Transcript
Speaker: Welcome to Critical Matters, a sound critical care podcast covering a broad range of topics related to the practice of intensive care medicine.
Speaker: Sound Critical Care provides comprehensive critical care programs to hospitals across the country.
Speaker: To learn more about our programs and career opportunities, visit www.soundphysicians.com.
Speaker: And now your host, Dr. Sergio Zanotti.
Speaker: Corticosteroids have been utilizing critical care for their anti-inflammatory and antifibrotic properties for decades.
Speaker: Both in sepsis and ARDS, debate over the role and efficacy of corticosteroids have lasted years.
Speaker: With the development of the COVID-19 pandemic, the potential role of corticosteroids once again became a topic of great interest and debate amongst intensivists.
Speaker: In today's episode of the podcast, we will take a deep dive into the topic of corticosteroids and COVID-19 ARDS.
Speaker: Our guest is Dr. Todd Rice.
Speaker: Dr. Rice is an associate professor of medicine in the Division of Allergy, Pulmonary and Critical Care Medicine at Vanderbilt University.
Speaker: He serves as the director of the medical ICU.
Speaker: As a physician scientist, he conducts clinical research in the ICU, specifically in patients with sepsis, ARDS, and acute respiratory failure.
Speaker: Dr. Rice's research has expanded in the last few years to using the ICU as a learning healthcare environment and conducting comparative effectiveness trials.
Speaker: In addition to having his own National Institute of Health NIH funding, Dr. Rice serves as the Vanderbilt PI for the prevention and early treatment of acute lung injury, the PETL network.
Speaker: Dr. Rice recently coauthored an editorial published in JAMA entitled Corticosteroids and COVID-19 ARDS, Evidence and Hope During the Pandemic.
Speaker: It is a true honor to have him on the podcast.
Speaker: Todd, welcome to Critical Matters.
Speaker: Thanks, Sergio.
Speaker: I would like to start with a
Speaker: brief historical context on the steroids and critical care, I mean, both in ARDS and in sepsis.
Speaker: Seems that we've been talking about them for decades now.
Speaker: I mean, I think it's been a little bit of a convoluted picture.
Speaker: And, you know, they, decades ago, were being used in these patients to decase the inflammatory cascade and try and, you know, overcome the body's inflammation to whatever the
Speaker: the etiology was of the ARDS or in septic shock, you know, the infection.
Speaker: And I think largely, as people know, there were lots of mixed results from those studies and a pretty convoluted picture that made it really hard to know what to do with steroids in patients that had either ARDS or septic shock.
Speaker: You know, then we had some
Speaker: reasonable randomized trials of high dose steroids that didn't seem to show a benefit.
Speaker: Some trials of late steroids in patients with ARDS where there may have been even potential neuromuscular weakness as a harm.
Speaker: And I think that for at least a significant proportion of the critical care world sort of turned people off to steroids and said, well, maybe they don't work.
Speaker: They do have some side effects.
Speaker: Maybe we shouldn't be using them.
Speaker: But as many know, there's always been an interest.
Speaker: And from a biological plausibility and mechanistic standpoint, they've always seemed to make sense that they're anti-inflammatory and pretty nondescriptive anti-inflammatory, meaning they broadly cover a lot of inflammation and mechanisms of inflammation.
Speaker: And so the...
Speaker: the concept of this is an inflammatory condition and it seems like the body's inflammation is what's doing a lot of the damage in this situation, both in ARDS and septic shock.
Speaker: It seemed like biologically it would make sense that if we could turn that inflammation off or depress that inflammation, the patient should do better.
Speaker: So there's always been this interest in them because biologically they make a lot of sense.
Speaker: And then, as you know, in the last three or four years, there have been a number of studies that have been very provocative in saying maybe with the right approach and the right dose of steroids, maybe there is some benefit to them.
Speaker: And there's trials in septic shock, like approaches in adrenal, that say maybe they do get patients off of pressers faster and get people off the ventilator faster, and maybe
Speaker: I don't think it's definitive in septic shock, but maybe they even have a mortality benefit.
Speaker: And then in ARDS, DEX-ARDS came out a few months ago and had a positive signal for mortality.
Speaker: There are many people, me included, that think that that signal is so big that it's probably not really that big of a mortality signal, but have to admit that there's a mortality signal and that maybe in patients with ARDS, texamethasone may,
Speaker: may improve mortality, may improve outcomes.
Speaker: I think part of the hard part, at least with ARDS, is that it's a very heterogeneous disease.
Speaker: And there's many, many, many different etiologies of it.
Speaker: And traditionally in the past, we've put all of those etiologies into the same bucket, called them ARDS.
Speaker: It's definitely a syndrome.
Speaker: It's a syndrome that can happen from many different diagnoses.
Speaker: And then we've tried to treat them all the same or try and study them all with the same potential intervention.
Speaker: And my suspicion is that that's what's contributed to the cloudiness and the muddling of this picture and made it harder for us to truly understand what the signal is of corticosteroids in patients with ARDS.
Speaker: It feels, Todd, that almost at the beginning of the year, like you said, DEXA-ARDS came out, but I think it was totally overrun by COVID and our focus on COVID, and we didn't really
Speaker: discuss it as much in many circles.
Speaker: But the feeling at the time was almost, steroids make sense.
Speaker: They probably work.
Speaker: We just haven't been able to figure out which are the exact patients that would benefit, right?
Speaker: Yeah, I think that's a big part of it.
Speaker: I think, you know, the other thing that trials like DEXA ARDS run up against is that there's some history with steroids.
Speaker: And so, you know, when a study like DEXA ARDS comes out and shows a
Speaker: what I consider to be a really big treatment effect, people say, well, if it's really that big, we should have, even if we had heterogeneity in the population and issues with trial design in the past, we should have been able to see something.
Speaker: There should have been something there.
Speaker: And if there wasn't anything there, then I don't know that there's any chance that the effect could be that big.
Speaker: And I think there may be some truth to that, but I also think that it's not
Speaker: it doesn't allow us to just dismiss the effect entirely and say, well, it can't be true because we haven't seen it in the past.
Speaker: And I think, you know, DEX-ARDS is swimming upstream a little bit from the past history of steroids and ARDS and, you know, no single study that showed a huge, huge effect.
Speaker: But I think that that's how progress is made and that's how we better understand treatments and diseases and,
Speaker: you know, as we're getting further and further into this and doing more and more steroids trials in ARDS, we're getting a better understanding.
Speaker: And DEXA ARDS is obviously the most recent part in non-COVID ARDS to give us information to kind of help clarify that better understanding.
Speaker: Before we started recording, we were just talking about the COVID pandemic.
Speaker: And one of the comments you made is that obviously as devastating as the pandemic has been for patients and as hard as it's been for healthcare providers,
Speaker: There's also always a several lining and one of them was the amount of effort that scientists around the world have put around trying to find answers.
Speaker: And specifically, we were just commenting on how much has been studied in steroids and ARDS due to COVID-19.
Speaker: So why don't we jump into COVID-19, ARDS and corticosteroids.
Speaker: And perhaps we could start Todd with a timeline that takes us back to February, March,
Speaker: when the initial recommendations from WHO and others were to not use corticosteroids in a widespread use, what the thought process was at that point?
Speaker: Yeah, the World Health Organization, the CDC all had comments and recommendations not to use them, not to widespread use them for the treatment of COVID-19.
Speaker: Obviously in those statements, they always say that if a patient has a condition that would otherwise be treated with corticosteroids, so they have an asthma exacerbation or they have, you know,
Speaker: flare with lupus or another connective tissue disease that you would treat with steroids, then you should use steroids, but not to use them specifically for the treatment of COVID-19.
Speaker: And, you know, I'm not part of those organizations, so I can't necessarily tell you with confidence why they did that.
Speaker: But I do know that the data for other viral illnesses, especially flu, MERS, and SARS, suggested that steroids increased viral replication.
Speaker: you know, made it so that the virus replicated more readily and for a longer period of time.
Speaker: And I think that kind of preliminary data, not necessarily clinical with clinical outcomes, but preclinical in viral loads and viral replication, I think that data scared people about maybe steroids weren't going to be helpful because we've studied them in ARDS before and haven't found a huge signal, save DEXA ARDS, which was very recent.
Speaker: And there's a potential downside of if it makes the virus replicate, maybe it's going to make the disease even worse.
Speaker: And therefore, no strong recommendation.
Speaker: In fact, the recommendation to avoid steroids for the treatment of COVID-19 specifically from the World Health Organization and the CDC.
Speaker: Interestingly, the Surviving Sepsis Campaign actually had a weak recommendation.
Speaker: They said, we think you should use steroids in COVID-positive ARDS.
Speaker: And I think they were really kind of the ones that
Speaker: that went out on that limb and said, you know, with the DEXA ARDS data, we think that there may be a potential here for some benefit, and we, you know, recommend that maybe you should use them in that situation.
Speaker: And then the IDSA was sort of what I call the political response, which is use them in the context of clinical trials
Speaker: We need to study them, but we don't think they should be used in routine clinical practice at this point.
Speaker: And that's sort of the widespread recommendations and the diversity of recommendations across different groups.
Speaker: But I think you're right.
Speaker: In general, the most common recommendation was we don't recommend using these in patients that have COVID-19 for the treatment of COVID-19.
Speaker: And I think an important distinction for intensivists is that in our world, obviously a great number of patients that we're seeing in our ICU with COVID, especially in places where they have a high incidence, are going to be patients with ARDS.
Speaker: But the hundreds of patients that we see in our ICU with ARDS are just the tip of an iceberg of millions of patients with much milder COVID-19 symptoms.
Speaker: And that is probably also what people were alluding to in terms of making these recommendations.
Speaker: Yeah.
Speaker: Yeah, I completely agree.
Speaker: So let's talk a little bit about the trials.
Speaker: I think that there's a lot that came out in the last couple of months, which I think, like we mentioned earlier, is a silver lining and a great effort from the scientific community.
Speaker: But perhaps we could start with the early retrospective data out of China.
Speaker: And then if you could share with us the idea of the WHO Prospero meta-analysis
Speaker: in terms of how it was set up, and then we could probably go into the individual trials and put it all together later with the results of PROSPERO.
Speaker: Yeah, so the early retrospective data were data that, you know, as the word says, were retrospective.
Speaker: They went back and looked at patients that got steroids and compared them to patients who didn't get steroids and saw, you know, hey, it looks like the patients that were getting steroids may have done better.
Speaker: Everybody on this call probably, or everybody listening to this podcast probably understands that there are lots and lots of biases and confounders in retrospective data.
Speaker: And ensuring that the population of patients that got steroids is the same as the population you're comparing it to that didn't get steroids is really, really, really difficult.
Speaker: And we do a lot of statistical things like
Speaker: like propensity score analyses or adjustments for baseline variables and a lot of those to try and make those two populations that we're comparing as similar as possible.
Speaker: But there's always unknown confounders and there's always issues with things that aren't measured that you couldn't possibly have adjusted for or you couldn't even recognize that they were a potential confounder to even know to measure them to adjust for them.
Speaker: So there's always some issues with retrospective data.
Speaker: But retrospective data are sort of the first body of evidence and kind of give us an idea of, well, maybe this is something we should look at further and maybe this is something that we should examine in more detail in a prospective manner and maybe even in prospective randomized trials.
Speaker: The World Health Organization, you know, their job is to try and better the health of the global population and they,
Speaker: I think recognized along with a number of others, but they recognized that to get reliable and good information, we were likely going to have to combine efforts.
Speaker: And during this pandemic, for a long time, there were lots and lots and lots of patients and places to study, but you got to get the trials up and running to do those studies if you're going to do prospective randomized trials.
Speaker: And there's always some
Speaker: lead time in that and some actual effort and infrastructure and time that has to be built and put into it in order to get the trial up and running.
Speaker: And the patients with COVID aren't waiting.
Speaker: They're continuing to just come into the institution.
Speaker: And so we heard pretty early on from colleagues in China that by the time they got their trials up and running and actually were starting to enroll, their cases were going away and they weren't completing their trials because they just didn't have enough patients anymore.
Speaker: which is a good thing from the population standpoint of COVID, but a bad thing because you need to have patients in order to run trials to get an answer as to treatments.
Speaker: And I think that signal coming from China and Italy and some of the early places really made it apparent that the best way we're going to get rigorous answers is to try and combine efforts.
Speaker: And that sounds like a no-duh, of course.
Speaker: I mean, that's how else would you do this?
Speaker: It sounds like the most
Speaker: obvious thing probably that'll be said in this podcast.
Speaker: But getting people to pull that off is hard.
Speaker: And getting people to networks, trialists, researchers to share their data, especially to share their data before they're already published, takes a lot of trust and takes a huge effort.
Speaker: Because, you know, we in the research world are always scared that we're going to get scooped on something or somebody is going to put our data out there before we publish it and then
Speaker: it'll make it non-publishable and we won't ever be able to publish it under our name.
Speaker: And so there was a lot of trust in the World Health Organization.
Speaker: The World Health Organization made a lot of promises and commitments to these networks to be very confidential with the data and to safeguard the data in a way that wouldn't inhibit or hurt
Speaker: potential future publication of individual trials and even worked with JAMA in making that so that they were all published together and so that there wasn't any scooping and there wasn't any pre-publication release of the data so that it would hinder the chance to publish it.
Speaker: So the concept of combining these trials, I mean, we've been doing meta-analyses, et cetera, forever, right?
Speaker: The concept is not new.
Speaker: The idea that we had to do this in order to get enough data to answer the question was readily apparent and apparent to a lot of people, but the machinery and the making it happen is where the World Health Organization really, really, really excelled and did a phenomenal job of getting large networks that were doing these trials to come together and say, yes, we will share our data with you so that we can have the most rigorous
Speaker: answer that we possibly can from all of the data that are potentially available to answer this question.
Speaker: And I think that also quite unique was, not for the first time, but not very common, the idea of doing this as the prospective meta-analysis, right?
Speaker: So getting everybody together on board from the get-go and moving forward, considering that time was a premium with the pandemic going on.
Speaker: Yeah.
Speaker: Yeah, I agree.
Speaker: That wasn't totally novel.
Speaker: It had been done a few times in the past.
Speaker: It certainly was in the early stages of things that have been done.
Speaker: And it increases the rigorousness of the meta-analysis because you know, prospectively, that you're getting these data.
Speaker: You can collect the data in a way that makes them a little bit more easily comparable and makes the analysis a little bit easier to understand and to do.
Speaker: Perhaps we could start with going through some of these trials in a brief description.
Speaker: I think that we'll put links to the trials for people to read in detail.
Speaker: And I think that ultimately it's the sum of this data and the meta-analysis that maybe guides us and gives us the final recommendations.
Speaker: But it's almost, I think chronologically, the first one that we heard about and that really caused the most impact was recovery.
Speaker: Could you just tell us a little bit about recovery, please?
Speaker: Yeah, recovery was an open-label randomized trial done in the UK.
Speaker: The UK really got all of their kind of hospitals, their public hospitals together and said, we're going to do research on COVID.
Speaker: We're going to do it in a manner that even if you're not doing, not a hospital that's used to doing a ton of research, you're going to be able to do it.
Speaker: And we're going to do it in mass force.
Speaker: And in fact, I think during that time, one-sixth of the patients that were in the hospital with COVID
Speaker: were enrolled in a trial of some sort through the recovery is kind of the name of the platform.
Speaker: So there's recovery steroids, there's recovery plasma, there's recovery hydroxychloroquine.
Speaker: And recovery steroids enrolled a lot, a lot, a lot of patients in the range of 6,000 patients.
Speaker: It was open label, so patients and clinicians knew if they were randomized to steroids or if they were randomized to just standard of care.
Speaker: There was no placebo, it was just standard of care.
Speaker: And there are some issues with open label trials and there's some biases that can be introduced with open label trials, but it's also probably the quickest way to get a trial up and running and the way they did it in recovery, which was similar in REMAP-CAP, you'll hear about that also, allowed it to be done as just part of routine care.
Speaker: And there didn't have to be a lot of special research specific interventions or people or
Speaker: or blood draws in labs, et cetera, et cetera.
Speaker: You could just do it as sort of part of your routine care.
Speaker: It works really, really, really well for medications and therapies that are already approved and already in practice and that don't need a ton of safety studying and don't need a ton of regulatory monitoring to be done.
Speaker: And that obviously, corticosteroids fit that perfectly.
Speaker: So, and then people I think know that the results showed, you know, they were released first in a press release, which was very interesting for those of us on the front lines to say, what do you do with a press release?
Speaker: You know, it's one thing we were just starting to get used to what to do with preprints and understand sort of how we deal with those.
Speaker: And then recovery was released in a press release.
Speaker: There was a little bit of data in it, but not really the data that lets you analyze the trial results and understand them in depth at all in the press release.
Speaker: So it was released in a press release and the data showed, and I think people recognize this, that there was about a 30% reduction in the hospital mortality for patients in the ICU, or I'm sorry, patients on mechanical ventilation, and about a 20% reduction in mortality for patients that were on oxygen, but not mechanically ventilated.
Speaker: You know, it's just a huge effect, a huge, huge, huge effect.
Speaker: And I think maybe importantly also, there seemed to be no effect in patients that were not on oxygen when they were,
Speaker: started on their steroids.
Speaker: And so those were kind of the results that we had.
Speaker: Big, big, big trial, an effect that was really quite large and potentially differential effect depending on how sick the patient was.
Speaker: And this changed the whole landscape and also had a big impact on the trials that you wrote the editorial on that we'll discuss next.
Speaker: Could you just tell what the impact of this was on the trials and why ultimately
Speaker: the meta-analysis proved to be super valuable in this case.
Speaker: Yeah, I think the good and the bad of this is that there's a trial, it was done, it was big, it had an effect, and they're putting results out there.
Speaker: The bad side of this is that it made it really hard for other trials studying steroids to know what should we do with this.
Speaker: Is it ethical?
Speaker: Is it...
Speaker: okay to continue to enroll patients into a steroid trial where there's a placebo arm and patients aren't getting steroids?
Speaker: Are these data, these recovery data, good enough that everybody should just be getting steroids, that we should stop our trial?
Speaker: Is there this term equipoise?
Speaker: Do clinicians still have enough doubt on whether steroids work or not that they would even be willing to put their patients in these trials?
Speaker: Or
Speaker: Do patients have enough doubt as to whether or not these work or not, that they would be willing to say, yes, I'm willing to participate in a trial instead of just saying, I think it works, just give me the steroids.
Speaker: Those are tough, having gone through those a couple of times, those are tough decisions to make in a trial when new information comes out that makes it so that you need to consider changing your trial.
Speaker: And in this place, in this situation, not just changing your trial, but pretty much stopping your trial.
Speaker: And so every single one of them, I think independently,
Speaker: looked at the recovery results and said they're good enough results that we probably can't continue we can't continue to give people a placebo or a non-steroid harm in order to study this further and so they ended up stopping their trials when the recovery results were were released talk could we just recap the uh a
Speaker: three trials that ultimately were published together and are published alongside the meta-analysis that did include data from recovery as well.
Speaker: But I think it's valuable for our audiences to know what these are.
Speaker: And you mentioned REMAP-CAP, and we can talk about that first.
Speaker: Yeah, REMAP-CAP is similar to recovery.
Speaker: It's an open-label trial, platform trial, so there are multiple actual interventions that you could get randomized to in REMAP-CAP, depending on which domain you fit into.
Speaker: So the remap cap steroids, corticosteroids, which is what was published in JAMA, randomized patients to their steroids or a standard of care in an open label fashion, again, without a blinded placebo to try and determine the effect.
Speaker: It was kind of in the US, kind of in the UK, kind of in Australia.
Speaker: So it's a multinational in Europe also.
Speaker: It's a multinational trial led by Derek Angus in Pittsburgh.
Speaker: It's kind of interesting to note that it was not a platform that was designed for COVID.
Speaker: It's actually an ongoing platform to study community-acquired pneumonia and treatments in community-acquired pneumonia.
Speaker: And then they essentially added a COVID domain when COVID was here.
Speaker: And so they sort of used already existing infrastructure in order to do this.
Speaker: And REMAP-CAP uses a Bayesian approach.
Speaker: So it's a little harder to understand because it's not a traditional frequent statistical approach.
Speaker: And the Bayesian approach just says sort of, it doesn't give you a p-value.
Speaker: Instead, it gives you an output that says, what's the probability that one arm is better than the other?
Speaker: And if we continue to go, what do we think the chance that one arm will be shown to be better than the other arm will be?
Speaker: And if you read the article, you can see that's how the outputs are from remap cap.
Speaker: There's a,
Speaker: a 90% chance, posterior probability, that steroids are better in patients that were randomized into, there's three arms in REMAP-CAP.
Speaker: There was a 90% chance they're better if they were randomized into the steroid arm.
Speaker: It's about an 82% chance they were better if they were randomized into the what's called shock-dependent or steroid shock-dependent arm.
Speaker: And in the steroid arm, the first arm of the 90%, they gave patients steroids.
Speaker: They gave,
Speaker: I'm going to forget actually what the actual steroid was.
Speaker: Sergei, you remember what the steroid was?
Speaker: They all ran together for me.
Speaker: Hydrocortisone.
Speaker: Yeah.
Speaker: Yeah.
Speaker: They and Cape COVID gave hydrocortisone.
Speaker: Kodak gave the dexamethasone.
Speaker: Yeah.
Speaker: So RemapCap gave hydrocortisone and they gave a fixed dose to patients in the steroid arm.
Speaker: They gave a fixed dose to patients in the steroid shock arm, but only when those patients were in shock.
Speaker: So
Speaker: Patients that were in shock at the beginning got randomized steroids, would get steroids, but patients who were not in shock, who got randomized in that arm, would only get steroids if they developed shock.
Speaker: So a little less than half of the patients in that arm actually got steroids.
Speaker: So it's a really complicated arm to understand.
Speaker: But even with that, it looked like steroids had a pretty high probability, a really high probability actually, of being better than the standard of care group.
Speaker: And this study, as the other ones,
Speaker: showed this probability based on the base and approach like you explained of 90% in the patients who got the steroids regardless and 83% of being better than those who got it in the shock mode but did not reach its primary threshold to call it a positive trial, correct?
Speaker: Correct.
Speaker: Yeah, correct.
Speaker: A big part of that, you'll hear that in all three of these trials, a big part of that is that
Speaker: they were all stopped early.
Speaker: They were stopped when the recovery results came out.
Speaker: So they have fewer patients enrolled in them than they had hoped they were going to have or that they were planning to try and enroll.
Speaker: And because of that, they don't have quite the confidence in their answers that they were hoping to have gotten had they been completed with enrollment to the planned numbers.
Speaker: The second trial is the CODEX trial, which is the one that gave dexamethasone, and I think it was conducted principally in Brazil.
Speaker: Could you give us a synopsis of that trial, Todd?
Speaker: Yeah, so the CODEX trial was in Brazil, another open label trial, and used dexamethasone, so very similar to recovery.
Speaker: Actually used a little bit of higher dose, 20 milligrams instead of six milligrams for the first, I think, five days, and then 10 milligrams for the second five days, if I remember correctly.
Speaker: So a little bit of a higher dose of dexamethasone in Brazil.
Speaker: And, again, didn't meet statistical significance for its outcome, but showed in general benefit to, I think their primary endpoint was ventilator-free days, which is days alive and off of the ventilator, and showed improvement in those that did not reach statistical significance.
Speaker: Also looked at mortality and showed a mortality signal that was, again, not statistically significant, but very, very, very consistent with what the recovery results showed.
Speaker: I mean, all three of these trials, REMAP-CAP codex and the next one we'll talk about, CAPE-COVID, have signals that are remarkably similar to the recovery signal.
Speaker: And that's, I think, where the value in the meta-analysis comes into is that it shows you how consistent the signal seems to be across all these trials, even if some of them, REMAP-CAP codex, for example, don't have enough power to reach statistical significance because they were stopped early.
Speaker: And the last one you mentioned is CAPE COVID, which I think had the particular aspect that was the only one that utilized placebo.
Speaker: Yeah.
Speaker: Yeah, CAPE COVID was a French trial.
Speaker: It uses hydrocortisone, just like RemapCap did, but it's the only one of the group, even including recovery, that's blinded and used an actual placebo-controlled arm.
Speaker: So its endpoint was this funny endpoint of death or death.
Speaker: respiratory failure, persistent respiratory failure at 21 days, or the opposite of that, which is getting over survival and getting over respiratory failure at 21 days.
Speaker: And essentially that is, it's a little bit complicated, but essentially that is coming off of the ventilator at 21 days and being alive.
Speaker: So it's a little bit akin to ventilator-free days.
Speaker: And, you know, it again got stopped early, but it actually had a signal in some of its outputs
Speaker: that was statistically significant and showed benefit of hydrocortisone in patients that had COVID and severe respiratory failure.
Speaker: And it's very interesting that these trials, if were to be published independently or different time spans, would all probably still add to the confusion because we would say that they technically show a signal, but they're not positive per se.
Speaker: And I think that that's where probably the value of this prospective meta-analysis by the WHO comes into play and giving us a little bit more confidence of what to do with these patients.
Speaker: So could you tell us what ultimately the Prospero meta-analysis showed and who did it include?
Speaker: Yeah, so the Prospero, I think, included the four trials we talked about, Recovery, Remap, GAP, Codex, and CapeCOVID.
Speaker: It also included three other trials.
Speaker: that I don't think are published yet, but are smaller and were also stopped because of recovery of steroids.
Speaker: One of those trials actually looked at methylprednisolone.
Speaker: So there's a couple nice things about Prospero.
Speaker: One is that it allows us to sort of look at the effect of different steroids, both hydrocortisone and dexamethasone, which are used in multiple trials, and then even one trial with methylprednisolone.
Speaker: and look at those steroids and kind of see is the effect consistent?
Speaker: Are we seeing anything that's different among the different specific steroids?
Speaker: The other thing I think that it kind of did was it just gave the full picture.
Speaker: What is the real consistency among these trials?
Speaker: And if we put them all together and get huge, huge, huge numbers, what does the effect size look like?
Speaker: And so in Prospero, they found that mortality went from 40%, and this is in mechanically ventilated,
Speaker: patients, so the patients with severe COVID-19, mortality went from 40% to 32% compared to from the standard of care arm to the steroid arm.
Speaker: So an 8% reduction, absolute reduction in mortality, which is a 20% relative reduction for the 40% in the standard of care arm.
Speaker: 8% is a pretty big reduction.
Speaker: That means that if you treat 12 and a half, so call it 13 patients with steroids, you'll save one life.
Speaker: One patient will
Speaker: survive that would have otherwise died if you treat 13 patients with steroids.
Speaker: So that's a pretty good effect.
Speaker: We don't see that effect very much in critical care medicine.
Speaker: And it's an effect size that is tangible and palpable and not such that you have to treat 100 patients in order to see one positive outcome.
Speaker: And considering the numbers that we're seeing, I think that we can all do the math real quickly.
Speaker: And it does add up to a lot of saved lives in this pandemic.
Speaker: Yeah, absolutely.
Speaker: Absolutely.
Speaker: Another aspect of Prosperol that you mentioned that I think is worth reemphasizing is that because it included all these studies with different steroids, that effect that you're quoting is on the use of steroids and includes dexamethasone, methylprednisolone, there's only one study, and hydrocortisone, correct?
Speaker: Correct.
Speaker: Yep.
Speaker: Yeah, I think it suggests pretty strongly that it's a class effect, that this is a steroid class effect, and that there's nothing specific about texamethasone from the recovery trial, that it's the only one that will benefit people, and that it's corticosteroids in general.
Speaker: And I think that also helps us with the mechanism that suggests that all of these are similar in that they're anti-inflammatory, and it's decreasing the inflammatory response from the body and the damage that's done from that inflammatory response.
Speaker: that's providing the benefit to these patients.
Speaker: One thing that Prospero looked at, sorry, Sergio, the other thing that Prospero looked at is dose and couldn't actually see a difference in dose either.
Speaker: High dose like was done in Codex versus a lower dose that was recovery and the other trials.
Speaker: So, you know, the hypothesis at least, maybe strong hypothesis from that is that you don't need a high dose and that low dose may actually be as beneficial as a high dose.
Speaker: I think there may still be a little bit of a lingering in that question, but at least for now, I think the data suggests that low dose is as good as high dose and you don't have to use high doses in these patients.
Speaker: Which I think is important in terms of side effects, which was going to be my next question.
Speaker: One of the aspects that perhaps we didn't get as much information as we would want because of the design of recovery and some of the other trials was what's the impact on side effects or potential complications from steroids?
Speaker: Yeah, I completely agree.
Speaker: Some of that you can read out of the editorial that the way these trials were done in an open label and in the middle of a pandemic, collection of side effects in a rigorous way was just not a high priority.
Speaker: And we still have a little bit of a void on truly understanding what might be the side effects in these patients.
Speaker: Many of you listening to this podcast probably are using steroids in your patients, so you'll relate a little bit to when I've used them
Speaker: I've found that there's obvious hyperglycemia.
Speaker: It's not a huge side effect because we usually can treat it with some insulin and treat it pretty well.
Speaker: But the bigger side effect that we've seen that we've really struggled with is this delirium and almost like a psychosis.
Speaker: It appears, at least to me anecdotally, it appears to be worse in elderly patients.
Speaker: And unfortunately, the elderly patients appear to be more affected by COVID and tend to get more severe disease.
Speaker: So we're seeing those in our ICUs and in our hospitals more.
Speaker: And the delirium is, at times, treatment limiting.
Speaker: And what I mean by that is, is that it's so bad that, you know, you struggle to keep the patient safe because they really are so confused and they're all over and they're trying to climb out of bed and they're pulling at things.
Speaker: And those, I think, are hard because we have,
Speaker: in-depth discussions at the patient's bedside on rounds about, you know, do we need to stop the steroids?
Speaker: Do we need to, is this side effect so bad that we can't continue this treatment, which we now have data suggest is really beneficial for the patients because we're hurting them with the steroids.
Speaker: And I think those are the hard things to understand and the hard, the data that we need, the question that we need more data on to truly kind of understand, you know, what to do in that situation.
Speaker: As the pandemic evolved, there was a big debate amongst people who were recommending all sorts of therapies with the argument that better to do something than nothing and with the argument that if we wait for the trials, we're never going to get the answer and people are going to die.
Speaker: I think that the steroid story is a great story to illustrate that when people get together and we have the numbers, that is the right time to do these trials so that we can get some answers.
Speaker: And I think that's a story that is worth also
Speaker: underlying because it's been a constant, I think, discussion amongst clinicians lately.
Speaker: In order to bring things together, Todd, based on all that we discussed, where do you think today the use of corticosteroids stands for COVID-19 ARDS, and what's the impact of these studies in COVID on other patients with ARDS?
Speaker: Yeah, I think in non-COVID ARDS, there are still some questions about what is the effect of steroids.
Speaker: DEXA-RDS gives us some data, but, you know, it's not like we have in COVID where we have thousands and thousands of patients in multiple trials that we put together and have a similar signal.
Speaker: So I think there may still be some randomized trials in patients with non-COVID ARDS to better understand the effect of steroids in that group.
Speaker: But I think it's becoming harder and harder because I think as the data are starting to come together and come out and we're gathering more and more of it, more and more of the data, I think the tide is kind of swinging to steroids may truly be a treatment of choice in patients with non-COVID ARDS.
Speaker: It's an easier question for me in patients with COVID ARDS.
Speaker: And, you know, you could tell from the editorial that Holly Prescott and I wrote,
Speaker: that both of us feel pretty strongly that steroids should be the standard of care for patients with severe COVID-19.
Speaker: Patients in the ICU with respiratory failure from COVID should get steroids unless there's just a huge, big contraindication and a reason they can't get it.
Speaker: I think the data are pretty convincing that they improve outcomes and you can pick some outcomes, come off the ventilator, keeping you off the ventilator, saving your life, getting you out of the hospital,
Speaker: I think it has a positive effect on all of those.
Speaker: And so I think steroids are the, I agree with the World Health Organization, I think they're the kind of new standard of care for critically ill patients with COVID.
Speaker: I do think that there are still a lot of unanswered questions.
Speaker: The question of, that we talked about earlier, what do you do when you start a patient on steroids and then they have such bad delirium that you have a hard time keeping them safe from themselves?
Speaker: I think that's one question.
Speaker: I think dose is a little bit of a question.
Speaker: Maybe that our preliminary data suggests dose doesn't matter, but I think we need more better data in that realm.
Speaker: And I think the population for steroids is another big question.
Speaker: It's pretty easy.
Speaker: We're in a good spot for the intensivist because I think it's pretty easy that the patient that's sick with respiratory failure from COVID, that population I think is pretty clearly defined that we should probably give steroids to.
Speaker: But we're starting to see articles that there's different phenotypes of those patients.
Speaker: And maybe steroids work for a predominant phenotype, but not for a less predominant phenotype, a phenotype that maybe is more vascular coagulopathy type than a inflammatory phenotype, for example.
Speaker: And it's never, a bunch of us have talked about this, it's never really made sense to me that the marker of will steroids work or not is whether or not somebody puts you on oxygen.
Speaker: That just doesn't seem to make a ton of sense to me.
Speaker: And I think work needs to be done to figure out better markers and better indicators of which patients truly benefit from, which patients with COVID truly benefit from corticosteroids and which patients with COVID either don't benefit or maybe even potentially get harmed by corticosteroids.
Speaker: Those biomarkers may be inflammatory biomarkers.
Speaker: They may be something we've never actually studied before, but I think there needs to be work in that area.
Speaker: And then the last kind of question to me, people have asked me, well, can you use it with remdesivir?
Speaker: Can you use it with antivirals?
Speaker: And that to me, isn't even a question.
Speaker: I think it's pretty clear that they work in different manners.
Speaker: The mechanisms of action is entirely different between remdesivir, which is an antiviral and steroids, which is an anti-inflammatory.
Speaker: And I think you both can use them together and maybe are helped by using them together.
Speaker: Maybe that's even better to use them together.
Speaker: So I think using them together is the right approach.
Speaker: The real unanswered question to me is duration of therapy.
Speaker: And some of this will resonate with, I think, on the podcast.
Speaker: We have a number of patients that we see that after 10 days of steroids that we've been using from the recovery trial, day 11 and day 12, the patient seems to be doing worse.
Speaker: Their inflammatory markers, like their CRP or their ferritin or their LDH or ASD or pick an inflammatory marker, are increasing.
Speaker: And they almost look like they got rebound inflammation from us stopping their steroids.
Speaker: In that population, should we give them their steroids back?
Speaker: Should we restart it?
Speaker: Did we just go too short?
Speaker: Do they need a longer course?
Speaker: Should we taper it?
Speaker: I think those are all questions that need to be answered.
Speaker: And I think they're questions that could have significant contribution to improving the care of these patients and improving the outcomes of patients that we're treating already with steroids, because we know that's the right thing to do, but we don't really know how to implement that in practice and get out of treating them with steroids.
Speaker: The patient who you treat with steroids, who got better, who is already out of the ICU and on their way home, that patient stopping the steroids, I think seems to be a reasonable thing to do.
Speaker: But that patient who's still critically ill and or getting worse, I think we still have that question to answer.
Speaker: And I think that, like you mentioned, these are very important questions that hopefully we will be able to continue to study and understand.
Speaker: But I think it's worth emphasizing once again
Speaker: the success story of really getting data in the midst of a pandemic and finding a therapy that seems to have an impact on important patient outcomes such as mortality and getting off the ventilator as a big win and I think a lesson for future pandemics if they were to come that the sooner we start organizing in terms of collaborating scientifically, the more likely we are to find answers that ultimately will help our patients.
Speaker: Yeah, absolutely.
Speaker: And I think I would take it even a step further, Sergio.
Speaker: I would say, and we say this, I think, in the editorial, that this may set the stage and the standard for collaboration even outside of pandemics.
Speaker: You know, there are multiple groups that have questions about the same things in critical care.
Speaker: You know, there's some big high-level topics that we'd like answers on that have multiple large networks doing studies on them.
Speaker: And although there have been preliminary talks between some of those networks that I know about, for example, in general, those networks work in an isolated fashion to try and answer that question.
Speaker: And I think, you know, the World Health Organization has shown us a path forward of a way of being more efficient and answering these questions in a faster manner and a more efficient manner for the benefit of our patients, honestly.
Speaker: And hopefully that'll carry over even, that'll be one of those silver linings, as you said,
Speaker: from the pandemic that will carry over even into non-pandemic research that we can be collaborative in that way.
Speaker: Absolutely.
Speaker: Todd, I really appreciate your expertise on this topic.
Speaker: We customarily will end the podcast with a couple of questions that are outside of the context of the clinical topic.
Speaker: And if that would be okay, I would like to go that direction.
Speaker: Yeah, that'd be great.
Speaker: So the first question, Todd, relates to books.
Speaker: Are there any book or books that have influenced you the most or that you have gifted more often to others?
Speaker: Yeah, you know, I'm not a huge book reader, so this is a little bit of a hard question for me.
Speaker: I read a lot of medical literature and not as much book reading, but one book that sort of has meant a lot to me and has sort of stood out to me is a book called When Breath Becomes Air that many of you, I'm sure, have potentially read by Paul Kalanithi.
Speaker: And Paul was actually a neurosurgeon who, during residency, developed EGFR-positive lung cancer and ultimately passed away from his lung cancer, but writes this autobiography.
Speaker: And there are a number of sort of stories in the autobiography that hit home with me.
Speaker: One is that, you know, we're on one side of this patient-doctor relationship, but we are not immune from becoming the other side, where we're the patient.
Speaker: And it only t akes a few minutes
Speaker: few times of becoming the patient to re-examine and look at that relationship differently when you're on the doctor side.
Speaker: I think many of us have experienced that and sort of can relate to that.
Speaker: The other thing that this book sort of sits with me well is that probably the biggest achievement, sounds kind of funny, maybe it's not the right way to say it, the highlight of my medical career so far is that
Speaker: my mentor came down with pancreatic cancer and asked me to take care of him and provide care at the end of his life.
Speaker: And although that was hard to do, it resonates with me that it was a very, very, very special thing.
Speaker: And it opened up so many additional thoughts about medicine and that it isn't just about care that's provided, it's about compassion and it's about empathy.
Speaker: And, you know, Paul says a lot in his book, it's about morals.
Speaker: and doing the right things.
Speaker: And I think that resonates a lot.
Speaker: The book came out shortly after I had cared for my mentor.
Speaker: And so I think it was sort of an open nerve that it kind of touched and really kind of rested and sat well there.
Speaker: So I think that's probably the big book that I like to talk about.
Speaker: It's an excellent read and we'll definitely link it in the show notes.
Speaker: The second question relates to something that you believe to be true in medicine or in life.
Speaker: that most people don't believe or act as they don't believe?
Speaker: Yeah, this one's a little nihilistic maybe.
Speaker: I think we in medicine want to do stuff and we like doing stuff.
Speaker: And sometimes doing things is not necessarily the right thing.
Speaker: So that's part of it.
Speaker: The other part of it is that in critical care, and this is specifically
Speaker: true for critical care, unfortunately.
Speaker: We've learned a ton of physiology, and then when we've taken that physiology and tried to apply it to the bedside, we've often been wrong in what we've been doing.
Speaker: And when we study things like higher tidal volumes, which improve oxygenation, but then when we studied it, it showed that it actually reduced mortality, or, you know, DVT, not DVT, sorry, GI prophylaxis,
Speaker: This seems to make sense, but when we study it says, well, it may not be saving lives and it could be harmful to people.
Speaker: And so in the ICU specifically, sometimes doing stuff, even doing stuff that physiologically is sound, may not be beneficial for our patients.
Speaker: And so, you know, I'm very much an evidence-based guy, and I think you got to have the evidence because the physiology all makes sense.
Speaker: And then when we apply it, it often doesn't happen in vivo like we think it's supposed to.
Speaker: So I think one of the, you know, it's a funny question, funny answer to a question, what do you believe to be a truth in medicine or life?
Speaker: Most people, other people don't believe.
Speaker: And I think the answer to me is that just doing something isn't necessarily good for our patients.
Speaker: It's something that I think was also highlighted during this pandemic and the discussions and the behavior of many clinicians and emphasizes the reasons why we need to really try to get the data
Speaker: and use the best available evidence always to try to answer those questions.
Speaker: Yeah, I don't know if everybody else ran into this like I did.
Speaker: At my institution, what happened at the beginning of this pandemic was that lots of non-intensivists became involved in intensive care.
Speaker: And what I mean by that was, is I had daily talks with rheumatologists and daily talks with my hematologists, and they were trying to relate things that they saw and recognized in their field to the ICU.
Speaker: And oftentimes, I knew they were wrong.
Speaker: So they'd say, this is an inflammatory process, right?
Speaker: We have to give an IL-6 receptor antagonist because IL-6 levels are high and we have to do that.
Speaker: And while I don't know if IL-6 receptor antagonists are good or bad in this disease, I was fairly confident that just giving them because IL-6 levels were high, we had tried that before in critical illness and it hadn't gone that well for us and that we shouldn't just
Speaker: say if A is abnormal, we should do something to make A normal again, because that's been shown numerous times that that's not what's beneficial for our patients.
Speaker: Not to mention, I think, the lack of perspective on time, and especially with IL-6 antagonists, remembering that they came to rheumatology after failing in critical care.
Speaker: Yeah, absolutely.
Speaker: And that's how they became big in rheumatology.
Speaker: Somebody salvaged them in another field.
Speaker: The last question, and I know that you have to go to take care of patients, have a clinical meeting, is what would you want every intensivist that's listening to know could be a quote or a fact to close?
Speaker: Yeah, I think the quote that was taught to me as I was growing up in my medical life was, don't just do something, sometimes just stand there.
Speaker: And, you know, that goes to what we talked about before, which is that just doing something isn't necessarily beneficial for the patients.
Speaker: And sometimes just giving the patient time and letting what you're already doing work is the answer to these problems.
Speaker: And sometimes not doing something but being with the family and being with the patient and being present there is the right answer, too.
Speaker: So that quote runs through my head a lot.
Speaker: And when I'm sitting there thinking, oh my gosh, I don't know what to do.
Speaker: What should I do?
Speaker: And then that thought comes through of, well, maybe the thing is don't do something, just stand here and let things develop and try and take in the picture and see what happens.
Speaker: So I think that's the quote that I would relay is don't just do something, sometimes just stand there.
Speaker: That's a perfect place to stop.
Speaker: Todd, I really want to thank you for your time and sharing your expertise with our audience.
Speaker: I look forward to talking with you about other topics related to critical care.
Speaker: in the future.
Speaker: Thank you very much.
Speaker: Sounds great.
Speaker: Thanks, Sergio.
Speaker: Thanks for the time.
Speaker: Thank you for listening to Critical Matters, a Sound Critical Care podcast.
Speaker: Make sure to subscribe to Critical Matters on Apple or Google Podcasts and share with your network.
Speaker: Sound Critical Care is transforming the way critical care is provided in hospitals across the country.
Speaker: To learn more, visit www.soundphysicians.com.
