SCCM Resource Library-ACCM Distinguished Investigator Award and SCCM-Weil Research and Discovery Grant Recipients

Video Transcription

So, good morning. My name's John Sevransky, and I have the privilege of moderating this session. There are disparate topics, but the theme  that holds this all together is today's speakers have won something, either an award or received a grant. And so, the people who you'll be hearing  from this morning will be presenting cutting-edge type material. I need to briefly mention a couple of housekeeping things. If your cell phone is on, please silence it,  and if there are any slides that say no social media, please don't tweet them. With that, why don't we jump right on in to our first presenter, Professor Menon,  who has won the ACCM Distinguished Investigator Award, and will be talking about traumatic brain injury. Thanks very much, John.  I'm very grateful to the ACCM for giving me the signal honor, and it's a pleasure to present stuff here. So, these are my conflicts of interest,  which are not really relevant to what I'm saying in the talk, which is really a summary of bits and pieces that I've picked out from the work we've done over the last 20, 30 years.  So, what does the current ICU management of traumatic brain injury look like? There's no mechanistic treatment for many diseases in critical care, this is true.  So, our current treatment targets secondary injury, prevents further injury after the initial insult of trauma. And the primary target here is inadequate oxygen  and substrate delivery. We try to prevent hypoxia, we treat raised cranial pressure, and try and keep the cerebral perfusion pressure high enough.  And this approach is supported by the fact that when you look at autopsy in patients who die of traumatic brain injury, ischemic infarction is common.  And there's some benefit that's been demonstrated from ICP-CPP targeted therapy, which is codified in this, which is the Brain Trauma Foundation Guidelines,  essentially talks about two lots of monitoring, ICP monitoring and CPP monitoring, provides a threshold for ICP treatment and a CPP for favorable outcome and survival of 60 to 70.  But doesn't provide any evidence for any intervention to treat any of these targets. But that doesn't stop intensivists. You guys know that we tend to find ways to treat people,  and this is our treatment algorithm for ICP and CPP, which consists of different stages of therapy based on increasing therapy intensity and side effects,  starting with basic intensive care, ventilatory management, perhaps hypotonic solutions. We tend to use hypotonic saline, a bit of hypothermia,  deepening sedation, and eventually taking the lid off the skull, decompressive craniectomy. And that's been summarized really nicely  in the Seattle Severe Traumatic Brain Injury Consensus Conference, and this publication is now out. It stays in press, but it's actually now in publication.  So the summaries are there, and lots of people have similar algorithms. The problem is that when you go looking at the evidence  that underpins these algorithms, it's really not very favorable. There's no EBM support for aggressive therapies. The BESTRIP trial suggested  that ICP-guided therapy was no better than imaging and clinical examination. DECRA suggests there was a harm from decompressive craniectomy.  Urotherm 3235 suggested there was harm from hypothermia. And POLAR, recently published in JAMA, suggested no benefit from hypothermia.  So this makes you wonder whether this is what we should be doing. But the counterargument to all of what we have here is this. I would say that this is with a single ICP threshold  of 20 millimeters of mercury, and all of these are interventions which are harmful and hazardous, but being used early and in all subjects.  And in response to all these high-level citations, I would go, as many people in the UK do, to Shakespeare and quote, diseases desperate grown  by this desperate appliance are relieved or not at all. And clearly, I've given you the citations for all of these, so the citation for Shakespeare is Claudius,  Hamlet, Act 4, Scene 3. So you can look it up and check that I've given you the right data. And the question here is, was the disease desperate enough?  Did we balance risk and benefit appropriately? How does that work? Well, let's look at this. This is raised intracranial pressure.  This is a young man who presents with an extradural hemorrhage, a deep coma, has been lying in a car park of a pub for the last 12 hours, so turns up too late and dies.  And this is at autopsy. You can see the uncus of his temporal lobe looking from below has herniated, caused a diuretic hemorrhage, and resulted in death.  Now, if this guy pitches up in the emergency room, the ICP threshold is not 22 millimeters. It's over 30. We don't know what it is. But here, this is associated with death.  On the other hand, this lady falling down the stairs after having had a bottle of vodka has liver disease, has her subdural hematoma evacuated,  but despite blood products, is still oozing, so they leave the bone flap out. But a post-operative PET scan shows hypoperfusion underlying the subdural hematoma, despite the fact  that the ICP and CPP are perfectly respectable. We don't know what the threshold for this physiology is. It depends on the CPP, but it varies from patient to patient.  Now, these have different risks, thresholds, and urgency, and different acceptance of iatrogenic risk, so we need to calibrate ICP treatment to the risk-benefit ratio.  Here, we would use every possible therapy because we are avoiding certain death. Here, we have to be much more circumspect with what we do because the hazards start having effects  compared to the disease itself. What about less desperate disease? Well, the BTF protocols have used averages for CPP targets. There's some refinement with age and auto-regulation,  but the optimal CPP varies across individuals, so if you're using an average target, some will be undertreated and some will be overtreated,  and the threshold for CPP, which is optimal, may be as wide as between 50 millimeters of mercury and 110 millimeters of mercury. So we need better tools to titrate ICP  and CPP targets in individuals, otherwise we wind up with this one-size-fits-none line, which is currently what we do in essence. So how can we refine these targets?  We can minimize the side effects of therapy. We can account for differences in age. We can target optimal auto-regulation, which can be measured at the bedside,  and then realize that the CPP doesn't exist as a number to be treated. It delivers cerebral blood flow, and that delivers oxygen, glucose, and when there's metabolic compromise  because these are inadequate, lactate goes up or the lactate-pyruvate ratio goes up. This can result in stress or injury markers as cells die, and you can have these biomarkers  shed into the circulation. So the way many neurointensive care units have done this is to combine tissue oxygen monitoring and microdialysis,  which measures lactate-pyruvate ratio and brain glucose, and PRX or some other marker of auto-regulation. It's not important what this is, but just to say that we target  to an optimal auto-regulatory range. And you can see that the numbers next to each of these are not single numbers. They tend to be two sets of numbers. How does that work?  Well, the more stringent numbers are for favorable outcome. The less stringent numbers are for survival. These do vary with age and gender,  and I've given you the list of papers here. Very happy for you to take a picture if you want. But the ones that have got arrows next to them  have opportunities to titrate not to groups of patients, but individual patients. So we can start trying to find out how these things are important to titrate  in individual patients rather than groups. And we can actually create little thought diagrams of physiology that's normal, desirable, and injury thresholds.  So I'll give you an example of ICP. You and I probably run below 10, maybe 5 millimeters of mercury. We want it to be less than 20, but once it gets well above 25,  we know it causes harm. Between the two, there's a bit of gray area. But what we can do by looking at these thresholds for ICP, PRX, jugular venous saturation,  brain tissue, lactate-pyruvate ratio, and brain glucose is to look at the acceptable treatment intensity. For this, we would use the treatments  that had an intensity that wasn't associated with significant hazard. But once we were getting to this, it becomes much more likely that it'll cause physiological harm.  And if it gets very bad, cause death. And so you can afford to use more hazardous therapies. We make the juice worth the squeeze. We use graded thresholds both to select treatments  and also to minimize harm. But these thresholds are now based on just associations between outcome and the physiological target.  And that's not good enough. But now there are a set of trials that are coming about. So BOOST-3, which I don't see, I don't know if Laurie Shutter is here in the audience,  but Laurie is one of the PIs on that, which is looking at brain tissue PO2 guided therapy in addition to ICP. Bonanza is the Australian counterpart of BOOST-3.  And Cogitate is a phase two trial that's just completed looking at the implementation of auto-regulation-based targets.  Now that's all very well, but this is saying that your only problem is a nail and what you're trying to do is to find how big a hammer you want to hit it.  This is not precision medicine. This is stratified medicine. What we have are problems that are nuts and bolts screws. Some are nails, but we need a variety of tools  so that we can hit these with precision.  And let's look at some of these. We've been talking about classical ischemia. And even if you look at classical ischemia, this is a paper we just published in JAMA  looking at 90 PET scans in 68 patients over time.  And we excluded any area which had a lesion. So this is not abnormal physiology around lesions. This is in the rest of the brain. And what you can find is that there's substantial  individual temporal and spatial heterogeneity. Classical ischemia occurs in some regions in some patients, often in the absence of raised ICP. And it's missed by bedside monitors.  So this is the summary data from it. Each column is one patient at a given time point. And you can see the different dots of the different areas.  So there's huge spatial heterogeneity and huge temporal heterogeneity across patients. But there's also a lot of mechanistic heterogeneity.  We've been talking just about classical ischemia, which is what we hit with CPP elevation and ICP control. But you can also have microvascular or diffusion  barrier ischemia or mitochondrial dysfunction, either because of injury to mitochondria or because nitric oxide acts as a competitive inhibitor at the cytochrome respiratory chain.  And this has been addressed by many authors, including us. Here you have the work from Paul Vespa, which is seminal in demonstrating energy crisis mitochondrial dysfunction.  But the most important thing is that the therapies for these could be potentially different. For the top, for classical ischemia,  you use CPP optimization, auto-regulation guided therapy, optimizing CO2, and reducing energy demand. Here we've got another set of potential interventions. Some are preventive.  So the work from Pittsburgh, from Ruchi Jha and Pat Kohanek is looking at edema reduction around contusions by addressing SIR1 and ABCC8-related mechanisms  and perhaps anti-inflammatory mechanisms. And here there's a whole host of interventions we might use, new substrates, which the UCLA group have been looking at substantially.  But it's... so we've talked about energy failure not just being classical ischemia, but beyond energy failure, there are other mechanisms. So this is work looking at inflammation  with microdialysis. What Peter Hutchison, when he was doing his PhD in our unit, showed was that there's a time frame to different cytokines across the brain,  looking at microdialysate with 100 kilodalton cut-off catheter. We can then use drugs that modify this. This is IL-1RA work from Adele Helmi,  showing that we get the IL-1RA into the brain, and detectable by microdialysis. That does what it says on the TIN, and there's a change on principal component analysis  on the cytokines in the treated versus the untreated patients. We can interrogate this further to find out which downstream cytokines are important.  We can look at the imaging correlates of that in terms of contusion evolution or other features in MRI. And then look at the cellular aspects. This is PK-1195-PET in patients,  showing that there is contusion... there's recruitment of monocytes to contusions, and then microglial activation around it. And then there are mechanisms which we commonly  think of in the chronic phase of TBI. This is work from our group looking at imaging amyloid deposition in patients with traumatic brain injury,  both in the acute phase and going out up to a year. This replicates patients' data from patients who've died, post-mortem data. But the important thing is being able to look at it  in life means that we can look at serial changes. But this also has shared biology with chronic neurodegeneration and may explain why patients  who have TBI may be at higher risk of late dementia. But amyloid may also be pathogenic in the acute phase. We know it causes vasoconstriction and can  promote inflammation and neuronal loss. So, so much for mechanisms, but in the end, it depends on what happens to the patient. And in most of critical care, we just worry about  whether patients die or survive. All of the RCTs look at that in general critical care. But it is possible to make patients with neurological  injury survive, but have very bad functional outcomes. Consequently, we use the extended Glasgow Outcome Scale as our tool, sorry. And we have a variety of functional states,  starting from vegetative state, lower severe disability, upper severe disability, lower moderate disability, and so on. The most common dichotomization in intensive  care is here, between patients who are independent or dependent. But we have, in a recent trial, used upper severe disability as something that we wanted as a target.  So, pre-specified that because, as I'll explain to you, it may be something that's worth having. So, in response to that, there was this editorial  from Peter Lundquist in the BJA two years ago, which was an overall stringent criticism of aggressive interventions that do not result in full functional recovery.  And he really said full. And one of these was our RCT of decompressive craniotomy for refractory ICP and TBI, which is rescue ICP. And this is important because deaths directly  attributable to TBI are due to refractory intracranial hypertension. And if you decompress them, they won't die. And if this is a healthy young person, otherwise  they will survive with very severe functional disability, if the neurological outcome is bad. And this is an irreversible intervention.  Rescue ICP used as a rescue treatment for really refractory intracranial hypertension showed improved survival and functional outcome, but  with the pre-specified threshold of upper severe disability, which is independent at home, for eight hours, up to eight hours. And the question which Peter Lundquist asked was  this, was this a life worth living? So our response was that we needed to acknowledge the perspective of patients and their families.  And we came down to saying that these are not easy questions. We need to acknowledge that all of our answers are imperfect ones, often resulting in the choice of  the least worst rather than a best option. Because if someone has a severe traumatic brain injury, complete recovery quickly within the six  month threshold that we often employ is not a realistic aim. We have to think more carefully. And this brings us to what kintsugi is. So we recently had, from the Center TBI  collaboration, a paper in Lancet Neurology, and Elena Beccabaroso put this on the cover. And what she showed was this pot. And kintsugi is the Japanese art of repairing  broken pottery by mending the areas of breakage with lacquer, dusted, or mixed with powdered gold, silver, or platinum, which adds to the beauty of  the pot that those imperfections make it more perfect. And what Elena showed was the fact that traumatic brain injury comes up with a kintsugi-like picture.  And within Center TBI, we have a haiku competition. This is my haiku, which I submitted. Broken China on our whole again, but can your kintsugi mend the me which injury shattered?  Those of you who want to try it, you can go and try and do it. Send it in to Center TBI, and we might enter it in there. But the thing is that when we're making these  decisions, it's fine for us to talk about the science, but really these are cultural constructs that we need to understand based on cultural inputs.  And our group, science, scientifically didn't have the expertise to do it, so we went to another group, and that group was Bon Jovi. So Bon Jovi, originally in one of their  publications, said, we've got to hold on to what we've got, because it doesn't make a difference whether we make it or not. We've got each other, and that's a lot, for love  and we'll give it a shot. And I wouldn't come and give the distinguished investigation lecture with data that wasn't appropriately cited. So this is it, Jovi J.B. Sambora, R.  Child D., living on a prayer. The album is slippery when wet, but they were very bad. They published this. They had dual publication. There was Mercury Records in the States and  Vertago Records in Europe, and if Howard Bauchner was here, he would decry that. But like many people do, this acceptance that you should try as hard as possible was  countermanded in the subsequent publication of theirs, and that's this, which says, it's my life. It's now or never. I ain't going to live forever.  I just want to live while I'm alive. It's my life. And that's a more fatalistic view. Get me as right as I want to be, otherwise I don't want to be, I don't want to survive.  Now, these two, I often use this picture because the cognitive dissonance from one group is what plays out in my brain every time I'm talking to patients' families, often at the  same time, because we have no knowledge of how people are going to do. If we had more precise prognostication, we might do better, but we don't as yet.  So, I'll finish with Bon Jovi, and thanks to all of the folk who've contributed to these studies, people from the Neurocritical Care Unit  at Cambridge, the broader division of anesthesia, radiology, the Wilson Brain Imaging Center, neurosciences, and beyond Cambridge, particularly  to Andrew Moss, Jeff Manley, and Adrian Owen. But all of this is team science. It's the Acute Brain Injury Group at Cambridge, the Center TBI and Icon TBI collaborations,  which I lead, and TRAC-TBI, GAIN, and INTIBIR. Thank you very much. Professor Menon, thank you for a wonderful and witty talk. We, unfortunately, are on schedule for the next talk.  If you're available, can people come up to you after the session? And, okay. So, moving on from the brain to the mouse, we have the winner of the WILD grant, Dr. Wendy Walker,  who's going to be talking about sepsis and surgery alter myeloid cells in the mouse peritoneal cavity.  So, I'd like to thank the Society of Critical Care Medicine for funding this research and for this opportunity to present my data. And I have no conflicts of interest to disclose.  So, I'm sure you're all aware that sepsis occurs when an infection induces a dysregulated immune response, culminating in organ damage. And this dysregulated immunity incorporates  components of hyperinflammation, as well as immunosuppression. So, one of the cell types that responds to infection during the acute phase is the monocyte, which  differentiates into macrophages. There are also tissue resident macrophage populations that respond to the infection. And these cells play both helpful and harmful roles  during sepsis. So, they are very important to phagocytose the pathogens and produce inflammatory cytokines and nitric oxide. They can also show components of immune suppression.  And these can be helpful or harmful depending on the timing and the nature of these changes. So, we're interested in the monocyte and macrophage populations.  It turns out if you look in the peritoneal cavity of mice, there are two distinct populations of tissue resident macrophages. And the first are the large peritoneal macrophages  here. These cells are F480 high. They're fairly easy to identify. And they're the most predominant population in the peritoneal cavity.  However, there's a second population of cells, these small peritoneal macrophages named for their smaller size. These actually only have intermediate levels of F480.  But they are MHC2 high. They're a little bit harder to identify. They require a comprehensive panel of antibodies for flow cytometry detection.  During inflammation, though, these populations change dramatically. So, the LPM appear to disappear from the peritoneal wash. And this has been described as the macrophage  disappearance reaction. You've probably heard of it. And it turns out that they don't die. But a recent report by Paul Cube suggests that  they actually migrate to sites of sterile injury. So, when they did sterile liver injury, the LPM traveled to this site and nibbled away at apoptotic cells.  And then also during peritoneal inflammation, monocytes infiltrate into the tissue. And if you have a systemic infection, you may even have emergency myelopoiesis generating very  high levels of monocytes. These might either be mature monocytes or, at some point, perhaps even myeloid-derived suppressor cells. I'll talk about those later.  And then these may differentiate into increased numbers of SPM to bolster the number of these cells. And the LPM and the SPM, they have a little bit  of different function in terms of the cytokines that they secrete. And there's some evidence that LPM may be more effective at phagocytosing apoptotic cells,  whereas SPM, they may be more effective at phagocytosing pathogens. However, the changes in these cells, they've been characterized in terms of sterile inflammatory  response in the peritoneal cavity, but not in terms of abdominal sepsis. So we wanted to find out, how do myeloid cells change in the peritoneal cavity during abdominal  sepsis and surgery? And to study this question, we used the mouse cecal ligation and puncture model. You're probably familiar with this model, where  you tie off a portion of the cecum, which is a loop structure of the intestine that contains a lot of commensal bacteria. Puncture it with a needle, allowing that bacteria  to escape into the peritoneal cavity. We do treat our animals with antibiotics and lactated ringer solution and get about a 30 to 50 percent mortality rate in this experiment.  And then, as a control, we just do the sham surgery, where the cecum is exteriorized and put back in and the animal is closed. No ligation or puncture to allow the bacteria  to escape. So we performed the CLP and sham surgery and sacrificed the mice at one, two, or 14 days post-surgery, harvested peritoneal lavage, and  stained the cells, and performed flow cytometry to see what was there. So this is just a brief view of the FACTS procedure that we used, the gating pipeline.  So I won't go through all of it, but basically, we gate the cells, the single cells, the leukocytes, exclude some populations that we're not interested in.  We look for CD11B positive cells. These will be the myeloid cells. We've already excluded the neutrophils in this particular gating strategy.  Over here, you can see a plot of F480 versus MHC2. So these F480 high cells are very abundant in the naive mouse, and then this less abundant SPM population.  However, after CLP, here in the bottom, you can see the LPM are virtually gone from the peritoneal cavity. There are still some SPM there, but we didn't see  that they increased dramatically in their proportion. And then in the naive mouse, there's very few monocytes, but after CLP, the monocytes become quite abundant.  So this shows sort of a plot of what we saw. I think it's about 9 to 11 or 9 to 12 mice per group. So on the left-hand side, the percentage of LPM in  terms of the over-leukocytes, and on the right-hand side, the absolute number of LPM that we found in the peritoneal cavity. So they're present in the naive mouse.  However, they rapidly disappear after either sepsis or sham surgeries. This seems to be perhaps a more general response to injury and not just to the infection.  In the sham mice, they had reappeared by day 14. However, that was not the case in the CLP mice. Now, note that we don't go back and resect the necrotic cecum.  The animals, so I say about 30 to 50 percent of the animals die. Most of that happens within the first three days post-CLP. The animals that survive out to here, they appear  healthy, but they do have a big adhesion within their abdominal cavity blocking off that necrotic cecum. So if you look at the monocytes, these rapidly  infiltrate, both in animals that have undergone CLP and sham surgery. In the sham surgery mice, they're pretty much gone by 14 days post-surgery.  But many of the mice that had undergone CLP and survived to this point still have monocytes in their peritoneal cavity.  So with reports of sterile inflammation, they found that these monocytes rapidly differentiated into SPM, but we did not find this to be the case after CLP.  So there was little change in the number of SPM at one or three days post-surgery. And in sham mice, there really wasn't any change at day 14 either.  In the animals that had undergone CLP, there was some increase in the number of SPM at 14 days post-surgery. So it seems like they do transition into SPM, but  this process takes a little bit longer. And remember, the majority of animals that die are going to do so during the early time point.  And at that time, there's not really a lot of SPM that have been generated. This gave us a clue that perhaps the monocytes are coming in and fighting the infection even before  they differentiate into SPM. So now we wanted to see if there were differences between animals that survive sepsis versus animals that succumb.  And here's a survival curve of some C57 black six mice that we did about the same time in the lab. They had 61% survival. And previously, Daniel Remick has shown that  interleukin-6 is a very good biomarker that can be used to predict whether an animal will live or will die after CLP. And we verified this in our own experiments and  used this to set the threshold. So we found that if you take serum at 18 hours after CLP, for the animals that are going to survive sepsis, they all have IL-6 levels less than  five nanograms per mil. But the animals that were destined to succumb to sepsis, the majority of them had IL-6 levels greater than five nanograms per mil.  And even these couple of mice that were below that threshold, they were animals that died late after CLP. So especially for acute mortality, it's a good predictor.  So we used this IL-6 to segregate our animals into the ones that were going to live versus the ones that were going to die. So we performed CLP in 18 hours post-surgery.  We took that serum sample to determine the group of mouse and then also harvested peritoneal lavage and stained it for the monocytes. So here you can see in the animals that  were predicted to live based on low serum IL-6 levels, they had a statistically higher percentage of monocytes in their peritoneal cavity  in comparison to the animals that were predicted to die. We also saw a slight trend in the absolute number of monocytes, but with the number of animals in our study,  that's not statistically significant. So we also noticed when we were doing this fax gating that among them, this is looking at F480 versus LY6C.  So among the LY6C positive cells, some of them seems to be F480 intermediate. And we thought that maybe this is a monocyte that's starting to transition into an SPM.  So when we quantified these F480 intermediate cells, we found that there were a higher percentage of F480 intermediate cells amongst the monocytes in animals  predicted to live versus animals predicted to die. Another thing that we wanted to look at was the possibility that perhaps these cells  were acting as myeloid-derived suppressor cells. So it seemed surprising to us that it was the animals with the low IL-6 levels that had the higher number of monocytes  in their peritoneal cavity. If these cells are some of the cells that are making IL-6, you might expect the trend would be the opposite way.  So we thought maybe these monocytes, somehow, they're not so inflammatory. And we looked to see whether they might be myeloid-derived suppressor cells.  These are cells that kind of get stuck partway in their differentiation. There are monocytic ones and granulocytic ones. The LY6C ones are the monocytic ones.  And one of the markers is CD31. So indeed, we saw a higher percentage of CD31 positive cells amongst the monocytes of animals predicted to live versus animals predicted to die.  So we had to do further characterization to see if those are truly myeloid-derived suppressor cells. So then we wanted to perform some functional tests  to see whether these changes in the cell subset actually, what we see trafficking after sepsis and trends in animals predicted to live  and animals predicted to die, did they actually have a functional consequence? So in order to determine the functional role of having more  monocytes, we wanted to see if the monocyte infiltration could somehow be reduced. Would that perhaps increase sepsis mortality? For this purpose, we used CCR2 knockout mice.  And notice that CCR2, it's a chemokine receptor that's on the surface of monocytes. And it mediates their trafficking from the bone marrow into the blood  and then from the blood into the site of infection, in this case, the peritoneal cavity. So we obtained CCR2 knockout mice and compared them to wild-type mice. We performed CLP.  And the first thing we wanted to do was determine, do they truly have a reduced monocyte infiltration? So here we found that they did. The CCR2 knockout mice had a lower percentage  and absolute number of monocytes in their peritoneal lavage relative to the wild-type mice. It's not completely gone. It's just reduced. And this is kind of akin to the levels  that we'd seen in the animals predicted to die, at least for the percentage over here. So then we repeated this experiment and looked at sepsis pathogenesis.  So in contrast to our hypothesis, we found that CCR2 knockout mice had a similar survival rate to their wild-type counterparts,  despite the fact that they have less monocyte infiltration. So we need to do further study to find out, with other approaches, to change the number of monocytes  and see, are monocytes still functionally important? Did this experiment just, was the approach just wrong? Or is it truly just that monocytes,  they reflect the outcome of the animal, but they're not functionally important? OK, so we also wanted to find out if the LPM affects sepsis pathogenesis.  Even though these cells do disappear, they disappear from the peritoneal wash, we know from the prior work of other investigators,  they're probably going somewhere and doing something functionally important. We thought maybe they're going to the site of the surgical injury or the punctured cecum  and helping to seal that off. So to test the role of these cells, we used these CD11B DTR mice. They expressed diphtheria toxin on the surface of CD11B  high cells, diphtheria toxin receptor on the surface of CD11B high cells. So you can treat them with diphtheria toxin to deplete the CD11B high cells.  And when we did this, you can see that the LPM disappeared from the peritoneal cavity. The SPM were maybe slightly affected, but not as big of an effect as we saw with the LPM.  And we found that diphtheria toxin treatment through abdominal administration did not affect blood monocytes. In fact, I'm not showing the data here,  but it actually caused a little bit of an increase in monocyte infiltration into the peritoneal cavity prior to surgery. So we treated these animals with diphtheria toxin or PBS  and then performed CLP to induce sepsis. And here we saw a dramatic phenotype. So with the animals treated with PBS, we only had a small amount of mortality,  but when the animals were treated with diphtheria toxin to deplete the LPM, this dramatically increased animal mortality.  So these data suggest that LPM are playing an important role in protecting the mouse from sepsis.  Okay, so to summarize, we found that the peritoneal myeloid cell composition shifts dramatically both after infection and after sterile surgery.  The LPM disappear from the peritoneal cavity by one day post-surgery. In the sham mice, they had reappeared by day 14, but this was not the case in the CLP mice.  We found that monocytes became very abundant at one to three days post-surgery in both CLP and sham mice, and they disappeared by 14 days in the sham mice,  but were still present in many CLP mice at 14 days. And the SPM, there was little change post-surgery in the septic and the sham mice,  but by 14 days, the septic mice showed an increase in SPM in the peritoneal cavity.  We found that mice predicted to survive sepsis had more monocytes in their peritoneal cavity versus animals predicted to die. They had more of these monocytes  that were F480 intermediate, which we think perhaps could be transitioning into SPM, but it will require further study to determine that.  And we found that they had more CD31 positive cells amongst their monocytes. We think that perhaps these could be myeloid-derived suppressor cells.  We know that they exert negative effects during chronic sepsis,  but perhaps during this acute phase, they're actually exerting protective effects. It will need more experimentation to determine if these are truly MDSCs  and express other markers of that population. We found that LPM play a critical role in protecting mice from sepsis, but we don't have any evidence yet  that the amount of monocyte infiltration influences sepsis, but we're working on monocyte adoptive transfer experiments and other approaches to see  if maybe somehow this still plays a role. So I would just like to thank the people that have contributed to this study, and as well as the Society of Critical Care Medicine  for providing funding for this. And if there's time, I would be happy to take any questions. Otherwise, see me afterwards. Thank you. And thank you for listening.  Thank you, Dr. Walker. And again, if you don't mind hanging afterwards, we're running up against time.  Our third winner, if you will,  is the awardee for the Discovery Research Grant, and he's going to be talking about the BEARD study. Dr. Pham. Thank you very much.  Thank you for the opportunity to present this work and for the award. But we are going to talk about mechanical ventilation right now.  So I have no financial conflict of interest for this, but we have an academic collaboration with a group of Dr. Louis Blanc who works in Barcelona,  and we used his software for the validation of reverse triggering that you will see in a few minutes. So we defined dyssynchrony by a discrepancy  between the breath delivered by the ventilator and the patient's neural time and needs.  Dyssynchrony might be important for the patient because we see that it's associated with the outcome. It's associated with longer duration  of mechanical ventilation and even with mortality.  It might be the underlying mechanism that explains better outcome when patients with ARDS are paralyzed because they cannot have any dyssynchrony when they are paralyzed.  And it might explain some of the results of the ARDS study, which was published two years ago in the JAMA and where they showed that patients  who had recruitment manoeuvres and high level of PEEP had a higher mortality than patients with lower level of PEEP. Because we see that these patients with higher level of PEEP  when they have some dyssynchrony and they have double triggering, they actually have higher pressure in the lungs than patients who have lower PEEP.  There is an association, but association doesn't mean causality. So we don't know if it's confounded by severity and in a little bit of a chicken and egg problem.  What we know is that we prefer our patients to be like this than like this, of course.  Dyssynchronies is a very wide topic because there are lots of different types of dyssynchronies. Dyssynchronies that happen when you have patients  with too much help for ventilation, patients who have too low support on ventilation.  There are dyssynchronies that happen when patients are awake, where some when they are sedated and they all have different mechanisms and consequences  and likely different risk factors.  They vary at different moments of the day and in the course of the mechanical ventilation of the patient and some also show that the clusters of dyssynchrony might be important.  So if a patient has a lot of dyssynchrony, even on the short period of time and then doesn't have dyssynchrony for a long period of time, this might affect the mortality.  And there is a huge inter-individual heterogeneity as well as within patient heterogeneity.  So the BEARDS study is a research program looking at breathing efforts during ARDS in invasively mechanical ventilated patients. It has three steps. The first step is to describe  the epidemiology of dyssynchrony, how we can detect it, monitor it, what is the incidence of the different types of dyssynchrony, what are the risk factors and mechanisms  and what is the association with the outcome. Then it will lead to a second step  in which we will develop algorithms to try to better manage dyssynchrony and then we will test these algorithms. And then we will design a cluster-randomized clinical trial  having some centers where you try to act on dyssynchrony and reduce dyssynchrony versus some centers that just monitor without acting on it.  So we will focus now on the epidemiology of dyssynchrony, which is the first step. So for our study, we included intubated patients  with a PF ratio lower than 200 and these patients had to be sedated, at least that inclusion.  For these included patients, we took some recordings of their waveforms. So we had airway pressure, we had flow and we had either electrical activity of the diaphragm  with a NAVA cassisor or an esophageal cassisor has a surrogate of pleural pressure. All these patients had 20 minutes recordings  three times a day and there were seven days recording.  We aimed at including 300 patients. For the moment, we have almost 200 patients. Most of them had esophageal pressure and one quarter of them had a NAVA cassisor.  The record is a days of recording in a median four days and we have more than 2000 tracings. It's a median of 12 tracings per patient.  So you can imagine that having more than 2000 recordings for now and then we will reach a sample size, we will likely have around 3000 recordings. It's not feasible to watch  and evaluate all the tracings  with just the vision and assessments. So we wanted to develop some automatic algorithms in order to detect and to have a better  epidemiology of all this dyssynchrony. So we wanted to validate an automatic detection of different types of dyssynchrony. So the group of Ruiz-Blanc in Sabadell  already has developed a software that is able to detect lots of different dyssynchronies that were already validated, mainly for ineffective efforts during expiration  and for the type of ventilatory mode.  They did not have a specific algorithm for reverse triggering, which I will describe right now. So reverse triggering is defined as a contraction of the respiratory muscles  that is happening after machine triggered breath. So it was described as a group of Laurent Brochard a few years ago in this publication. And when you see that on these tracings,  you have the flow, the airway pressure and here is electrical activity of the diaphragm. You have a passive insufflation. Here it's time triggered. The machine gives the flow.  And then here, a few milliseconds after, you have contraction of the diaphragm. This is the same with esophageal pressure. So when you have a negative swing,  it means that the diaphragm contracts. Again, you have a passive insufflation and then diaphragm contraction. And this happened also in volume assist control.  Here again, you have machine insufflation and here diaphragm contraction.  So there is a passive insufflation followed by a diaphragmatic contraction. This is the same here in the tracing when you have airway pressure, esophageal pressure and flow.  So you have passive insufflation and then diaphragm contraction that you can see as a negative swing. This happens again here. And you see it can lead to a second breath  that is delivered earlier.  And this is a concept of breath stacking. So if you have a breath and inspiration here, and then you don't have enough time for exhalation, and then you have another breath  that is triggered by the diaphragm contraction, you will think you ventilate your patient as six or seven ml per kilo, but in fact, it will receive twice the volume,  which can lead to 10, 12 ml per kilo, which can be deleterious for the patients.  We don't know for sure the impact of reverse triggering. As I just said, it can increase tidal volume. It can lead to an effect of panderluft.  Panderluft means that you have some air that it moves from some part of the lung to the other part of the lung during the contraction. For instance, here you have reverse triggering  and here is with EIT, so impedance metric of the lung. You can see that from the non-dependent region to the dependent region,  you have some sucking of the air during the inspiration. So in the non-dependent, when you have a passive breath, you have all of these volumes that goes  in the non-dependent part of the lung, but when you have reverse triggering, some part of this volume goes to the dependent part of the lung.  If the reverse triggering happens during exhalation, it can lead to eccentric diaphragm contraction. And as we already said, it can lead to double or triple insufflation.  We can even hypothesize that this could be protective. If you have reverse triggering, so you will have some diaphragm activity, you will maybe prevent some diaphragm misuse  and some loss of diaphragm function if you have some residual contraction without being deleterious. But this is an hypothesis. We don't know this yet,  and that's why we developed all this research around reverse triggering. So the first part of the research was to try to evaluate the performance  of an automatic detection of reverse triggering using a software, so the BetterCare software. And this software only uses airway pressure and flow,  so you don't need any additional monitoring or carcasser for this detection. And we compared this to a gold standard, which was visual evaluation of all these tracings.  But to make it easier for the gold standard, we had access to airway pressure, so we could actually see the diaphragm contraction.  So we selected 20 tracings. Most of them were during assist control, so volume or pressure assist control. As a control group, we have some patient on pressure support ventilation.  So if you are on pressure support ventilation, you are supposed to trigger all the breaths, and so you don't have reverse triggering.  And we had other control. We had four tracings with paralyzed patients. So when they are paralyzed, they also don't have reverse triggering. We analyzed almost 5,000 breaths,  and one quarter of them displayed reverse triggering. If you look at this confusion matrix, so you have the gold standard here.  So yes, visual assessment, there is reverse triggering. No, there is no reverse triggering. And here is the comparison with the automatic detection.  So you have most of the reverse triggering, which are here, that were detected by the automatic detection. And when you don't have reverse triggering,  most of the time, the automatic detection found that there was no reverse triggering. So this led to an accuracy of 96%, with a good sensitivity of 82%,  and a very, very good specificity,  positive predictive values and negative predictive values. If you look at the discrepancy, this happens here. So you had 4% of the breaths that are false positive.  So the detection says, yes, there is reverse triggering, whether when the expert said no, and the 17% of missed reverse triggering. We think that this is not a big deal,  to have some of the missed reverse triggering, because we didn't want to have over-detection,  I think if you miss some of the reverse triggering, but then the patient is monitored for several hours, you will capture when this happens. But this would be a problem  if you capture too many reverse triggering when they don't exist.  Some, we reviewed the discrepancies, some of them were obvious, it's because it was difficult for the investigators of all the detection to see exactly when was the trigger,  when the, here, you don't really see a patient triggering, but the contraction happens right at the same moment of the passive insufflation.  And you had some issues with the detection sensitivity, so the automatic algorithms provides a score, and they decided that above a certain score,  the output says, yes, there is reverse triggering, and their threshold was 35%. So here you clearly have reverse triggering, but the score was lower than their threshold.  So the conclusion of this first step of the research program is that automatic detection of reverse triggering is feasible, it's accurate,  and this is the first step that allows continuous monitoring so when you have continuous monitoring, you will be able to have a more precise idea  of what's the epidemiology of the different desynchronies. So next steps will be to describe the whole epidemiology of different types of desynchrony,  not only reverse triggering, but ineffective efforts, prolonged inspiration, short cycling, and so on, to propose some management for the types of desynchronies  that we think might be harmful for the patients, and maybe to shape the future of mechanical ventilation. And the ultimate step will be to have some algorithms  that will detect in advance before the patients are awake or have desynchrony,  that they could have them and to act ahead.  So of course, this is a collaborative work, and we have a lot of different countries participating, and I thank all the investigators from the BEARDS study,  and I thank you for your attention. Thank you, Dr. Pham. We, the presenters, or you've gotten through in such a time that we probably have time to,  maybe if I can ask all three speakers to come up and join me, and if people have questions for the presenters, we do have a few moments right now.  Dr. Menon, thank you. I'm not sure if it's on. We can hear you. Can you hear me? It's a small room, and I can project if I need to. So thank you for your presentation this morning,  and congratulations. One question. When you talk about desynchronization, what do you mean by desynchronization? What do you mean by desynchronization?  Well, desynchronization is when the patient is awake, and the patient is awake, and the patient is awake,  and the patient is awake. One question. When you talk about the fact that, you know, looking at quality of life outcomes,  we also have information that if you look at 12 or 24 months after injury, that sometimes things continue to change.  Any advice or comments on how to address the fact that many sponsoring bodies of research want us to always use six-month outcomes as our markers,  as our endpoints, when in a disease or a situation where you may see results on a longer time frame? How do we address that with our sponsoring bodies?  So I think we just have to provide the data that shows that we need longer outcomes. So I'd say three things about outcomes following most neurological injury,  but in this case, TBI. One is that the GOSC is not adequate. We need to get psychological health outcomes and cognition. They don't quite map on the way I thought we would.  We've just submitted 1,500 patients where we've looked at cognitive function and functional outcome to Lancet Neurology.  But what surprised me there was that the cognitive deficits are most driving functional outcome in the mid-range between lower-severity disability,  upper-severity disability, and moderate disability. At the top end, it's psychological health and PTSD, depression, anxiety, resilience. Those things play a bigger part.  So we need more instruments. And both CENTER and TRACK-TBI, I think, will come out with some kind of a combined outcome assessment. The second is the point you make.  We know that patients continue to improve, and the TBI model systems in the states have shown that people continue to improve up to 10 years.  Even at the worst outcomes, the vegetative state patients, Joe Giacino has written a lot about from TRACK-TBI, up to 30% of them may be  in gainful employment if you follow them up for long enough. These are people who pitch up to the rehabilitation groups.  And the third thing, I think, is to try and, which is reverse of what I'm saying, is to try and reduce the burden of collecting outcome data,  because patients, when they've come for these big studies, have spent four hours doing outcome assessment. At the end of the four hours, I'm sure they're not functioning quite as  well as they were at the start. So, Dr. Pham, I wonder if you could talk for a moment about the relationship between breathing dyssynchrony and the size of your tidal volume.  Is that something that we need to worry about? I think we really need to worry about, because this is a kind of iceberg.  We don't see it, because it's not really displayed by the vent. The vent resets after exhalation. So, if you have 6 ml, inspiration of 6 ml per kilo, then you don't have  complete exhalation.  Next, you will see, again, 6 ml, but the total volume actually will be more than that, and you have to wait for total passive exhalation to see  the whole volume that goes out of the lung of your patient. And I have some slides here. Yes. So, this is data I got from Marcelo Amato from the heart trial.  And so, a subgroup of patients had the monitoring, some tracings, because not all of the investigators sent, actually, recordings, strips to the investigators for the analysis.  And you can see here clearly that you have double breaths. So, we don't know. I think it's reverse triggering that you have.  Or maybe not, because you already have patient-triggered breaths. But actually, you have some activity, and you see that you have breath stacking here.  And this doesn't have the same impact on a patient with low PEEP. You can see here that you have, even with a tidal volume of 12,  the plateau pressure stays in what we suppose is a safe limit, lower than 30. But in this specific patient, it goes up to 15, and you have plateau pressure of 45.  If this happens a lot in the course of your patients, you create ventilation-induced lung injury. And this might not be seen.  If you don't pay attention at the screen, you will not detect that. And the respiratory therapist will charge the 6 ml as usual, and you will miss this point. Great.  Questions for Dr. Walker. Thank you for a nice presentation. I don't know if you've considered a translational study to accompany what you're looking at. Yes.  Actually, we have an IRB to get, well, first to get waste fluid from peritoneal dialysis patients, just to start to see what's there.  Now they're not really healthy patients, but at least we'll collect some samples. And then I would love to get clinical samples from patients who've  undergone abdominal surgery, either for reasons of sepsis or nonsepsis. That would be wonderful. Yeah. The question that I have is, and this is one that we struggle with when  we're looking at biomarkers of any sort, is that have you done any studies where you look at peripheral blood?  Is peripheral blood in any way useful in looking at some of these changes? Yeah. We actually have another study where we're looking at biomarkers in the  peripheral blood, not exactly the same thing we're looking at here. I do want to back up and see, is the trafficking pattern through the  blood changed in the animals that are destined to die versus destined to live? I guess the rationale behind starting this was that if we can identify  changes in cell movement that represent a functional response versus a non-functional response, then maybe one day we could develop a therapy  to try to tweak things in the correct direction. So I think when I started the study, it was more from that mindset than  using it as a biomarker, but I do think it would be really interesting to see if we could detect something in the blood.  We have a separate study where we're looking at biomarkers in the blood that they are markers of monocyte and neutrophil changes that could be used  for sepsis diagnosis and prognosis, perhaps. Yeah, because for many of these patients, we're sort of flying blind.  We're not really looking at any sort of immune modulation or immune change other than very basic white blood cell counts or differentials, which is not very helpful.  So it's very promising, very interesting. And the nice thing about flow cytometry is it can be done on a very small  blood sample and fairly rapidly as opposed to, say, like blood culture. Great. Thanks. Thanks. So I'd like to congratulate the speakers on their awards and grants  respectively and thank the audience for participating in this session.  Thank you.

Video Summary

The speakers in this session discussed different aspects of critical care medicine. Professor Menon presented on the management of traumatic brain injury and the challenges in current ICU treatment strategies. He discussed the need for better tools to titrate ICP and CPP targets on an individual basis. Dr. Walker discussed the detection and impact of dyssynchrony in mechanically ventilated patients with ARDS. She presented the results of a study on reverse triggering, which is when the patient's diaphragm contracts after the ventilator delivers a breath. Dr. Pham presented on the BEARDS study, which aims to better understand breathing efforts and dyssynchrony during mechanical ventilation in ARDS patients. He discussed the feasibility of automatic detection of reverse triggering using airway pressure and flow data. The study also plans to develop algorithms to manage dyssynchrony and test their effectiveness in a clinical trial. Overall, the speakers highlighted the importance of understanding and managing dyssynchrony to improve patient outcomes in critical care settings.

Asset Subtitle

Neuroscience, Sepsis, 2020

Asset Caption

The Society of Critical Care Medicine's Critical Care Congress features internationally renowned faculty and content sessions highlighting the most up-to-date, evidence-based developments in critical care medicine. This is a presentation from the 2020 Critical Care Congress held February 16-19, 2020, in Orlando, Florida, USA.

Learning Objectives

Review clinical care after traumatic brain injury and the resultant disabilities
Report findings related to sepsis and surgery changes in mouse peritoneal cavity myleoid cells
Explain BEARDS study results

ACCM Distinguished Investigator Presentation: The Critical Care of Traumatic Brain Injury: An Exercise in Clinical Kintsugi?

David K. Menon

SCCM-Weil Research Grant Recipient: Sepsis and Surgery Alter Populations of Myeloid Cells in the Mouse Peritoneal Cavity

Wendy Walker

Discovery Grant Recipient: BEARDS: Breathing Efforts in Acute Respiratory Distress Syndrome

Tài Pham

Meta Tag

Content Type Presentation

Knowledge Area Neuroscience

Knowledge Area Sepsis

Knowledge Level Foundational

Knowledge Level Intermediate

Knowledge Level Advanced

Membership Level Select

Membership Level Professional

Membership Level Associate

Tag Traumatic Brain Injury TBI

Tag Inflammation

Tag Oxygenation

Year 2020

Keywords

critical care medicine

traumatic brain injury

ICU treatment strategies

ICP targets

CPP targets

mechanically ventilated patients

ARDS

dyssynchrony

reverse triggering

	Society of Critical Care Medicine 500 Midway Drive Mount Prospect, IL 60056 USA Phone: +1 847 827-6888 Fax: +1 847 439-7226 Email: support@sccm.org	Contact Us About SCCM Newsroom Advertising & Sponsorship DONATE	MySCCM LearnICU Patients & Families Surviving Sepsis Campaign Critical Care Societies Collaborative	GET OUR NEWSLETTER
© Society of Critical Care Medicine. All rights reserved. \| Privacy Statement \| Terms & Conditions The Society of Critical Care Medicine, SCCM, and Critical Care Congress are registered trademarks of the Society of Critical Care Medicine.