Welcome, everyone. My name is Mitchell Buckley. I'm a clinical pharmacy specialist in the medical ICU at Banner University Medical Center in Phoenix, Arizona. And it is my distinct pleasure to be talking with you today at the SECM's Critical Care Congress. And the topic that I'm going to be discussing is the never-ending controversial topic of looking at corticosteroids use in ARDS, which it seems like it particularly waxes and wanes over the past several decades, whether either for or against its use. But more importantly, I'm going to be focusing on a little bit more of the certain aspects, such as why would we use them, when would we use them, so in other words, the optimal timing, and which agent might be more beneficial, and what particular cost might we be expending to some of our patients as far as exposing them to potential adverse events. Just a quick note of disclosures. I do serve as a paid consultant at Walther's Kluwer. However, with that being said, none of the information discussed in this presentation is related to any past or current work. In my capacity, it is my role as a consultant. The learning objectives today that I'd like to cover, first and foremost, is just basically review some very common differences between the corticosteroid characteristics between some of the agents we commonly employ for ARDS. Secondly, evaluate the clinical data on corticosteroids' impact on the ARDS and its potential role for acute respiratory syndrome, so more specifically, what are some of the clinical outcomes that it potentially impacts, and then lastly, based on this information, can we recommend an evidence-based approach for corticosteroids and ARDS, specifically answering some of the questions within the title, such as, is there compelling data to use it commonly and widespread in a lot of our ARDS patients, whether it be COVID or non-COVID related, and at the optimal timing, as well as the optimal agent, if we can answer some of those questions. And unfortunately, even though we are in a virtual environment, and I'm not able to poll the audience, here's a question, if nothing else, to at least ask yourself. Which of the following corticosteroid agents do you prefer in ARDS? And for the time being, we'll think about this as more non-COVID related. Is it 1, hydrocortisone, 2, methylprednisolone, 3, dexamethasone, or 4, I'm just not sold on corticosteroids altogether. And then ask yourself, does my answer change if it's COVID related ARDS? So without wasting any time and jumping headfirst into this controversy as far as answering the question, which of the following agents is the agent of choice in ARDS, and again, we'll kind of concomitantly try to answer the question as far as, is there compelling data to use steroids at all, altogether. So quickly to cover some basic slash boring information as far as corticosteroids, I'm not going to go into this a lot in depth, because most of you probably have seen this table before in some form or fashion, but as far as the different agents that we have available, and again, not being all inclusive, but just kind of covering some of the common agents that we have seen studied in ARDS, mainly hydrocortisone, methylpred, as well as dexamethasone. Clearly, we can see here the relative glucocorticoid activity as we move away from hydrocortisone, further derivatives away from it, the glucocorticoid activity becomes more potent, i.e. needing less of that amount of drug compared to hydrocortisone, given that potency. As far as the mineralcorticoid activity, again, moving away from hydrocortisone, specifically methylpred, as well as dexamethasone, have little to no mineralcorticoid activity, although the relevance of this in ARDS remains somewhat questionable. One other point to make here is the duration of action. As far as hydrocortisone, compared to again, the more potency as we go down, or excuse me, go up in the potency, the duration of action lasts longer. So the relative clinical significance of this remains somewhat questionable, however, with that being said, it might be a matter of more convenience as opposed to methylpred, the dosing strategy, compared to what we would use, say, with dexamethasone, in other words, giving it once a day as opposed to multiple times within the day. There is one other point I want to make here, and that is the question as far as are there any subtle differences between some of these agents that might explain why one might be particularly better than the other, specifically in ARDS. So unfortunately, looking at dexamethasone versus methylpred versus hydrocortisone, that question hasn't been particularly answered, however, with that being said, there was an older study that looked at the lung disposition of methylprednisolone compared to prednisone that was published in 1983 in Lancet. So really what they found was there was a potential theoretical advantage of using methylprednisolone as attributed because there was a higher correlation looking at the serum concentrations within the concentrations of that drug specifically in the BAL fluid with methylprednisolone, and they found that there was more concentration of that drug with methylpred over prednisone. However, keep in mind, they didn't look at outcomes, they were just looking at how much of that drug was actually concentrated within the lung, and of course, you can make the insinuation that if there's more drug there, that could potentially, theoretically, although not proven, potentially play a better role as far as disease state mitigation or managing the syndrome of ARDS. However, with that being said, I do want to emphasize that that has not been studied directly between methylprednisolone and dexamethasone, so while that benefit was found over prednisone, that has not been so with dexamethasone. And cutting right to the chase and answering one of the questions as far as the clinical impact that corticosteroids have in ARDS, a very high-level overview, because clearly we don't have the time to cover all the bases we need to go over each individual trial and critique each one individually, so very high-level overview, and looking more specifically at the mortality impact that steroids play, and this actually trial, or excuse me, meta-analysis did a great job as far as highlighting steroid use in ARDS, both COVID and non-COVID related. Now I'm bringing your attention to, this is actually Contemporary Findings published in 2021, and one of the other things I also want to include as far as this particular meta-analysis is that the inclusion criteria included RCTs that had strict criteria for ARDS in their definition, so in other words, they only included RCTs that were consistent with the AECC or the Berlin definition for ARDS as we know it today. So I'm going to bring your attention to the bottom green graph here, and as you can see here, all the trials that were included, the pooled data did show a signal favoring the use of corticosteroids and the beneficial effect it had on mortality, i.e. lowering the risk of death. And they further broke it down between two different subgroups, and first looking at just those trials that looked at COVID-19 ARDS, and as you can see here, two trials were included. However, what they found was that there was no significant difference on the impact of mortality in ARDS with steroids compared to the control group, whether it be usual care or placebo. With that being said, the non-COVID ARDS trials that were included were a little bit more numerous, and it did include the more recent DEXA ARDS trial, and again, looking at the pooled data overall, there was a signal consistent with a beneficial use of corticosteroids. However, with this being said, we'll point out that there was a limited amount of information for the COVID-19, so even if there was a true signal of a mortality benefit that was not found here, possibly due to the low number of trials and sample size of patients in the pooled data. So you can see here that the overall signal for reduction in mortality with steroids was largely driven by the non-COVID-19 ARDS clinical trials. Also further emphasizing the impact it has on clinical outcomes, a different source recently also in Critical Care published in 2021 by a different group of investigators did look at some other outcomes as far as clinical outcomes, and the one important outcome also that they evaluated from pooled data was the mechanical ventilator free days, specifically at 28 days. And of all the clinical trials that were included here, as you see listed, again, including the DEXA ARDS trial that was recently published, pooled data did favor the use of glucocorticoids as far as increasing the number of mechanical ventilator free days, again, at 28 days compared to the control group. They also did look at kind of more, although not as robust clinical marker, but more of a surrogate marker as far as like the PF ratio impact that steroids may have played. Again, those that were reporting that data were included, and of that, it did show a significant actually improvement of steroids impacting the PF ratio as well, although the clinical significance, again, remains somewhat debatable. Here's yet another independent source that does suggest that there is a positive clinical benefit of using steroids in ARDS, not just with mortality, but actually helping improve mechanical ventilator free days. And so if we're making the argument that, yeah, there's some data to support the use of steroids, and if you're using it in your practice, and widespread use is probably potentially warranted, and if you're buying into that, then the next question you might have is if I'm going to use a steroid, which agent am I going to pull the trigger and actually employ within my practice? And so here it was actually a great work done by the REACT Working Group that was published in 2020 in JAMA. And so I do want to emphasize that this was actually a systematic review made analysis that looked at steroid RCTs in COVID-19. And I do want to emphasize another caveat that not all of these trials that were included in this analysis were ARDS. These were critically ill patients that were employed the use of steroids, but not specific to ARDS. However, with that being said, there were two trials that did look at ARDS specifically and not just an ICU population. And as you can see here, the overall impact that steroids had in COVID-19 in the ICU barely missed the mark as far as like demonstrating a risk reduction in death with corticosteroids. Again, overall from all the trials that were included in their pooled data. However, the great thing and the reason why I include this slide within the presentation is because they actually made an attempt to figure out like, okay, where's the benefit if there is one that's being driven potentially by a specific agent. So here what they did was they actually separated out between dexamethasone, hydrocortisone and methylpred. And looking at the dexamethasone, the subgroup pooled findings was that it actually did reduce the risk of death using the dexamethasone. Again, these were very large, robust, more contemporary reflections of ARDS management practice in the dexamethasone trials. Nonetheless, it did find a favorable response in dexamethasone in COVID-19 in a quickly ill population. Again, two of those reflective specifically in ARDS. Now hydrocortisone, three trials in total pooled data from three of those. They did not demonstrate a significant difference between the control groups and methylprednisolone. There was only one trial. So it's really kind of hard to interpret the true benefit just from the limited number of trials but also the number of patients. And again, they did not find in that specific trial any difference in mortality rate between steroids or excuse me, methylprednisolone in the control group. So if you're wondering if the steroids really impact COVID-19 mortality rates, it may appear that it might and largely was driven by those trials using dexamethasone. So again, is that really making the argument that we should be using dexamethasone possibly? Is it truly something unique and specific to that particular agent? I think that question remains unanswered at this point. Nonetheless, that is kind of moving the needle to favor the use of dexamethasone as being more reflective of the contemporary trials that use that particular agent in COVID-19 critical illness. And although one can make the argument that yeah, we could probably consider dexamethasone being the flavor of choice at this particular time based upon what we know, it definitely keeps the argument open that if you use an equal dosing strategy of another agent, could you get the same result? And again, that remains unanswered. Now moving on to the timing. More specifically, when should steroids be initiated in ARDS, if at all? And so that's the next burning question we're going to try to answer. So before moving on to the data, a quick audience poll question to ask yourself, what time period do you typically start steroids? And again, this is making the assumption that you do commonly employ them after the onset of ARDS. And again, we'll break it down between non-COVID. And you can also ask the same question, is there anything changes with those COVID patients? And so is it number one, do you typically start them if you do early? So in other words, within 72 hours. Is a little bit later, do you try other modalities for employing those first, and then if the patient doesn't turn the corner, say like a little bit later, i.e. after a week, then you'll consider steroid use. Or is it three? Is it really late, like after a couple of weeks? Or is it maybe four, somewhere arbitrarily in the middle of some of these options? And now let's take a look at kind of more of a high level overview of the use of early corticosteroids in ARDS. And again, these were findings published in 2020 from pooled data. And looking at the overall utilization of corticosteroids, the RCTs that looked at this in an early onset capacity, in other words, within 72 hours, looking at the endpoint of its effect on 28 day or 30 day mortality, and those trials that reported this information, as you can see individually here listed. And the pooled data showed that altogether that they did favor the use of corticosteroids. And what about 60 day mortality rates? And three of the originally four trials did report these as well. And again, there was a signal suggesting that that risk of reduction of death at 60 days was lowered with corticosteroids. So again, I apologize for the high level overview of this data. For the sake of time, we really can't get into the weeds. But again, the point being is that there, yes, there is some information out there suggesting that corticosteroids may be beneficial from a mortality benefit using early use within like 72 hours in ARDS. And while the previous slide kind of made the argument that, yeah, using it early on, so in other words, within about three days after the onset of ARDS, there might be a beneficial effect in regards to mortality. Well, what happens if we use it like later on? And so most of us are probably already familiar with the clinical trial, the Lazarus trial published by Steinberg in 2006, one of the more controversial trials out there. And so again, I'll just briefly kind of review their particular findings and how it relates to the timing of steroids. And so their primary endpoint looking at mortality at 60 days, patients were randomized to either methylprednisolone or placebo. And one of the things that I really want to emphasize to the audience is that patients were not eligible for randomization within this particular analysis. They had to be within seven to 28 days after the onset of ARDS. So again, this was more like later on persistent ARDS. So again, excluding the early onset use of corticosteroids. And so seeing if that can play a role as far as answering the question in regards to the optimal timing of using steroids, and does it actually make a difference? And so kind of cutting through the chase here, the overall cohort, there's no differences between the methylpred and blue and the placebo and gold, between the mortality rates between the two, and just shy of about 30% in each group respectively. And further breaking it down between the exact timing, 132. So the majority of the patients were within about one to just shy of two weeks after the onset of ARDS. And again, no significant differences between the mortality rates. And here's the finding that actually kind of raised eyebrows at the time. Those patients that were randomized after 14 days of the onset of ARDS, there was a significant difference of a higher rate of mortality with methylpred compared to the placebo group, which had a mortality rate of 8%. Now, this was unadjusted crude mortality rates. And I do want to emphasize a couple points here, is that this was an association suggesting, and some of you may have interpreted this, is there a harmful effect that steroids may play as far as like the higher rate of mortality compared to placebo? And I do want to point out also, there is the controversy is that if you look here, the mortality rates across the board were just either hovering below or above about 30%. Visually, you can see here, the unusually low rate of 8% in the placebo group should potentially hit the pause button and ask us to revisit this question, which we will do so a little bit more in detail in the next slide. But before we move on, a post-topic analysis of 180 days, again of the same analysis, looked at similar findings across the board for methylpred versus placebo, and again corroborated the earlier finding at 60 days that at 180 days, the steroid group actually still sustained a higher rate of mortality, and again, the placebo group again was unusually low. So on the surface, that could beg the question, well, if we know early use of steroids may be beneficial, and here it's suggesting that there was really no mortality difference if we started between 7 and 13 days, so a little bit later, as long as it's under two weeks, we didn't see any signal for harm, and contrast that with, well, after two weeks, maybe that's the point of no return, and if there is caused a harmful effect, maybe the sweet spot is do not start any later than two weeks if you're going to use steroids at all. So I think this is an important question that we're going to go into a little bit more detail on the next slide. Now on the previous slide, clearly there's a cause for concern that if you're buying into that, that steroids could be potentially harmful if you're using them after about two weeks. So I think it's important that a deeper dive is reviewed for those specific details, because again, the whole point of this presentation is to identify the optimal timing. So while the 7 to 14 days really didn't provide a compelling reason to that was kind of a sweet spot when to initiate steroids, I think a bigger question is, do we absolutely avoid steroid use after two weeks at all costs? So really I want to look at some of the reappraisal that was published on the Lazarus trial that I think deserves attention, and so one of the biggest arguments was there's an imbalance at baseline characteristics of the trials within that particular cohort after the 14 day enrollment. So really what they were doing is reanalyzing the data to try to mitigate some of that imbalances. And so bringing your attention to here, one of the adjustments that they did on the mortality rate was looking at some of those baseline covariates such as age, acuity, number of organs that failed, plateau pressure, and alveolar to arterial oxygen difference. And so they redid that analysis, and what they found was that after adjustment of those baseline covariates, that significant difference was lost from the original publication. And so again, providing an argument that maybe that difference in mortality was attributed to erroneous baseline characteristics that were not accounted for in that analysis. They also took it one step further and looked at an adjustment on mortality with the incorporation of the Murray Lung Injury Score that was included within this model. And by doing so, they also found that there was, again, no differences between the mortality rate specifically in those patients that were enrolled after about two weeks. And again, continuing to use enhanced interrogation techniques on this data to try to confess the truth as far as like, is there really a risk of using steroids in late ARDS, specifically after two weeks, they did yet a third analysis on the impact of mortality. In this one, they actually looked at pneumonia, gender, and creatinine and added it to their model as well. And then again, they found no specific difference. Now, one thing to point out was there was a test of interaction between the treatment group, so in other words, methylpred and control, as well as the onset of duration in ARDS being the 7 to 13 duration and the greater than 14 day duration of ARDS. And that was in the original analysis using an a priori unadjusted plan test for interaction, and that was found to be significant in the original trial. Now, actually using an adjustment, again, for that Murray Lung Injury Score, again, significance was lost through the use of log regression repeated analysis with the five baseline covariates that were mentioned above. So again, the point of this slide is maybe not taking that interpretation that was originally published at a surface value, and again, we could potentially contribute that significant finding to just simple imbalances that were not adjusted for, and by doing so, that difference in mortality was potentially lost. And I'd be remiss if I didn't look at or present with you another, again, high-level overview of pooled data that was published in 2009, so again, after the publication of the Lazarus trial. So of the trials that were included at that particular time, which unfortunately was about like 10 years ago, they specifically looked at the question of early versus late and trying to answer the question on that debate for the impact of mortality. And again, as you can see here, looking at the early, less than seven days, and after the later days, again, there was no differences in the risk of death between those. Now the caveat being is even though it's not suggesting that starting it after seven days, it really didn't answer the question or corroborate after two weeks specifically, and again, there was no other data other than the Lazarus trial, so they didn't try to pinpoint that as far as is there, again, harm or was it equivocal as far as like after two weeks of starting steroids, would that impact mortality either neutrally, beneficially, or deleteriously? And yet, Maduri tried again with another publication in 2016, so again, a little bit more contemporary of answering the question of early versus late, so really kind of beating a dead horse here. The only weakness was that they were trying to answer this question, but also they did take it down to the patient level data of all the RCTs, and pooling that information, they did not find a signal of early versus late, and so again, the only trial to date that actually looked at the use of steroids after 14 days was the Lazarus trial results, so again, they were not answering specifically, focusing on that particular question, but nonetheless, it didn't really show a signal that whether it be early or late, that that really had made a difference on the mortality rate as far as figuring out for us as the clinician to apply at the bedside, is what is that optimal timing of using steroids? So the point to drive home is up to this point that we have not seen any evidence other than the Lazarus trial, which hopefully I've made the point that that was potentially erroneous, and the exact timing has not been found, but furthermore, it's been specifying that you've got to use it within 72 hours, otherwise if you don't, then potentially that beneficial effect if there is one, was lost. So ultimately, the takeaway is that the exact timing, whether it be early or late, does not appear to make a significant difference on the data that we have currently. And if you're still not convinced or still skeptical if early versus late really makes a difference, this actually looked at a different take rather than just mortality, but what about the effect of early versus late on the impact of essentially giving them off some sort of ventilator assistance, so unassisted breathing rates, was that they looked at from pooled data of this trial from Maduri, and what they looked at was looking at corticosteroids here marked by the blue line, and placebo in the green, and looking at the early use, so in other words, within 72 hours, there was actually a signal suggesting that getting them off the unassisted breathing was a little bit more impactful if you use steroids early on. But what about late? And so here they defined it as beyond day five or later, and again, same similar findings that the beneficial effect had, so again, here we have analysis suggesting that mortality was probably not negatively impacted, but there's actually a beneficial impact as far as like getting them off ventilator support if steroids were actually started a little bit later. And again, unfortunately, I'm going to emphasize that they were not able to analyze beyond the 14 day mark if that was capable, because again, the only study that we have at this point was the Lazarus trial that has done so. So assuming if I haven't lost you as far as like answering the question as far as the optimal timing, I'm going to try to really drive it home. There really is not any data to suggest that steroid timing, whether it be within 72 hours or even after day five or day seven, appears to make a beneficial or deleterious impact on the outcome per se of mortality and or unassisted breathing. Now again, the one thing to go back onto if you're fixated on it is the Lazarus trial. That is the one signal that tend to do so. But again, pointing out the limitations of that interpretation of that higher risk of death with the steroid group compared to the placebo after 14 days, again, should bring to light as far as some of the questionable interpretations of taking that at face value. So if we're going to try to answer the question, leaving the audience of, well, what should I do? I believe in steroids. I'm going to use them. Does it really matter? Can I use it at day two, day five? And here's one thing I really want to kind of emphasize too, is again, from a theoretical standpoint, if we're looking at the progression and the timeline of ARDS between the exudative phase and the fibroproliferative phase, here's also one other thing to reflect on. And looking at the maximal effect of each of these entities and the syndromes associated with them, the pathogenesis of say like edema typically peaks within the first probably two to three days at most. And again, the progression may begin as far as that interstitial inflammation as well as those fibrotic phases, but it typically doesn't peak on average, or at least we believe so in most patients until probably at least after a week, if not a little bit later. So here kind of begs the question, do you really want to wait until the point of no return in which the patient actually has gotten to this point? So in other words, starting it after say like day 10 or even like waiting for day 14, which most of us I would imagine probably don't. So maybe this will emphasize the fact that if you do believe in the data with steroids having a beneficial effect on outcomes, maybe why wait until we get here? And so maybe we should incorporate it a little bit earlier, again, based on this theoretical data. But again, emphasizing the fact that we don't have robust data to guide that optimal timing of corticosteroids. But again, leaving you more with this food for thought that again, if we're going to try to use it to prevent the fibroproliferative phase from peaking, maybe we should consider it a little bit earlier in the course of therapy. And lastly, one of the other questions is, what cost are we doing it? So I'm probably not going to get into the financial costs, but more specifically, what are the maybe costs or in other words, exposure to risk if we use steroids in our aortase patients? And lastly, there is a benefit, is there actually a negative side of using steroids commonly in ARDS? And clearly again, we don't have the time to go over every single trial and what their adverse events were reported for each one, but kind of just fast forwarding to the more recent dex aortase trial and looking at the adverse event rates. So interestingly, even for hyperglycemia, they didn't find a significant difference even though they're high in both. And more importantly, new infection and barotrauma, there's no differences between dexamethasone and the control group. So again, a good sign that we weren't exposing potential unnecessary risk to the patients if we're employing the use of corticosteroids or more specifically dexamethasone in this trial. So cutting through the chase and overall globally, what are the risks associated with steroids? Hands down, hyperglycemia was the most commonly reported increased risk of event compared to the control group, whether it be usual care or the placebo groups in those clinical trials. However, with that being said, hyperglycemia is relatively common in the ICU and again, not to belittle its clinical significance, but we do have modalities in which we can at least mitigate and control some of the glucose levels in a lot of our ICU patients. With that being said, I really do want to emphasize probably maybe the more salient points is that there's no evidence that steroids and ARDS increases other serious adverse events such as GI bleeding, new infection, or neuromuscular weakness. So the bottom line, hyperglycemia is probably going to be the one side effect that we'd probably significantly see when we do incorporate steroids to no one's surprise. However, with that being said, there really hasn't been a strong signal to suggest that the more serious adverse events are increased when we do use corticosteroids. So again, not a very compelling argument as a reason not to use steroids. And the major takeaway is to kind of wrap things up. The optimal agent dose timing of steroids remains still even today controversial and that goes for both COVID related and non-COVID related ARDS. However, with that being said, not to leave the audience hanging. So the question is, well, what should we use and when should we use it? And so really one of the kind of major takeaways and you may or may not disagree is that consideration of early dexamethasone may be preferred. And that's truly in the absence of knowing definitively what that optimal timing and if there is truly a drug class effect versus one particular agent is more beneficial on the outcomes. However, with that being said, there has been a lot of variability as far as the dose, the agent, the timing, but the point being is that the more contemporary clinical trials, the more robust data that we have today, and the more reflection of contemporary management practices of ARDS altogether all suggest that if we're going to follow the evidence, those breadcrumbs tend to lead us to the conclusion that if you're going to use it, maybe consider earlier use and with the more robust data using dexamethasone, that might be the agent of choice based on a reflection of the data and not necessarily the fact that there is truly a benefit with that particular agent that we know of, at least at this point. And lastly, the impact of survival and ventilator-free days with being a potential signal for that, that clearly outweighs any potential adverse events that we could be potentially exposing our patients to. And with that being said, there's still a lot of unknowns, but I will leave you with this. The only known fact is that I guarantee a pharmacist in the ICU should and probably will not badger you or your ICU team about the costs associated with steroids. And with that being said, I want to thank you for your time and attention. And if you should have any questions, feel free to email me on the title slide, and I would be happy to correspond with you at my earliest convenience. Thank you so much.

corticosteroid use

acute respiratory distress syndrome

optimal timing

agent selection

potential risks

glucocorticoid activity

mineralocorticoid activity