambridge. It was one primarily out of Europe. It was a multi-center trial. Seventy percent of the patients came or were recruited in England. Here you could do either a hemipraniotomy or a bifrontal craniotomy. Now interestingly, about 60% of the patients had a bifrontal craniotomy. In my experience, I've hardly ever performed that. In the United States, most patients get a hemipraniotomy. You could also do this when there was hemorrhage, whether it be a subdural, contusional, ventricular hemorrhage or a little different from DECLA. And here the patients were recruited if they had intractable intracranial hypertension. It was not responsive to treatment and the recruitment occurred up to 10 hours. 408 patients were included and randomized in this trial, so a larger sample size. The principal hypothesis was that in patients that raised in refractory ICP, you'd see improved outcome compared to optimal medical management. So you had to go through these stages. You had to do initial stage one treatment, stage two treatments, where you could not use barbiturates. And only then, if you had intractable intracranial hypertension, were you randomized. You still had to continue your medical treatments. What they found was an improvement in outcome, albeit sort of hidden among the data. So fewer patients had, or rather, there were fewer hours of raised intracranial pressure with decompressive craniotomy. I think the median amount of time with surgery was five hours versus 17 hours with medical management. You went out of hospital sooner, 15 days versus 28 days. So efficiency of care was improved. Your mortality was improved, 27% versus 49%. But what happened was you saw this shift in what was called the upper severe disability. All right? You were more likely to have less of that. And that was a difference of between about 13% and about 4%. But that was statistically significant. Now, severe upper disability means the patient is independent at home, but not outside of the home. So the implications of rescue ICP, well, it will help patients. Not only mortality, but you see some improvement, albeit difficult to tease out, in functional outcome. There were many implications from this trial. We really need to understand how to select patients. We need some kind of decision-making tools. We need to engage patient surrogates because how quality of life is determined varies with patients. Let's just look at those differences between DECRA and rescue ICP. DECRA enrolled 4% of patients and included 155 patients. Rescue ICP enrolled 20% of their screen patients and included 400 patients. A large number of patients in DECRA had bilateral unreactive pupils, universally recognized as a bad prognostic sign. These were excluded from rescue ICP. You could have a mass lesion. This was simply diffuse edema. The key difference is this was a trial about early ICP, not refractory ICP. This was about refractory ICP. There was crossover that occurred in both studies, which does limit sometimes how you can make conclusions. Well, those trials have led to a lot of thought about how can we select patients because that's the key issue. Who should we be doing this on? What's associated with the outcome, not unexpectedly, the worst in neurologic state, bilateral fixed dilated pupils, higher ICP tend to do worse. Technical issues make a difference. Smaller bone openings and bifrontal craniectomies tend to do worse. It's important to continue to monitor the patients afterwards because what we do see is that sustained intracranial hypertension or a failure for ICP to decrease is associated with bad outcome. You might be able to define this with CT and a combination of factors, younger age, better clinical condition to start with, larger bone flap and a low ICP after surgery are associated with better outcomes. As I said, here are the two techniques, a hemicraniectomy or a bifrontal craniectomy. They are very distinct technical considerations. You have got to get back to the coronal suture. And quite frankly, those cruciate incisions that you saw in Decker don't work. You have to divide the folks up front and let the juror flap back. This decompressive hemicraniectomy is associated with better outcome than bifrontal. When you look at a variety of clinical studies, you must recognize most of those are retrospective observation studies. But you do better with a hemicraniectomy than a bifrontal craniotomy. That may be an epiphenomenon based on indications. Whatever you do, you have got to do the procedure right. And importantly, the bigger the craniotomy, the more effective. So you need to have a diameter of the craniotomy that is probably close to 15 centimeters in diameter. That is a big opening. Those small openings don't work. They make outcome worse. We have to recognize that like any other surgical procedure, there are complications. It's about 13%. And to put that in perspective, the complications that you see if you do elective meningioma surgery, that's about the same. So it's not unique to this procedure. But hematomas can occur either on the same side or elsewhere, wound complications, and fairly unique issues related to CSF flow dynamics, either surplus, dural collections, and the syndrome of the trephine. There are ways to prevent some of these. So here you can see a patient with a decompressive craniotomy and a large contusion. Well, he's going to have a bad outcome. He's going to be left with a dysphagia. Here's something that we see fairly frequently, this extracerebral herniation. Both of those can be prevented by doing a large enough craniotomy and using an augmentation duroplasty. So if you don't do a big enough craniotomy, you hurt the patient. So technique is important. CSF fluid dynamics can be effective. So a large hygroma can evolve, a sunken skin flap or the syndrome of the trephine, and hydrocephalus. Again, technique is important. A large craniotomy with an effective duroplasty, augmentative duroplasty, can limit but not eradicate these. There is some data that suggests that these also may depend on how close you get to the midline. So if you're within two centimeters of the midline, the incidence of these problems is higher. That may have to do with disturbance of the arachnoid granulations where a lot of the CSF transport occurs. Here's syndrome of the trephine. This is about 10% of patients. Look at the sunken bone flap. These patients manifest by decline in function, whether it be motor function, cognitive function, or consciousness. Now this lady had a decompressive craniotomy. Her bone went back. It got infected. It got taken out. Now you can see the sunken shape. You cannot put anything back in. So she's given a tissue expander, which is a breast implant, to try and stretch the scalp. These patients can do badly. You can see she's on a tracheotomy still. Can we prevent it? This is a still unanswered question. In all patients with decompressive craniotomy, so you look at stroke literature and TBI, the concept of an early cranioplasty does not change the incidence of hydrocephalus. But if you look at only the TBI patients, there is a suggestion that early cranioplasty reduces the risk of hydrocephalus. That, however, is not demonstrated in all trials, and this is the big question. Because you cannot look at decompressive craniotomy in isolation. You have to then consider the next step. That is the cranioplasty. So this is a bifrontal craniotomy, and you can see the cranioplasty afterwards. You can use the own bone. You can mold it at the time of surgery or pre-made. I like the own bone. I store it in a freezer. I will not use it if it's been stored more than six months. I do not like putting it in the abdominal cavity. I don't like making it at the time of surgery, particularly up front, because you do not get a nice cosmetic result. So I use a pre-made cranioplasty based on 3D reconstruction, and the cosmetic result is good as well. One has to recognize that this operation exposes the patient to other complications, somewhere between 5% and 25%. The medical complications are more frequent. The mortality is not insignificant. They're different types, obviously hemorrhagic, infectious. There's abnormal CSF flow, cosmetic issues, or seizures. Whatever the complication is, it aggravates the outcome. So you have a favorable outcome. It's rare that you have a complication, whereas those patients with an unfavorable outcome tend to have a lot more frequent major complications. What are the risk factors? Well, big implant size. Well, that's a bit of a catch-22, because we know you want to do it big to actually make it effective. Older age, bleeding disorder, infection, not surprising. An acute subdural is the etiology for your decompressive craniac in the first place. That may, again, represent what an acute subdural hematoma is. That is a horrible injury to the brain versus just the hematoma. Whenever you do a combined shunt and cranioplasty, so you're trying to correct abnormal CSF hemodynamics at the time you're reconstructing the cell, what you put in there doesn't matter. How or when you do it, we don't really know what it is. So the timing of surgery remains debated. And there are different ways of looking at it. Some people say within three to six months. Others say you can do it much earlier when they improve. Others say do it as quick as you can, because atmospheric pressure makes a difference. And there is no question that doing a cranioplasty does improve cognitive function and improves brain perfusion. Others suggest, well, you only have to do it if they develop the syndrome of neutrophil. Now, the effect of your timing may vary based on what it is. So done early, between 15 to 30 days, may reduce the risk of infection and seizures or bone flap absorption. Others suggest if you wait longer, you increase the risk of seizures. And as I said, this whole question of hydrocephalus is unanswered. Some say waiting a long time reduces the risk. Others say doing it early reduces the risk. Bottom line is we don't really know. We don't know because it's a complex problem. Both of these patients have hydrocephalus. But look at this. Sunken hydrocephalus, hydrocephalus with external herniation and large hygromas. As I said, it is important to use a shunt before you reconstruct. And a programmable shunt can make a big difference in these circumstances. Bone resorption is a problem that can occur. You can see the bone fragment. So I will store these in a bone bank in a freezer. Longer than six months, I will not use it. I will not use bone, prefer not to use bone that's stored in the abdominal cavity. Just to make sure you can think outside the box sometimes, decompressive laparotomy will also work in some patients. And this is another consideration. This may be applicable in the obese pigwickian patient or the patient with multiple abdominal trauma where you get a significant increase in intra-abdominal pressure, you can measure the folate and intrathoracic pressure. It works. Maybe it reduces the risk of craniectomy and cranioplasty. So, there's evidence, the clinical series, the ethical considerations. The big issue is, should we do it? Because those that advocate that it's not worth doing, say what decompressive craniotomy and traumatic brain injury does, is it increases survivors who remain neurologically non-functional with a poor quality of life. So what you do with surgery is you increase the likelihood of survival, but you don't necessarily increase the number of favorable outcomes. It's been studied in stroke, not in traumatic brain injury. So folks have been asked retrospectively, either the survivor or the caregivers, would you give consent knowing what you know? And most, 60 or 80% say yes, they would. This is despite many patients having a decrease in quality of life or depression. And this is important to recognize because the quality of life perception, that's ultimately what we want, are very patient-specific. So how a patient defines their life as worth living, whatever that means, is very dependent on their context, their experience, their environment, their cultural beliefs, their religious faith, even the medical economics of where they live, all have an impact. So we can use evidence-based medicine to guide us. And the definition of evidence-based medicine, which comes from Sackett, back some 25 years, is the judicious use of current best evidence to make decisions about individual patients. But that external evidence can inform, but can never replace individual clinical expertise. So in other words, we use evidence and individualize that to each patient. There are challenges to looking or obtaining evidence in TBI. We rely on randomized clinical trials. It's difficult to do that in traumatic brain injury, particularly in surgery, because there is a variation in clinical equipoise. As I said, best trips occurred in Bolivia and Ecuador, not in the United States, because people in the United States did not have clinical equipoise. There's an imbalance in surgical technique or expertise, and how you do the procedure makes a difference. There's difficulty with blinding. You know, boost two, all patients were gonna have oxygen monitoring and ICP monitoring, independent of what arm they were randomized to. And the trial was set to go ahead. And at the onset, when the investigators got together, they realized, well, that's gonna create a bias, because the nursing staff will see the oxygen data and change their treatment even if the patient's not in the treatment arm. So then they said, okay, well, let's just cover it up with some black tape. Even that was not regarded as good enough. So the manufacturers of Lipox came up with this little plastic thing that you could lock on with a lock and key. So you couldn't bias how people manage people, how they manage the patients. But perhaps most importantly is this great heterogeneity in how we care for patients and in the patients themselves. So we use all these therapies. These are some therapies that we use in severe traumatic brain injury. And hardly any of these have been really rigorously tested in randomized clinical trials. So a lot of what we do is not evidence-based in the first place, but based on clinical experience and understanding of physiology. So the big question in critical care trials is, do we even need evidence-based medicine? It's important to recognize that we in medicine are the only group that grade our inquiry. You read an astronomy journal or a geology journal or a physics journal, we're not grading the science. And what we do in evidence-based medicine is really confuse statistics with science. And that's largely based on simply one issue, bias avoidance, that's randomization. Now there's good in that, but that's what evidence-based medicine is, simply bias avoidance. It's not necessarily scientific, it's not necessarily physiologic. Important to recognize that that grading system is an opinion. And according to evidence-based medicine, opinion is the least valid of the evidence. And if you look at what's happened in critical care, very few trials have actually made a difference. This was a study a couple of years ago that looked at randomized trials in critical care, and very few of them, less than 10%, were actually positive. So most of what we do may or may not be supported from data from a randomized trial, but we really understand pathophysiology and then understand how what we do influences that pathophysiology. So there has been this question, do we need randomized control trials? This is an economist from Stanford who actually said that almost all randomized trials are simply a reflection of our prevailing bias and conventional wisdom at the time, since the vast majority of randomized trials, even the most cited and most reliable, have been challenged and refuted over time. So for example, a couple of years ago, we had intensive insulin therapy. We don't have that anymore, but that was demonstrated in a randomized clinical trial. We have this concept of what might be labeled practice misalignment. So for example, the TRIC trial, which was all about transfusion strategies, doesn't necessarily apply to traumatic brain injury, yet it's extrapolated. And so that is one of the issues we have to be careful about, that you cannot generalize the study findings beyond the group of patients it was before. So years ago, steroids were given to traumatic brain injury because steroids work in metastatic brain tumors. But now we know you should never give steroids. That's one of the treatments that have been demonstrated to hurt people in randomized clinical trials. Thank you. We also will not necessarily do some trials for certain conditions. This is a patient with epidural hematoma. Nobody's gonna do a randomized clinical trial. So do we have evidence that it works? Just as nobody's gonna do a randomized trial about whether parachutes work or not. There is a misalignment or an incongruency often between clinical practice and randomized trials. Andreas Kramer is a very thoughtful intensivist in Canada, did a big population-based study and looked at the patients in his system. I think this was in Alberta, Canada, who were having decompressive craniotomy. Less than 20% of the patients were eligible for ductal rescue. So it's then, as you can see, difficult to extrapolate that data into your clinical practice. The evidence-based medicine permit is dependent on background, animal studies, and expert opinion. It then drives us to the opinion, to the randomized trials and reviews of that. But take away this physiology. Take away what an expert thinks. The pyramid starts to fall over. So is there animal data? There is. White matter injury is synonymous with bad outcome in traumatic brain injury. In a mouse model, you see less injury to the brain. If you look at physiology, brain hypoxia, which is bad for the brain, is decreased by decompressive craniotomy. Not surprisingly, so is endocrine pressure. What's called the therapeutic intensity level, which is the effort needed to treat a patient, is significantly decreased by decompressive craniotomy. One has to recognize every therapy we have has good and bad effects. We can use advanced biological metrics, perhaps to decide who should we treat or when should we stop treating. So those patients who, for example, have brain oxygen monitoring and have a favorable outcome have a much greater increase in brain oxygen compared to those that don't. The same is being seen if you use microdialysis. With mitochondrial dysfunction, defined by the lactate pyruvate ratio and the pyruvate ratio, are significantly more frequent than those that have a poor outcome. So we can use advanced monitoring to select patients not only for a decompressive craniotomy, but who should continue to have care. Now, this is aggressive. This requires a lot of resources. But these are marker modeling studies done that demonstrate that at any age group, the more aggressive your care is, the better your outcome, including in the 80-year-old. It may not be cost-effective in the 80-year-old because the life expectancy is short. This may not apply to every environment. So a low-resource environment, this was a neurotrauma conference in Delhi, the resources available in India are not the same as the resources available in the United States. Here's a recent guideline that comes out of Columbia where they take the evidence and then a consensus in their group. It's different from what you might expect in the United States. How we do things may vary in the military environment as well. There seem to be economic benefits. So there's a certain cost savings when you do a decompressive craniotomy. And that also is seen even at an older age. So what can we take home from this? First of all, you want to optimize care in a specialized center with advanced neurosurgical and neurocritical care. You have to recognize that our guidelines and our protocols have to be individualized. In other words, it's not one size that fits all. Decompressive craniotomy will decrease end cranial pressure. It makes care a lot easier. You have a lower mortality. It's cost-effective. Should it be done in everybody? It's probably best if you haven't lost your capillary reflexes or you don't have evidence of herniation. Those patients are not going to do well no matter what you do. And similarly, if you do a decompressive craniotomy and you still have raised end cranial pressure and herniation-induced infarction, withdrawal of care may be appropriate. We still don't know the correct answer. Who should be the surveyed adult exactly when the real ramifications of technique, how do we best monitor these patients, what defines the outcome? The trial definition, physician definition, or patient definition. So there are some guidelines that exist. This is the consensus statement from the International Consensus Meeting on decompressive craniotomy led by Peter Hutchinson. Recommendations for primary decompressive craniotomy, secondary, which is the focus of what we've talked about, perioperative care, surgical technique, cranial reconstruction, and in low- and middle-income countries. So primary decompressive craniotomy, for which we don't have much evidence in these ongoing trials, is really done as, you take out a large subdural, the brain is swollen, you decide to leave the bone out. They do recommend that if the bone is beyond the inner table, if it's very relaxed, or it's an epidural hematoma, you don't need to leave it out. The intermediate's all about surgeon judgment. This is subjective. This is a judgment issue. But again, the bone flip has to be big. If you've done a small craniotomy and leave the bone out, you're going to make worse outcomes. It has to be big. And ideally, you should put in an intracranial pressure monitor in these patients afterwards. Secondary decompressive craniotomies occurred if you've done a craniotomy, let's say. This is not just for raising the cranial pressure. If you've done a craniotomy, put the bone back, and they then get an increase in the cranial pressure. Usually, they had a higher ICP, combined lesions, early secondary cerebral insults drive this. But secondary decompressive craniotomy, the recommendations from the consensus conferences, you need an ICP monitor to make a decision. But it's not the only thing that drives your decision-making. Other tools may be important. And it's important to understand the mechanism behind that, increase in cranial pressure. You should look at the selective choice. In other words, the ideal candidate is a patient where the ICP elevation is the primary reason why they're going to have a poor outcome. And that should be taken with other tools, whether they be monitoring or imaging or clinical finance. You should involve the family in that discussion, and you should continue to monitor the patients afterwards to understand the effect of a decompressive craniotomy on physiology. There's uncertainty with technique. In this craniotomy, I don't do, I don't think it works, I don't think it should be done. We don't quite know the optimal period for a gyroplasty, which may have an effect on the cranioplasty that's later done. Bifrontal decompressive craniotomy is perhaps less successful than a hemipraniotomy that may depend on how you expose or divide the superior central sinus. So the most recent guidelines for traumatic brain injury, which were published two or three years ago, did not have recommendations for decompressive craniotomy. They've now been updated. And they include a secondary decompressive craniotomy can be performed for late refractory ICP. And this can improve mortality and favorable outcome. You do not necessarily have to do it for early, in other words, recapitulating death rate. But that data cannot be extrapolated when you're just leaving the bone compound. Whatever you do, you want to do a big opening, all right? And also, this can improve ICP, but whether that makes a difference to outcome remains uncertain. As I said, selection is important. You don't need to have a person who's got evidence of herniation or infarction. They're not going to do well. You're not going to take out the subdural and later on say I'm going to do a decompressive craniotomy and this guy with a duré hemorrhage and evidence for herniation. He's not going to do well. So selective is important. So your surgical decision-making is dependent on experience. It's dependent on the competence of the surgeon. It's dependent on the risk and benefit to the patient. Evidence can help. Individual clinical experience in pathophysiology can help. So your decisions need to be made in the context of what's indicated for that patient and a consideration of what that patient and you are going to be guided by the family often in these circumstances, what their expectation for the quality of life is. Thank you. Thank you, Dr. LaRue. That was an excellent presentation. I think everyone was at awe of your presentation throughout the process and really appreciative of the evidence-based guidelines you gave, but also medicine is not a black and white practice. So we have a few questions. There's one, similar approaches to decision-making in peds versus adults. Do you feel like there's a significant difference? They are similar, but there are differences. Obviously the pediatric brain is very different. One of the trials that I didn't mention was by Taylor, which is really more in pediatrics. You could argue that perhaps the pediatric brain is more likely to recover than the adult brain and hence it's better to do. Secondly, the brain is tight to start with and hence you're more likely to see raise in the cranial pressure. Third, there's often a hyper-remake response and so there's value to it. But there are distinct differences and we do not have as much data on children as adults. There also is not a huge amount of data on the role of intercranial pressure monitoring in children. And the percent of patients who get monitoring is very small relative to adults. So we don't have as much evidence. The guidelines are different. I do think your decision-making is somewhat guided by the adults and it's individualized. And importantly, the discussion has to occur with parents because they're usually there. And a very clear conversation be had. It is difficult to extrapolate outcome data in adults to children. And I certainly know that there are some parents and I've had conversations with who are quite happy to have a child who they're taking care of at home, who's talking to them versus is out playing in New York. They look at that as a really good outcome. I know for one of my kids, if they were eight or nine and they were gonna be stuck inside and I had to take care of them, I wouldn't want that. That's me. But other parents do. And I think that's a very good point. But other parents do. And that's important to recognize patient and family specific outcomes might be different from our biases as a physician. Especially. We have a comment about floating a bone flap in a pediatric population especially. What your thoughts are? I'm not a fan of the floating bone flap. I'm not a fan of the hinge craniotomy. I know that people use the rationale as I don't have to go back and do a cranioplasty. But depending on the age of the child, you may have to go back and reconstruct it because you have a horrible cosmetic result. So I like to be black and white when it comes to this. If I'm gonna do a decompressive craniotomy, I take the bone out. I don't like the hinge or the floating bone flap. There is no good data on that it actually helps. There's really no good data on that it really hurts. But you're sort of making a real judgment decision on that. My personal belief, which is really driven by physiology, is it doesn't work. Another comment on decadron and subdural hematomas. Should it not be used? So in chronic subdural hematomas, different condition, there are some people who suggest that it may be helpful. In the acute phase, all of the data we have, if you look at crash, it doesn't work. It increases bad outcome. So in the acute phase, there is no role for steroids in large populations. Now, could there be a role in select patients? Yes, and that's the importance of individualizing care. I think you have to differentiate the use of steroids from acute, subdural, and chronic. There is some role in chronic. I very occasionally use it. If, for example, I need to hold a patient for a day or two before they get their blowhole. But that's a different pathology. Last question. Do you feel a pupillometer use is important in all patients and for how long? I do think it's valuable. I have no clue how people can tell me that the one pupil is three millimeters and the other is two. I have no clue how that is done. I can tell you it's big versus small or not reacting or reacting. That's a very crude measure. A pupillometer does allow you to quantify that. And there is very useful information that comes from that. That's being maybe better described in say cardiac arrest and the non-surgical neurologic emergency. So Mauro Otto and others in Europe really demonstrated the value of it. So if you've got a pupillometer, I use it. It does allow you to quantify the pupillary response and also gives you some useful prognostic information. My opinion, there's some things that if you've got a neurointensive care unit, you're taking care of a lot of patients with severe brain injury, whether it be head injury or subarachnoid hemorrhage or intracerebral hemorrhage. Buying a portable head CT and a pupillometer, great investment. Those are wonderful tools to help you make decisions. Thank you, Dr. LaRue. We're gonna end it here. We're a little over time. We've got about a 20 minute break and then we'll be back with Dr. Prasad, spinal cord injury evaluation and treatment. Thank you again, Dr. LaRue. It was an amazing talk and we're getting a ton of chats coming through saying how thankful they were that you were able to give this one for us today. Well, thank you. I think I've got to come back during the discussion for a brief period of time. So happy to take- You are, you'll be back with us hopefully. Yes. Yeah, I've got about a 30 minute window that I can do that. So happy to help. And as I said, delighted to work with PAs and nurse practitioners. They make my life a lot easier and patient's outcomes better. So thank you. Thank you.

decompressive craniectomy

traumatic brain injury

evidence

patient outcomes

individualized care

techniques

complications

decision-making process

family input

holistic approach