Everybody, thank you so much for having me feel naked without my mask. All right, so I have a couple of disclosures, or one disclosure I'd like to make. The top half is just research funding, but I do want to highlight one item on the bottom, which is, as was mentioned, in the PI for a trial, an industry-sponsored trial from Sedona Medical. And prior to coming onto that role, I did do some consulting with them, and they're a company that makes a device that facilitates ICU ventilator adaptation to enable giving inhaled sedation. So just to acknowledge that, I don't endorse their products, I'm not a spokesperson, and the opinions are my own, and I don't mention their product anywhere in this talk. The other disclosure, which I want to emphasize, is that volatile anesthetics are not approved in the United States for ICU sedation, so please don't go try this at home if you're in the US. But it is approved in other countries, in Europe and beyond, and there's a fair amount that I think we can learn from their experiences as we start to gain experience here. So this first part won't go over fast since it was touched on with the first two speakers, but obviously when you're thinking about, or giving any talk on ICU sedation, one of the first questions to ask is whether it's really even necessary to give continuous sedation. And as we heard from the early presenters, there's a fair amount of data to suggest that patients will do better if you avoid unnecessary sedation. But what if the sedation is necessary? Then the question is, well, which one might you think about choosing? As we heard earlier, benzodiazepines are probably, or should not be your first line agent in most cases, in part because they're delirogenic, and they may delay timely extubation, and potentially time to wake up, though it's an inconsistent finding in RCTs as we heard about. And so that really leaves us with Propofol and dexmedetomidine, which are the two agents that are recommended in the current SCCM guidelines. And I'm not here to refute those recommendations at all, but to acknowledge that those drugs also are far from perfect. Both drugs, we all know, have side effects, including hypotension and bradycardia, the latter especially with dexmedetomidine. Dexmedetomidine also is probably not the best agent if you're targeting deep sedation. I think it's hard to achieve that sometimes if dexmedetomidine is a monotherapy. And as a culture, as we seem to give dexmedetomidine at higher doses for longer periods of time, I think we're likely to encounter more and more issues with the draw if the agent is abruptly stopped, and we've certainly seen that in our ICUs as well. For Propofol, there's at least a theoretical risk born in some lab-based studies of line infection because of the lipid-rich emulsion that it comes in. And then obviously there's a rare but very serious risk of Propofol infusion syndrome. So these are not the perfect agents. And so if you're spinning your wheels in the field saying, oh, we can try this one, or we can try that one, or we can try the third one, I think another thing that would be reasonable to do would be ask, well, what other drugs might we already have available that might be worth studying? And I think one of them is volatile anesthetics. So volatile anesthetics are a widely available class of medications. They've been around for over a century. A couple of the medications that we use to this day have actually been used for over 50 years. And we have extensive experience in the operating rooms using it for deep sedation during surgery, you know, with general anesthesia. But it hasn't been thoroughly investigated in the ICU. And I think a lot of that has to do with technical limitations of the differences between an ICU ventilator and an OR anesthesia machine. So one of the first things I think we need to do as a field if we're going to explore this is to stop using the word volatile anesthetics. It's really volatile inhaled sedatives. I'm a pulmonary critical care physician and I'm not giving general anesthesia. I'm qualified to titrate sedation in the ICU setting. And that's really the point of this type of modality is not to give general anesthesia in the ICU, but to give sedation. And I think the word volatile is actually very helpful for understanding how this works. So it's defined as evaporating readily, often encountered temperature and pressure. So what that means is that you have a drug that can be dispensed as a liquid and it evaporates very readily. So when you put it in line with the ventilator circuit and you have warm air blowing by it, it will evaporate. And once it's a vapor, it can easily be absorbed into the lungs into the patient. And there are really three medications, three volatile sedatives that are used in the OR setting. The first one, desflurane, is probably not appropriate for the ICU setting. And as far as I know, it's not actively being studied. The reason for that is the side effects are things you probably don't want to encounter in your critically ill ICU patient with restorative, like bronchospasm, extensive coughing, or increased secretions. You know, bronchorrhea would probably be problematic in a lot of the patients that we're ventilating. But sevoflurane and isoflurane both, I think, have fairly favorable properties that would make them interesting to translate into the ICU setting. And both are actively being investigated currently. One of the great advantages of volatile anesthetics, compared to a lot of the other sedation that we currently have available, is that they're eliminated by the lungs. So if you can breathe, even if it's breathing on a ventilator, as long as you're getting air in and out, you can be sedated with volatile anesthetics. And there's very, depending on the particular drug, there's very minimal metabolism that happens in the body. So essentially, you breathe it in and out, reach a steady state as your body absorbs some of it, and you don't have to worry about liver function, you don't have to worry about kidney function. It should be relatively unaffected in terms of the drug characteristics in patients with multi-organ failure. That being said, there's another definition for volatile that I think is worth highlighting, which is something that's characterized by rapid or unexpected change. So the stock market might be volatile, or I think pertinent to this talk, our knowledge of the long-term side effects of these agents, I think, is very likely to be volatile, right? We haven't really studied this in a rigorous way before, so it's quite possible that we'll discover new problems because surgeries don't last for three weeks, but ICU stays do. And as a specific example of new knowledge that's come out relatively recently about that is, as I said, in Europe, where some of these setups that we're about to talk about are approved by the EU and used clinically, patients are getting these sedatives sometimes for several weeks or even over a month on end. And there have now been a couple of case reports independently that have come out, and actually case series, that suggest that prolonged exposure to sivoflurane might increase the risk of nephrogenic diabetes insipidus. Now, does that mean you shouldn't give it? Does it mean there's something you should be monitoring? You know, I think we haven't worked it out yet, but that's something that never showed up in the OR literature that is a new discovery as we gain more experience with giving these drugs over time. So we don't know, but we don't know because we haven't studied it yet. So in terms of the mechanisms of inhaled sedatives, they target all the same receptors that other medications do, but their characteristics are slightly different. So unlike other sedatives, they hit multiple presynaptic and postsynaptic effects, and that's probably because they bind to ion channels, and so can both augment inhibitory signals, like from GABA or glycine, and inhibit excitatory signals, like your nicotinic acetylcholine receptors, glutamate, serotonin. So they have a multitude of effects on different receptors that might be relevant for sedation. So how do we take this experience that's quite extensive in the OR, as somebody there personally titrating the drug around the clock, to a situation where there's nobody at the bedside around the clock, and things sort of have to fly on their own to some extent? Well, I think in order to understand how this works, it's worth just, sorry, I can't resist this as a ventilator guy, very briefly reviewing the anesthesia loop circuit, and then you'll understand how simple that adaptation is to the ICU ventilator. So, I guess you can't see my mouse, but if you look at the top of the screen, you see there's an anesthesia machine with a vaporizer within it, and so gas leaves from the anesthesia machine, travels to the right, down towards the patient, right? So that's your oxygen, and your anesthetic gas going into the patient. Now when the patient exhales, it goes into the expiratory limb of the circuit, which is down at the bottom of the screen, and travels all the way back in a loop to the anesthesia machine. It passes through the loop and goes right back into the patient. So they're actually re-breathing the anesthetic gas that they exhaled, and so you achieve a steady state with just a slow, continuous infusion because they're constantly re-breathing the gas within the circuit. Now, the reason they're able to do that in the operating room is because the anesthesia machine also has a CO2 absorbent, so it binds the CO2 so they're not re-breathing it, and then also they can titrate in fresh gas flows, giving us a fresh oxygen, because the patient will obviously be consuming oxygen while they're breathing. So those sort of two adaptations, the fact that they're re-breathing the anesthetic gas, but they have a CO2 scrubber, and they get fresh gas flow are some of the unique properties of the anesthesia circuit. Now, if you look at just any regular ICU ventilator, you can make some simple adaptations, now with the current technology, to achieve a very similar effect. So, as you'll see in the top of the screen here, we have air going into the patient as the top limb, and then when the patient exhales, the limb on the bottom of the screen there is the expiratory limb. And so I want to highlight a few things here. There we go. So the first is, if you're going to adapt the ventilator, you need to give the drug, right? So you have drug and whatever, a syringe, and a pump to administer the drug, and that runs into the ventilator circuit, as depicted here. Now the next thing is, you need to modify the ventilator circuit so it can receive that drug. And so there's two aspects. One is that it needs to be what's referred to as a dry circuit, not a heated wire circuit. So a dry circuit means that you're used, like what you would use on a transport ventilator. So you have a heat and moisture exchanger, you know, that disposable thing that you change every 12 to 24 hours that sits by the Y connector of the ventilator. And the reason it needs to be a dry circuit for this particular setup is because of that third adaptation, which is the modified HME, the modified heat and moisture exchanger. And so what a heat and moisture exchanger does, the respiratory therapists in the room know this very well, and I hope I'm not embarrassing myself, but basically, when you're breathing air in and out, it needs to be warm and humidified, right? Because if it's cold and not humidified, it's going to dry out the airway and cause bleeding, irritation, et cetera. And so you can either do that by having the whole circuit and have warm, humidified air, or you can have a little contraption that sits there that sort of traps the heat that you're exhaling and retains the moisture, and so you can continue to have heated, humidified air breathed in. And so with that HME setup, if you modify it slightly by putting in a carbon filter, it can function as a HME, but also block the anesthetic gas from passing through it. And so that's the big adaptation, the big modification that was a technological advance that needed to happen. So in this case, if you look at the pump, number one up at the top of the screen, it's situated so that the infusion runs on the patient's side of that HME. And so when it goes into the patient and then the patient exhaled, all of the CO2 that they're exhaling freely goes into the expiratory limb, but that drug will bind to that specialized HME, if you will, and get stuck there. And then when they breathe in on the next breath, they're rebreathing back in that anesthetic drug. So it creates that same kind of rebreathing effect that you see with the loop anesthesia circuit, but without the loop. It just creates a loop only for the drug with that modified HME that has a carbon filter in it. Now it's not perfect. Roughly 10% of the drug of the volatile anesthetic or volatile sedative can pass through that filter into the expiratory limb of the ventilator. And so that's why we make the last modification, which is number four over there. So on the back of the ventilator, it's terrifying to think of this after a COVID talk, but on the back of the ventilator, there's an exhaust port that just spews all the air out. And so, I know, right? It's in New York, tell me about it. So, but in order to do this, you know, you don't want to spew anesthetic gas out into the room. So you can make an adaptation that just has a tube that literally sticks on the back of the exhaust and takes that, whatever's coming out of the exhaust down into a carbon reservoir that will bind the drug. And so that's all it is. And there's no ambient anesthetic leaving the room, even if it passes through that, into the expiratory limb of the ventilator. So the one other modification that's not depicted here, because it's optional, would be to include a gas sampling line, and you can measure the end concentration drug, of the drug if you want to. But I want to emphasize, in places that do that still, this is supposed to be given as a sedative. You should be titrating it to a target RAS, not to a target entitled concentration, which is actually not all that well correlated with RAS between patients. So that's the modification. It's just a simple modified HME, and something on the back of the exhaust of the ventilator, and then a pump to give the drug. So what's the clinical trials experience we have with this? Well, the most robust study done so far was from Andreas Meiser's group in Germany, where they studied 301 patients and randomly assigned them to isoflurane versus propofol. And these were patients who had some sort of acute need for the ventilator, so they were newly intubated. And the clinical team had decided they needed a goal RAS somewhere between minus one to minus four, and needed continuous sedation to achieve that target RAS. Both arms received opioids for analgesia, titrated to a pain scale. And the study treatment was just for the first 48 plus minus six hours. So basically just two days of comparing these two agents. The main outcome was just asking, can you do this to effectively achieve some level of sedation short of general anesthesia? And the answer was yes, so that target RAS between minus one to minus four, you're equally able to achieve that with propofol versus with isoflurane. The mean RAS score, though, in both arms tended to be on the deeper side, so I don't know if this study answers the question of whether you can target a RAS minus one or minus two very well, because both groups had deeper the RAS than perhaps what you would target in most of your patients if you decided to give continuous sedation. But I think one of the exciting findings, potentially from this trial, was the wake-up time with interrupted sedation. So while there was no difference on the first day, on the second day, patients who were randomly assigned to isoflurane, which is that top line, that purplish line, woke up significantly faster. Now the median difference was only 10 minutes. It's hard to know if that's clinically significant, but if that same type of effect accumulates over time, you could imagine this could be a clinically relevant endpoint. The other interesting part of it, sort of relevant to our earlier talks, was the opioid requirement actually diminished with patients that were assigned to isoflurane compared to propofol, and that decreased opioid requirement occurred despite no change in pain. So they measured that behavioral pain scale and saw that it was similar between patients, one group needing less opioids to get there. Other things, just to highlight in this article, were that there were no meaningful difference in mortality, no meaningful difference in need for vasopressors, and multi-organ dysfunction seemed comparable between the two groups as well. But importantly, long-term cognition was not assessed. So what about in patients with severe lung injury? That was patients that, for the most part, had reasonably healthy lungs. Maybe they had pneumonia or something, but it seemed like they had probably adequate gas exchange. So there was another trial done by Matthew Jovedone's group in Clermont-Ferrand Fronts that enrolled patients with moderate severe ARDS. And now we're talking about patients with a P to F ratio of 150 or less. It's 50 patients, they randomized them to sivoflurane versus midazolam, and they targeted a RAS of minus five with the intent then of giving paralytics. So if you can give sivoflurane to hit a RAS of minus five, then clearly it works in this patient population. And the objective actually wasn't to evaluate its effects on sedation. There's some animal data that, in the interest of time, I'm not gonna go over now, that suggests that it might have lung protective effects in terms of the alveolar epithelium and endothelial barrier function. So RAS minus five was readily achieved with sivoflurane. There was no issues doing that, which is not surprising, since that's what anesthesiologists do every day. But the effects on P to F also suggested perhaps P to F ratio was improved in patients on sivoflurane. Hardly a patient-centered or even clinically meaningful endpoint, but I think when you pair that with the biological data that they have, it raises some intriguing hypotheses. So they looked at soluble rage, which is probably the current biomarker du jour for alveolar epithelial injury, and found that both plasma and BAL levels of soluble rage were diminished in patients assigned to sivoflurane, suggesting potential lung protective effects. They also looked at inflammatory cytokines and saw similar signals both in the lung and in the plasma. So that's sort of the multicenter clinical trials data that's available. There are some smaller studies that, in the interest of time, we're not going into, but showed relatively similar signals. But I do wanna highlight, there is one serious safety signal, or safety risk that can happen with volatile anesthetics in general, and that's malignant hyperthermia. The clinical presentation typically, the high fever, even though it's in the name, is one of the later manifestations. But things like hypercapnia, muscle rigidity, signs of rhabdomyolysis, lactic acidosis, some scary stuff that can happen with patients and that anesthesiologists are intimately familiar with. But when you compare malignant hyperthermia to PRIS, it doesn't seem like it's such a bad trade-off if the question is should we give propofol or should we give volatile inhaled sedatives. So the bottom line is really the take-home, which PRIS has a mortality rate in the literature of roughly 50%. With malignant hyperthermia, it's less than 5%. Malignant hyperthermia has an antidote. These are patients in the ICU that are having continuous vital signs. Chances are you would pick it up and be able to intervene. And the MH incident is substantially lower. We're talking one and a quarter million people. And as we do in our trials, you would screen the patient or the surrogate to ask for risk factors related to MH, which you can't really do for PRIS. I do wanna touch on briefly also staff safety because this is something that often comes up. Is it safe for me to be caring for this patient? And there are several studies that have looked at this as well. I'm gonna highlight one in particular from Peter Saki's group. So they looked at basically patients who were using one of the setups similar to what I showed in the prior picture. They looked at ambient drug exposure within the room. And they had a sensor in that measured the level of exposure, the amount of isoflurane that was in the room. And in no patient room did it ever cross the limit that the US government says it shouldn't cross. In fact, it didn't really even come close to that. The US limit is two parts, the US nine hour limit is two parts per million for one hour. They experienced like one part per million for 10 minutes. And that was in the setting of things like a circuit disconnect. So of course, if you're completely disconnecting the ventilator circuit, there's gonna be a little gas that leaks out, but it's not gonna cause the nurse or RT to faint if there's a little bit of gas that leaks out. Having been in the room and done that multiple times myself, I can tell you haven't fainted yet. But also in the study, I did something else I thought was really neat which was had the ICU staff caring for the patient wear lapel pins would actually measure directly for that individual, what kind of exposure they had. And you can see again, the NIOSH limit is two parts per million over an hour. And the cumulative exposure that these folks saw was like 0.1 parts per million, like not remotely close orders of magnitude less than what would be considered possibly an unsafe exposure with the US government limits being even more stringent than what the European limits are where the study was done. So I think there's really negligible staff risk for this. And the sort of the warnings around pregnancy exposure, again, you're thousands of times off from risk to somebody who's pregnant and caring for that patient. So but there are obviously many key unanswered questions. I highlighted a few here. In my mind, I think the most important one is really the long-term cognitive effects. But I also think there are some practical questions about set up a maintenance of the anesthetic reflector. Because now if you think about ICU workflow, it's not just the nurse potentially who's responsible for the sedation setup, but also potentially, at least the US environment or places where there are respiratory therapists, it would be the respiratory therapist possibly managing that sort of modified HMB. So that makes the management of the setup a little more complicated, although presumably it would be the nurse titrating the medication, as is the case in the trial that we have ongoing now. There are also several therapeutic questions I think need to be answered before we would consider this for prime time. But on the bright side, there are several multi-center trials underway, including two here in the US. One that I and Brian O'Gara from Beth Israel Technics are leading, and another one that Tina Bonsack and Chris Hughes, Brenda Pun and the Vanderbilt team are leading that are FDA registration trials to ask if we can do this safely here in the US. And those trials importantly have the rigorous battery of long-term cognitive assessments that are being done at Vanderbilt that will help answer some of these safety questions. There's also some other interesting trials going on in Europe, one actually looking at giving this to children. And then there's another ARDS trial trying to flush out that potential safety signal, or sorry, that potential beneficial signal for lung protection in patients with moderate to severe ARDS. So in conclusion, inhaled sedation is technically feasible. I think we've overcome that barrier. And there are studies going on now in the US to ask whether this is something that might be beneficial here. I think there are preliminary intriguing, the preliminary findings are intriguing, including faster wake up, diminished need for opioids, and maybe some biological data suggests there's benefits to both pulmonary and extra-pulmonary organ effects. But those are all extremely preliminary and absolutely need to be validated before we sort of hang our hat on this. There are definitely key knowledge gaps to be addressed. I think the biggest one is what happens with prolonged exposure to these drugs, because much like the dexmedetomidine experience, it'll be FDA approved for 24 or 48 hours, and then somebody's gonna give it for three weeks. And so I think we need to start to understand that before we can do it responsibly. And then the last thing is that, clearly this is not ready for prime time, but I think if these studies show not just non-inferiority versus propofol, which is not a particularly interesting question, but if they show a patient-centered benefit, then perhaps it's worth trying to change the ICU culture and set up around what agents we might reach for for our first-line sedatives, but we're not there yet. Thank you for your time.

volatile anesthetics

sedation

ICUs

benzodiazepines

propofol

dexmedetomidine