Hello and welcome to today's Journal Club Critical Care Medicine webcast. We understand that these are challenging times for everyone, so we very much appreciate you taking time out today to participate. Today's webcast, which features articles from the March issue of Critical Care Medicine, is hosted by the Society of Critical Care Medicine. It is sponsored today by the Penn TTM Academy, which is an educational program led by expert faculty from the University of Pennsylvania. It focuses on post-arrest care and targeted temperature management. For more information, please visit pennttm.com. That's p-e-n-n-t-t-m.com. The webcast will be available to registrants on demand within five business days. Log on to mysccm.org and navigate to the My Learning tab. My name is Tony Gerlach, and I'm a clinical care pharmacist at The Ohio State University Wexner Medical Center in Columbus, and I will be moderating today's webcast. Just a few housekeeping items before we get started. First, during the presentation, you'll have the opportunity to participate in several interactive polls. When you see a poll, simply click on the bubble next to your choice. Second, there will be a question and answer session at the conclusion of both presentations. To submit questions throughout the presentation, type into the question box located on your control panel. If your question is for a specific presenter, please indicate that. Third, if you have a comment to share during the presentation, you may use the question box as well. Fourth, everyone joining today will receive a follow-up email that will include an evaluation. Please take five minutes to complete the evaluation. Your feedback is greatly appreciated. And finally, please note the following. This presentation is for educational purposes only. The presentation is intended to represent an approach, view, statement, or opinion of the presenter, which may be helpful to others. The views and opinions expressed herein are those of the presenters and do not necessarily reflect the opinion or views of SCCM. SCCM does not recommend or endorse any specific test, physician, product, procedure, opinion, or other information that may be mentioned. And now I'd like to introduce today's presenters. Dr. Nick Johnson is an emergency physician and intensivist. He received his medical degree from the University of California, San Francisco, and completed residency and chief residency in emergency medicine at the University of Pennsylvania. He then completed a fellowship in critical care medicine at the University of Washington. He cares for patients in the emergency department and intensive care units at Harborview Medical Center. Dr. Johnson studies cardiac arrest with a focus on oxygenation, ventilation, lung injury, and extracorporeal support. He is interested in determining how mechanical ventilation strategies might impact outcome after cardiac arrest and how extracorporeal technologies will be used to support patients with failing hearts and lungs. He's received funding from the National Institutes of Health and Medicine One Foundation. Additionally, he serves as an associate program director for the critical care medicine fellowship and has received numerous teaching awards. Dr. Kelly Sawyer is an assistant professor of emergency medicine at the University of Pittsburgh Medical Center. After her emergency medicine residency, she completed an emergency cardiac care and resuscitation fellowship at the Virginia Commonwealth University Hospital in Richmond. Her research has focused on optimizing in-hospital TTM practices and more recently, bringing awareness about the need for systems and research to improve survivorship after cardiac arrest. Thank you for both joining us today. Before we begin, could each of you tell us if you have any disclosures to note? Nick? I have no financial conflicts of interest. I do receive some research funding, as you said, from the National Institutes of Health and the Medical One Foundation. Thank you very much. And Kelly? I have no relevant disclosures. Thank you very much. Okay, then let's get started. Nick, we'll start with you. Great. Thanks, Tony. Thank you to Tony and to SCCM and to the Penn TTM Academy for the opportunity to discuss our recent paper, Targeted Temperature Management at 33 versus 36 degrees, a retrospective cohort study. Again, my name is Nick Johnson. I'm from the University of Washington. I think we already covered disclosures, so I'll skip this. First off, I'd like to thank a really excellent team who contributed in a massive way to this paper. Like all research studies, this was a big effort and I couldn't do it without the people listed on this slide. So I think actually both of us today are going to be talking about this concept of Targeted Temperature Management Dose, and when we refer to dose of Targeted Temperature Management, we're generally referring to the depth or temperature goal times the duration. So we'll be focusing on the dose of TTM today. Just as some background information, there are four key randomized trials that have looked at this question of Targeted Temperature Management Dose. The first two were seminal trials that were published in the same edition of the New England Journal all the way back in 2002, the Bernard and Haka studies, and both of these studies compared a goal temperature of around 33 degrees to normal thermia, which was not actively maintained, so just usual care, for a duration of 12 or 24 hours, depending on the study, in a specific cohort of patients without a hospital cardiac arrest due to ventricular fibrillation or tachycardia. And these studies included a total of 350 patients, and both found that cooling patients to the range of 33 degrees led to an improvement neurologically in tachysurvival, the Bernard study at discharge and the Haka study at six months. These two studies really informed clinical practice and guidelines for about the next 11 years, until the TTM trial was published in 2013, which compared the goal of 33 degrees to 36 degrees in a much broader cohort of patients with out-of-hospital cardiac arrest from all cardiac rhythms. The TTM trial was much larger, including 939 patients total, and as you all likely know, found no difference in their primary outcome of all-cause mortality at 90 days. And most recently, the Hyperion trial, published just this last year, compared in a cohort of patients with non-shockable cardiac rhythms, so PEA and asystole, a goal temperature of 33 degrees with 37 degrees. They included 584 patients total, very high overall mortality as expected in this non-shockable rhythm group, and found a 4% absolute difference in their primary outcome of CPC1 or 2 at 90 days. But the story of our study really has to do with publication of the targeted temperature management trial. After this study was published, there was pretty wide uptake of a 36 degree goal, as evidenced by a number of survey studies and also studies of clinical practice showing that institutions were pretty rapid to take this goal into clinical practice. So that leads to our first polling question. So what goal temperature do you currently use after cardiac arrest? And we'll say all types, out-of-hospital, in-hospital, all rhythms, just on average. So we'll give people a minute to answer the poll here. Great. Well, the group is quite evenly divided. A third saying 33 degrees, a third saying 36 degrees, and a third saying individualized to the patient. So I have a suspicion that this is a group of enthusiasts who's really read the literature on this topic. And I think if you looked at overall landscape, there was pretty wide uptake of that 36 degree goal after the TTM trial was published. So for some more background, a couple of years after the TTM trial was published, studies like this one started to trickle out in the literature. This was a before and after study done at the Alfred Hospital in Melbourne, Australia, where they compared outcomes and processes before and after they implemented a 36 degree protocol. They looked at just 76 patients with ventricular fibrillation out-of-hospital cardiac arrest and found a number of interesting things. After they changed their local protocol from 33 to 36 degrees, the proportion of patients receiving active cooling in the ED dropped by 23%. The absolute percentage of patients who received cooling in the ED dropped substantially. The time spent below target temperature also decreased. The incidence of fever increased by about 20%. And there was a trend toward a decrease in survival that didn't reach significance. Not far after that, a large study was completed using the Australia-New Zealand Resuscitation Outcomes Consortium database. This study included over 9,500 patients with out-of-hospital cardiac arrest. And they also compared both processes and outcomes before and after the TTM trial was published. The study included about 4,400 patients before the TTM trial was published in December of 2013 and 5,100 patients after. On the process side, they noted a couple of interesting things. After the TTM trial was published, average temperature increased quite substantially and quite abruptly by about one degree. And there was an increased incidence of fever as well. In this figure here, the blue line represents publication of the TTM trial. And note the axes. So these are pretty big spaces between temperatures. The temperature had been decreasing on average, and then after publication, an abrupt increase. Similarly, they also looked at outcomes after publication of the TTM trial. And what they had noted is that there had been a secular improvement in mortality from out-of-hospital cardiac arrest in Australia and New Zealand. And the slope of that improvement abruptly changed after publication of the TTM trial. Obviously, these are observational data, so we cannot infer causation, but certainly a concerning trend. So second polling question, did you or did your institution change goal temperature after publication of the TTM trial? All right, so the vast majority, 70% or so, said there was some change made with a handful of nos and some others. Maybe the protocol was modified to some degree but not changed completely, and this is consistent with what's out there in the literature. There was quite a rapid uptake of the 36-degree goal after publication of the TTM trial, perhaps much more rapid than a lot of evidence is translated into clinical practice, which was an interesting phenomenon. So my institution was no different than 70% of yours. Our TTM protocol, and we've always had only one TTM protocol with a single temperature goal. Back in 2002, at time of the publication of the Bernard and Hacke studies, we had a 33-degree protocol that ran all the way up until April 7th of 2014, where we made an institutional change to 36 degrees. During both of these periods, there was only one protocol with one goal temperature active, and all of our closed-loop temperature devices, and we used both surface and endovascular devices at our institution, were pre-programmed to the goal temperature. Our hypothesis was that a TTM goal temperature of 36 degrees would be associated with no difference in neurologic outcome at hospital discharge following out-of-hospital cardiac arrest compared with 33 degrees in keeping with the findings of the TTM trial. We conducted a retrospective before-and-after study of patients admitted with non-traumatic out-of-hospital cardiac arrest over a seven-year period to our hospital, which was an urban academic level one trauma center. We included only patients who received targeted temperature management in our final analysis cohort. Our primary exposure was whether patients were treated before or after we changed our protocol from 33 degrees to 36 degrees, and our primary outcome was favorable neurologic status at hospital discharge, which we defined using a cerebral performance category score of one or two. We had a number of secondary outcomes, including hospital mortality and a number of TTM care processes. Looking at our patient characteristics, we treated a total of 782 out-of-hospital cardiac arrest patients during the study period, but 151 of those died in the ED, and another 178 did not receive targeted temperature management, leaving 453 patients for our analysis cohort. Those patients were pretty evenly split into those who were treated during the 33-degree period and those who were treated in the 36-degree period, and we'll talk about outcomes a little bit later. Looking at our table one, just basic demographics, I've highlighted a few select ones here. The TTM 33-degree cohort was slightly older, by about five years on average, and there was no difference, notably, in shockable rhythm, but there was a difference in cardiac etiology, and this variable was coded after the fact by one of our expert-trained coders. Interestingly, these usually track pretty tightly together, and there was maybe a trend toward more shockable rhythms during the 33-degree period that didn't reach significance. In terms of other aspects of cardiac arrest care, the groups were pretty balanced. EMS response time was quite fast in Seattle at four minutes. Almost half of patients received bystander CPR, the majority were intubated in the prehospital setting. About a third or so received coronary angiography, and mechanical circulatory support was fairly infrequent, no statistical difference between the groups. The older cohort got more balloon pumps, because we were doing that at the time. And then a few differences in terms of neurological consultation and in-hospital DNR status. This is a graph plotting lowest temperature achieved over time. You can see that the dotted line here denotes when the clinical protocol was changed. You can see that lowest temperature achieved over time jumped fairly abruptly from the TTM 33-degree period, where patients clustered around 33 degrees, to the 36-degree period, where patients clustered around 35 or 36 degrees. So moving on to our outcomes, these are unadjusted outcomes, so our unadjusted primary outcome of neurologically intact survival, CPC 1 or 2, there is a 10% absolute difference in neurologically intact survival, with neurologically intact survival being higher during the 33-degree period. A number of unadjusted secondary outcomes, survival to hospital discharge was not statistically different between the two periods. TTM was initiated in the ED much more frequently during the 33-degree period. The time from EMS call to initiation in hours was also much faster in the 33-degree period, 1.9 versus 3.5 hours. And the time to lowest temperature in hours was actually longer during the 33-degree period. Most patients came into the emergency department with a presenting temperature of 35.5 degrees on average. Looking at our multivariable models, we found that TTM at 33 degrees was associated with a higher odds of favorable neurologic outcome at the time of hospital discharge with an odds ratio of 1.79. This is after adjusting for a number of covariates that you see here, age, sex, cardiac arrest location, witness status, bystander CPR, shockable rhythm, and EMS response time. Interestingly, when we substitute primary cardiac etiology for shockable rhythm in this model, it essentially is unchanged. TTM 33 was not, however, associated with any difference in survival to hospital discharge. The odds ratio was 1.46, but the confidence intervals crossed 1. We adjusted for the same covariates. This study has a number of key limitations. First of all, this is a retrospective study. And just based on the nature of a before and after retrospective study, there's a large potential for residual confounding or secular change that we could not account for despite our attempts to adjust. This was conducted in a single hospital in a single EMS system, and therefore, our results might not easily be generalized to other systems. We also didn't have a granular way to measure adherence to the protocol. Many of the TTM randomized trials track hourly temperatures. We were not able to do that in the QI database that we used. And then our primary outcome was a CPC score, which is potentially subject to bias, although we did have a single trained abstractor abstracting all of the data over the seven-year period. So in conclusion, TTM at 33 degrees was associated with a higher odds of favorable neurologic outcome in hospital discharge than TTM at 36 degrees. There are a number of possible explanations for this. One is that we found that there was earlier, more aggressive application of targeted temperature management in the ED. One concern that we have is that if TTM isn't applied early, there's a window, a lag between the time the patient arrives in the emergency department and they end up in the ICU, where the TTM definitive device might be placed. Even if the patient arrives at 35.5 degrees, that opens a window for fever, which we know is detrimental in all types of brain injury, and especially hypoxic ischemic encephalopathy, the type of brain injury that exists after cardiac arrest. It's also possible that TTM protocol tracked with other secular changes. There could have been other things about our care or about our patients that have changed over time that we couldn't account for. One possibility, although we didn't see this in our data, is that there were more opioid overdoses, for instance, in more recent years. It could also be that TTM was a marker for other care practices changing over time, especially withdrawal of life-sustaining therapy. We didn't see this in the data, but it's possible that those practices changed as well. And lastly, it could be that there's some biological effect of lower temperatures, and that was different in this study than in the TTM trial. One possibility is that our cohort actually looked a lot more like the Hyperion trial, that study of non-shockable rhythm patients, than it did the cohort in the TTM trial, which was overwhelmingly populated by shockable rhythm patients. So it could be that our patients had much more profound ischemia reperfusion injury, and therefore benefited more from being cooled to lower temperatures. Importantly, we did not find a difference between target temperature and survival of hospital discharge. So what are we going to do in the future? Well, based on looking at these data back in 2017, from a quality improvement perspective, we actually made a change again in our clinical protocol and went back to 33 degrees. So now we have the opportunity and have set up a nice little natural experiment to take a look at whether outcomes have changed after we went back to 33 degrees, and we plan to do that soon. So our final polling question, for me at least, is what goal temperature should we predominantly use now after cardiac arrest? And the options are 33 degrees, 36 degrees, 37 degrees, which would be controlled normothermia, or should we just individualize to the patients, or some other response? All right, well, it looks like I've convinced some of you, at least, that 33 degrees might be the right temperature. I think there's no right answer here. Again, this is an observational study that should not and cannot overturn randomized data, but I think it's a compelling paper to perhaps argue that a one-size-fits-all approach is not the right approach. Well, thank you all for listening. I'm going to hand off to Kelly now. All right, thank you so much, and thank you to our sponsor and the Journal of Critical Care Medicine for this opportunity. I'm going to build upon what Dr. Johnson has just so elegantly discussed from a clinical standpoint, and I'm also going to start by giving additional background and discuss some of the pilot work that led to this study, the relationship between duration of TTM, ischemic interval, and good functional outcome from cardiac arrest. And then I'll walk through the methods and results of the study and finally summarize and give future directions. I have no personal conflicts. Some of my co-authors, of which there are many, and I am grateful to the Resuscitation Outcomes Consortium for the data available, had some disclosures listed on this slide. So just some major summary takeaway points regarding the concept of TTM in general after cardiac arrest. We know it's become commonly accepted that optimum treatment dose is unknown. Timing, depth, duration, and rate of TTM are difficult to study, and investigations typically choose only one of these dimensions for clinical intervention. Data from preclinical studies seem more consistent. Earlier, faster, and colder confers more benefit in a lot of preclinical trials. Translating preclinical data to the clinical space, however, has not been easy, and this is often a result of variable arrest and treatment conditions, as well as heterogeneous patients with comorbidities. Improving institutional protocols for uniform TTM procedure has improved post-cardiac arrest care, especially at centers with less expertise or low volumes. These same hospital factors, however, may make individualized treatment more challenging as well. Finally, optimizing individual patient treatment and tailoring TTM will require specific treatment targets that are easy to measure and reliable. Thus far, I think we can have an impact on cardiac arrest survival by preventing cardiac arrest and optimizing cardiac arrest systems, reducing ischemic reperfusion injury, also reducing delayed neurodegeneration and inflammation. The literature discusses several tools that may help impact ischemic reperfusion injury and inflammatory thrombotic mitigation, and certainly novel targets are needed. The idea that hypothermia as a treatment exerts influence on many pathways has deep roots, and this is a concept that is in contrast to just simply avoiding hyperthermia. Next, I'll highlight a couple of preclinical studies that really help build the foundation for some of the additional work I'll discuss today. In this study, using a rat model of asphyxia, there were several experimental groups. There was a normothermia group, and then several hypothermia groups that had therapeutic hypothermia initiated at various time intervals after ROSC, including at time zero, one, four, and eight hours. And then there were also two different durations of hypothermia studied, 24 and 48 hours. People with good neurologic function was observed in the hypothermia groups initiated within four hours of ROSC, regardless of duration. However, benefit was observed in the hippocampal CA1 neurons in all the hypothermia groups compared to normothermia. In this case, 48 hours demonstrated benefit over the 24-hour duration. So as this study, I think, highlights, the outcome of interest matters, and we find varying results, for example, good neurologic function versus what happens in the hippocampus when we measure the impact on these different outcome measures. This study, I think, really highlights the potential therapeutic benefits of temperature management across the spectrum of early resuscitation. Again, this was a rat asphyxia model, and there were four treatment groups studied. They had a normothermia group. Second group was hypothermia initiated at one hour after ROSC and maintained for 24 hours. The third group had hypothermia initiated intra-arrest, so very early, but maintained for only one hour, and the fourth group had hypothermia initiated intra-arrest and maintained for 24 hours, and it's in this group that the researchers observed the best survival as well as the most highest level of good neurologic function as well as the least neurodegeneration. So the authors' observations led them to postulate that perhaps the early initiation helps reduce neuronal death and thereby mitigate ischemia reperfusion injury. In combination, the expanded duration of hypothermia treatment allowed for reduction of neural death and amelioration of inflammation. So thinking about this in humans, perhaps we need to consider the interplay between depth and duration as it pertains to timing in order to optimize both windows of opportunity, and this would require a three-dimensional study, again, with a reliable treatment target. So this is our first question for me. What would you target to tailor TTM for your patient? If you had something available, would you choose something that then allow you to tailor therapy, whether depth or duration, to your individualized patient? There we go. All right. All right. A nice spread of answers, and we'll come back to this towards the end. Some of these have been studied in small trials, and then letter A, time pulses, is actually the one that we used in our study. All right. So in the absence of another biomarker or treatment target, it does seem reasonable that a clinical target would be useful to titrate according to patient phenotype, and so ischemic interval or time to ROSC seems like a possible target, and that's what we used in our pilot work here, cited here, published in 2014. We used data from a single center in Richmond, the Virginia Commonwealth University Arctic Database, and because of preclinical studies, as mentioned, talking about or demonstrating that earlier and quicker TTM was beneficial, the city of Richmond at the time was providing cold saline in the field, and our institution had a single protocol with a target temperature of 33 degrees, and then we cooled our cohort for 24 hours and then had controlled rewarming, usually over about 36 hours. So in thinking about how to tailor therapy, we hypothesized that there was a relationship between the hypothermia duration and the ischemia duration, and we called this the hypothermia to ischemia ratio. We only evaluated patients presenting with a shockable rhythm, given the clear survival differential between presenting rhythms and patient phenotypes in our cohort, and our pilot results ultimately suggested that there was a significant association between a larger hypothermia to ischemia ratio and survival from adiposphalocardic arrest, suggesting that patients with a longer ischemia time may benefit from longer hypothermia time. And this brings us to the study that I'm going to go through more in depth. Building on the pilot work, we aim to investigate the association between the hypothermia to ischemia ratio and functional outcome in a much larger and more diverse population. We considered for inclusion those out-of-hospital cardiac arrest patients screened for the Resuscitation Outcomes Consortium ELPS trial, and those patients were screened between May 2012 and October 2015. Patients were included if they survived a hospital with a recorded time of return of spontaneous circulation and documentation of TTM. This figure highlights a lot of the exclusion criteria, and remember this is a retrospective secondary analysis of a database, so we excluded patients who died within four hours of hospital arrival, taking that as a surrogate for those who re-arrested or had early withdrawal of therapy. We also excluded those who had no intubation documented, again as a potential surrogate for being awake, and then of course those who had no TTM documented or missing TTM status. Our outcome of interest was survival to hospital discharge with a good functional outcome, and this was defined by the modified Rankin scale and abstracted routinely by structured chart review. The ischemia time, and this figure sort of highlights the time point and ratio makeup, return to ROSC was defined as time of arrest or 911 call to the return of spontaneous circulation, and total hypothermia treatment time was measured from initiation of cooling through TTM treatment time and re-warming, and that is laid out here at the bottom. This table highlights the main patient characteristics. The box highlighted in yellow are those who received hypothermia compared to those who did not receive hypothermia at all patients. Those receiving TTM had a higher proportion of initial rhythm of ventricular fibrillation or pulseless ventricular tachycardia, they were more likely to have a public location of arrest and a lay rescuer witnessed arrest, age, gender, time from 911 call to first EMS arrival and performance of lay rescuer CPR were similar between those receiving TTM and those who did not. Among out-of-hospital cardiac arrest patients who received TTM, approximately 29% were discharged with good functional outcome. These figures, these histograms show us the distribution of our variables of interest, those that make up the hypothermia to ischemia ratio. Median interval from 911 call to ROSC was 21.1 minutes, and median duration of hypothermia was approximately 33 hours. While participating hospitals were given best practice guidance, hospital processes for TTM delivery and management were not strictly controlled in the parent trial. This is a table of a lot of our main or primary results, and the all arrests column is our main results. You can then see that we conducted additional sensitivity analyses based on known factors relevant to survival, such as initial rhythm and witness status. And this highlights that across these sensitivity analyses, there's really little change in the odds ratio of hypothermia to ischemia ratio across these groups. This figure plots our data according to ischemia and hypothermia durations, and the lines represent different hypothermia to ischemia ratios. And you can see with an increasing HI ratio, there's a higher proportion of good functional outcome compared to those at lower ratios. Our data are summarized here in another way, and this table shows the proportion of good functional outcome according to terciles of ischemia and hypothermia durations. You can see that the data suggests a non-linear relationship since the medium TTM group fared best regardless of ischemia duration. This may be related to patient cohort characteristics or unmeasured confounders, or may just represent the natural way that clinical data and patient data present themselves, and we can't really conclude much more than that, but it's an interesting finding in this data set. This table is important, and it presents a sensitivity analysis with results of the model that includes hypothermia time and ischemia time as main effects, and then the HI ratio as an interaction term. And I've highlighted the odds ratios for these three variables, and you can see that the odds ratio for the interaction term, the hypothermia to ischemia ratio, is actually not significant, and it seems that in this cohort, the denominator does drive our results as far as we can conclude. So the strength of this investigation is its generalizability. We have patients that represent many different sites across North America with significantly larger variations in inclusion criteria, duration of ischemia, and methods of TTM practice. The ROCK-ALPS trial was large enough to allow for more meaningful sub-analyses that were not possible earlier in our pilot work. However, it was not that the ROCK-ALPS trial was not a TTM intervention trial, and so we are limited based on process data and granular details regarding TTM. And then, interestingly, we have observed a non-linear relationship between ischemia duration, hypothermia duration, and functional outcome, giving some signal that this is worthy of prospective study. This trial, or this study, was recently published and evaluated the same hypothesis. This is a sub-study of the TTH48 trial that was the 24 versus 48 hours of duration of TTM published a few years ago, and you can see the figure of their results is nearly identical, at least in gross form, with our figure as well, showing that perhaps there is this signal for a relationship between ischemia and hypothermia durations and outcome. The poll we had gone over, and I posed the question of how you might tailor your therapy, and there were some answers regarding other clinical data, and these four studies actually took a look at some of those same questions. In the top left corner here, a sub-study of the TTM trial evaluated the relationships between time to ROSC and treatment arm and outcome. On their secondary analysis, they found no interaction between temperature and time to ROSC and mortality, although if you look at their data, the fourth quartile was quite broad, ranging from 40 minutes to 170 minutes. More study is needed based on data available from that trial and the patients they included. So perhaps the time to ROSC isn't the best marker for severity of post-cardiac arrest syndrome. This study down here in this corner was a small study that evaluated the interaction between hypoxic changes on CT and targeted temperature management regarding good CPC outcome, and they concluded that in the absence of CT imaging evidence, TTM is better than normothermia. This potentially suggests that CT evidence of hypoxic encephalopathy indicates that perhaps TTM won't help ameliorate the damage that's been done, but certainly more study is needed. And then this top right corner, this group of researchers looked at using lactate as a potential marker to then stratify therapy, and so this was a post hoc analysis of a Japanese registry, and patients were stratified by lactate level after ROSC, and they also then stratified based on target temperature group. So they had three groups of lactate, severity, mild, moderate, and severe, and the two temperature groups based on their registry, 32 to 34, and their 33 to 36 group. And so what they found actually was that with a severe lactate group, which is defined as greater than 12, a deeper TTM target of 32 to 34 degrees conferred benefit. So perhaps titrating TTM temperature or depth based on initial lactate may be a consideration in the future. And then finally, interestingly, this is another sub-study of the TTH48 trial that looked at their two subgroups of patients, and those who survived with CPC1 and 2 were invited back for neuropsychiatric testing at six months, and they found that patients in the 48-hour group performed better at six months with regard to elements of cognitive testing compared to those treated in the 24-hour group. This is our final question. And so the question is, there are many limitations to using databases, and what sort of limitations do we face? I would be surprised if the answer is not uniformly single. There we go. And this is true. This is true for the study I just presented with regard to the hypothermia to ischemia ratio within the ROCK-ALPS database. Okay, so all these are issues, and certainly there were unmeasured confounders, including clinical factors that we could not address or account for, which may have had some influence on our primary outcome. So what are our future directions, regardless of where we stand on TTM, we do need better process data in order to study it more consistently across sites, across studies. The combination of consistent and target temperature and inconsistent definitions makes it difficult to synergize registry studies and prior observational trials. Contemporary clinical protocols tend to start a 24-hour duration clock from the time that target temperature is reached, such that if the induction phase is long, the total duration of TTM may be significantly greater than 24 hours. So the definition of TTM needs to be consistent, and we need to start thinking about how to think about it in a three-dimensional way. We need to modify our animal models so that we can better model clinical scenarios, and we need a more immediate and less time-dependent measure of ischemia reperfusion injury so we can replace the limitations of time to RAS as a treatment target and attempt to study tailoring post-critical RAS therapy. In conclusion for this discussion, a larger hypothermia to ischemia reissue was associated with good functional outcome, but in this cohort, it does seem that the denominator is playing a large role. Thank you so much for your attention. Thank you both for your interesting presentation. I see we have a few questions, and for the audience, feel free to send us more while we work through the first couple. The first question I have is, can you both comment on the 2015 American Heart Association recommendations regarding early cooling? Specifically, they recommend against routine pre-hospital cooling of patients after ROSC with rapid infusion of cold intravenous fluids, and specifically, they wanted to know, are there other methods we should be looking at for pre-hospital cooling, specifically intranasal cooling? I guess I can lead off there, Tony. I think that's a really good and interesting question. I think I can just re-summarize, so if we think that time to target temperature or time to cooling is important, why not push that out even further into the pre-hospital setting? I think it's a totally logical argument. Unfortunately, the studies to date have not entirely borne out that pre-hospital cooling is beneficial. The one conducted in my system by Francis Kim in Seattle in 2014 is the one that informed some of the 2015 guidelines. I think part of the criticism of that study is the method of cooling, which was infusion of cold intravenous saline. Patients got a fair amount of volume in the pre-hospital setting, and there was no difference in the primary outcome of the study and also an increased incidence of adverse events in the flavor of pulmonary edema and complications from volume overload in the group that got randomized to pre-hospital cooling. The question also refers to other modalities. There are a number of others that have been studied or are being studied. One of them is a transnasal cooling device that's been studied in Europe in a study that showed also no difference in primary outcome, but there were some interesting signals in the subgroup of patients with shockable rhythm. So stay tuned. I suspect we'll learn more about pre-hospital cooling and that these series of studies is not completely over. Go ahead, Kelly. Tony, I think the only thing I would add is that this issue highlights the challenge we've had with translating pre-clinical findings to the clinical challenges that we have with pre-hospital transport and patient phenotypes. It certainly seems there is evidence to continue that study, and I know that investigators will. Well, thank you, and Dr. Sawyer, there's also a question specifically for you. Is there a target HIA ratio that we think we should utilize specifically in a prospective trial? That's a great question. I don't think we have enough information to target one specific hypothermia to ischemic ratio yet. The couple analyses my colleagues and I have done in the past have looked at small and then a slightly larger database, but it's been done retrospectively without granular TTM data available. So I think this really, the next step is to study prospectively and to identify a more reliable measure of injury, whether that's a clinical variable, a laboratory test. We need some more work there. Thank you very much. And this question also comes from our audience, but I'm very interested since I work in a surgical ICU. What are your thoughts on both post-traumatic arrest and cooling in those patients? So for indications other than cardiac. Yes, I think, first of all, it's really parsing out based on the patient's history whether the trauma caused the cardiac arrest or the cardiac arrest was the precipitating event for the trauma. We see this a lot in the emergency department where a patient comes in as a car crash, but they were a single vehicle crash for no other explained reason. And it was really the ventricular fibrillation that led them to crash their car. In that case, I think the situation is to ensure that there's no severe or life-threatening hemorrhage, which could be a contraindication to targeted temperature management and to proceed as you usually would. I think for a true traumatic cardiac arrest due to exsanguination or to severe traumatic brain injury, the data are much more tenuous and we are not routinely using TTM in that patient population. Very good. Kelly, did you have anything to add? I don't have much more to add beyond Dr. Johnson. I think certainly there are traumatic models of cardiac arrest and temperature management but clinically, our job is to sort of parse out what came first and what the individual patient would be most benefited by. And my understanding of the situation, you can correct me if I'm wrong, is even if it's truly due to exsanguination and the trauma population, they're actually looking at least the one study out of shock trauma in Maryland is looking at really cold temperatures, so even less than 36 degrees Celsius, so it might be a different population altogether. Now my next question goes, is there definitive evidence that longer time to ROSC, total ischemic time corresponds to poor neurologic outcomes? I'll take that one first. Certainly there are many studies that have looked at this and seem to come to that association. I think we have to recognize the potential for confounding based on withdrawal of life-sustaining therapy practices and other individualized patient factors that we may not always be able to account for in the studies. So it does seem that there's a relationship, certainly there are other case reports of people who do well, but it seems like the data is potentially muddy and we may not be able to always account for all the variables that are important for this particular question. Do you have anything to add, Nick? No, I think that was pretty well covered. And then this is a couple questions here. If pre-hospital is still being studied, do you both institute TTM as soon as possible once the patient arrives at your facility? And based on your studies, what are you guys doing differently at your own institutions? First part of the question is, do we institute TTM as soon as possible? We do. First steps, obviously, are always ABCs and to ensure that any precipitating cause of the cardiac arrest has been addressed and reversed, but not long after that for us is placing a definitive cooling device and starting TTM in the emergency department, which is how our protocol is written and intended to be used. Based on our study, well, our study really started initially as a quality improvement effort. And based on a look at our QI data back in 2017, we changed our goal temperature back from 36 degrees to 33 degrees. So it had a very real and very immediate impact on a clinical protocol. And now we're evaluating that in ongoing fashion. So we don't know if that led to any change in outcome or any other processes yet, but we're hopeful that it was the right move. I think for us at my institution, the goal is to always initiate as quickly as possible. That doesn't always happen in the emergency department. If we can get a bed quickly in the ICU, then that's where everything takes place. I think the keys going forward are to be able to track that very clearly and consistently so that we can always look back in a QI fashion and then study it more definitively in the future. Thank you very much. And one of the questions I had is, especially NICS patients, it looked like those that had a witnessed arrest had a higher odds of improved outcomes, and it could be a surrogate for a short and ischemic time, especially if you're only doing hands-only CPR. And besides improved CPR, are there any other strategies do you think that might help improve outcomes in these patients? Yeah, I think, you know, looking at the kind of global out-of-hospital cardiac arrest literature, some of the most effective, all of the most effective interventions are really the ones that are performed very early. And so those patients who had a witnessed arrest have a distinct advantage of having a higher likelihood of getting bystander CPR. And for us in this region and for many regions in the country, it's telephone-assisted bystander CPR where 911 dispatchers provide real-time instructions to lay people who are near the patient to immediately get hands on the chest and start good quality CPR. And then also unwitnessed arrests, you know, tend to have longer times to 911 call and longer EMS response intervals as well. So I think building systems that optimize those things as much as possible is really what the EMS folks here in Seattle and King County have done so well for over 40 years now. And those are probably the main contributors, and all the other stuff we're doing in the hospital probably contributes in a relatively minor way compared to that stuff. Did you have anything else to add, Kelly? No. I think that continuing to enhance the system of care and everyone involved, as cardiac rest is not an experience in isolation, will help improve outcomes for all in the future. So I think all these things we've talked about today are very important to continue those conversations. Well, thank you very much, both of you. I see we're all out of time, and that concludes our question and answer session. And many thanks, and thanks to our presenters and the audience for attending. And a very special thanks to our sponsor, the Penn TTM Academy, an educational program led by the expert faculty from the University of Pennsylvania, which focuses on post-arrest care and targeted temperature management. For more information, please visit pennttm.com. That's P-E-N-N-T-T-M.com. And again, everyone who joined us for today's webcast will receive a follow-up email that will include an evaluation. Your feedback is greatly appreciated. And on the final note, please join us for our next Journal Club webcast on April 23rd. Thank you very much, and taking time out in this crazy time. And have a great day. Bye.

webcast

targeted temperature management

TTM

cardiac arrest

retrospective cohort study

neurologic outcome

hospital discharge

survival

hypothermia to ischemia ratio

functional outcome