Good morning. Thank you, Chairperson, and thanks to SCCM for the invitation and the opportunity to discuss this topic with my dear friend and colleague, Dr. Prescott. Just a brief acknowledgement to my fund, as I'm mostly funded through UK public funding programs like the National Institute of Health Research, MRC, and the Wellcome Trust. I do have an intellectual conflict of interest. I run a precision medicine immunology program. I do not have any financial conflicts to declare. I think you know that you're getting a pro-con debate because there are more systematic reviews, opinion pieces, and editorials than original research publications on a topic. We've got more of these than randomized control trials. And to kind of frame this for a pro-con debate, what I thought I would do is to give you an overview of why do we think about immunomodulating patients with coming-to-acquire pneumonia and ARDS, and a brief overview of the mechanisms of corticosteroid therapy in this context, and then the clinical evidence summary, which I've essentially chosen two of the most recent systematic reviews and the corresponding guideline documents that comes with those systematic reviews, and then some take-home messages before handing over to Dr. Prescott for the con version of the talk. So biological rationale, and this is an important kind of few minutes of the talk. There are a lot of things that we do in critical care that doesn't have any biological rationale, and this is one of the things where we do have some rationale, but what we haven't done thus far is to define what we mean by the dysregulated immune response during ARDS and pneumonia. I'm just going to kind of start with that and then get into the lung biology as to why we should do this. So we've just published this paper. This is going to come online either end of this month or early in February, where we talk about what we mean by a dysregulated immune response in the context of sepsis, but pneumonia is an infection, and therefore some of these things that I'm going to talk about is relevant, and ARDS, the most common etiology, is pneumonia or infection. So dysregulated immune response in those scenarios is defined as a pathological disruption or alteration of the homeostasis of the immune system, and it involves three things really. Immune-mediated resistance, alterations in the disease tolerance, and resolution mechanisms. The way to think about immune-mediated resistance is the kind of common pass-away statement of there is inflammation, and that is what we mean by immune-mediated resistance. So if you take this sort of framing of a dysregulated immune response, infection or sepsis or ARDS state, you have two things, three things going on. Abnormal resistance mechanisms, and impaired disease tolerance mechanisms, and impaired resolution mechanisms. And I'll come back to what is specific for ARDS in the next few slides. If we focus in on the abnormal resistance mechanism, we have two things that are going on. There is excessive inflammation in ARDS, and there is also concomitant immunosuppression in both ARDS and pneumonia. And this has been known for a long time, but we haven't actually thought about it like this, and that helps reframing as to how we should manage immunomodulation in these patients. So this is an example of the biology of community-acquired pneumonia from one of the Nature Reviews in Disease Primers by Tony Torres and colleagues. If you think about the lung and the alveolar structure, within the lung there are alveolar macrophages. Those are the primary kind of cell types that respond to an insult. And when the alveolar macrophages respond, they produce some chemokines and cytokines that activate the neutrophils. And these neutrophils kind of migrate across the endothelium and the alveolar capillary membrane into the alveolar space, and kind of contributes to the inflammation. And it also kind of activates the coagulation system and the platelets. And on top of it, there is a respiratory epithelium that has got some inbuilt resistance mechanisms that are able to kill pathogens. So in pneumonia, all of these things are activated. And the way to kind of address some of these excessive inflammation that happens in pneumonia is to give a drug, and that could be corticosteroids as one of the drugs that we could use. And the immunobiology of ARDS also has a very, very similar kind of paradigm, where this is a review, again, by... from Nature Reviews. It's Michael Mathieu and colleagues, where they talk about, in an ARDS lung, there is the alveolar capillary membrane that is disrupted. So you get exudative pulmonary edema. There is activation of all the cells that I discussed in the previous slide, the alveolar macrophage, the neutrophils. And there are new cells coming into the alveolar space. And alongside that, you get the disruption of the capillary membrane, and all of the tight junctions are gone. So ARDS biology can be thought of as an extension of the pneumonia biology that I highlighted in the previous slide. So if you kind of take this basic premise of what the disease that we are trying to treat is, corticosteroids seem to have biologically plausible roles. But there are many things in critical care where biological possibility hasn't led to interventions in the future. But let's kind of give corticosteroids a try. It's got two different actions, genomic effects and the non-genomic effects. Corticosteroids don't work immediately through the genomic effects. I'll... I'll explain what I mean by that in the next slide. The reason why I put these two broad categories of action is the side effects that you see with corticosteroids, like immunosuppression, GI bleeding, muscle wasting, et cetera, is mediated through its genomic effects, whereas some of the early cytokine suppression responses are through the non-genomic effects. And if there is a way to kind of use the corticosteroids in such a way that we can enhance the non-genomic effects whilst reducing the genomic component of its effects, you might actually get the risk-benefit profile sorted. And I'm not saying that we have done that yet. And as I said at the introduction, there is potent anti-inflammatory and pro-resolution effects for corticosteroids. So coming back to the genomic effects, broadly, you have two mechanisms that are going on. There is transactivation. With transactivation, what you're trying to do is to activate the inhibitory molecules that suppress the inflammation. For example, IL-10 is a good example where you enhance it with corticosteroid therapy. There's also transrepression, which is the kind of mechanism we always think about, which is anti-inflammatory by suppressing pro-inflammatory molecules like IL-6 and IL-8, transcription of those. So there is broadly two mechanisms. And different corticosteroids have got differential effects on these mechanisms. And we don't even remotely consider that at the current kind of paradigm of using corticosteroids. As I said, there are loads of non-genomic effects. And they are broadly mediated through stabilizing the cell membrane, influencing the arachidonic acid metabolism. So things that contribute to your pro-inflammatory cytokines and other mediators that are anthromboxanes, that could be affected by corticosteroid therapy. And it also kind of affects the T cell responses that you see. And those are the mechanisms that actually gave us benefit during COVID-19 patients early on in the pandemic. So it's not as if we don't know how to use steroids. It's just the fact that we haven't quite figured out what's the right thing to do for chymotoquandamonia and ARDS. So just to go back to my figure and summarizing this whole idea of immune response could be dysregulated in patients with ARDS. And that dysregulation could be broadly divided into excessive inflammation and immunosuppression. That is your abnormal resistance mechanisms that contribute to some of the early and late deaths that you see. There is impaired tolerance and resilience mechanisms. There is an uncertain association with early or late deaths. And there is impaired resolution mechanisms in ARDS that contributes to your persistent respiratory failure, long-term fibrosis, post-ARDS recovery. And all three mechanisms are relevant in the context of ARDS. And if you were to think about glucocorticoids, they have a positive role in reducing inflammation, as I highlighted in the last few slides. They worsen immunosuppression that is known, especially if the dosing is inappropriate. And they also have pro-resolution mechanisms which we never think about. And in terms of timing of therapy, all of these mechanisms come into play, especially the pro-resolution mechanisms are much later on in the disease process. We haven't really honed in on how to treat using corticosteroids yet. So with that sort of basic biology background and, you know, fundamental biology says this should work. And we are still trying to figure out how to do it. So let's think about the what's the evidence summary here and what kind of, what do guidelines say as we speak. So if you think about community-acquired pneumonia, I have decided to group patients here into severe-cap and ARDS because I think if we don't do that, we have a risk of extrapolating evidence from what we think is community-acquired pneumonia evidence into ARDS evidence. And that is probably the wrong thing to do. So if we group these patients into two, severe-cap and ARDS, we may actually start to make a really true pro-debate for this therapy. So if you think about community-acquired pneumonia, a mild is community-acquired pneumonia treated on an outpatient basis and moderately severe as those who require hospitalization and organ support respectively. So the gradation in the pneumonia severity could have a differential treatment effect with corticosteroids. And this is one of the many meta-analyses. If you have just blinked in the last few minutes, there'll be another one published. So this is one of them. And this is led by Pete Steyer and colleagues where they identified 18 randomized control trials to date on community-acquired pneumonia and steroid therapy. And what they show here is group the patients into less severe disease and more severe disease. So less severe probably reflects mild community-acquired pneumonia. And there, the evidence for supporting corticosteroid therapy is minimal, at least from this systematic review. The risk ratio is around 1, and the confidence interval crosses 1. In contrast, in a much more severe form of community-acquired pneumonia, your risk ratio is much lower, and it's around 0.6 with the confidence interval that it doesn't cross 1. And with this level of risk reduction, you would be kind of, as a clinician, you would find it hard to keep equipoise for not giving somebody corticosteroids. But the challenge here is the trials have used different steroids at different doses. And probably the only consistent signal that you see here is with hydrocortisone in the middle, where you see a beneficial effect. And that, I think, reflects some of the guidelines, which I'll come on to in the next few slides, where if you think about community-acquired pneumonia with a degree of shock, even the surviving sepsis campaign guidelines do support that as an intervention in that context. Whereas if you extrapolate this evidence into somebody who's mechanically ventilated with ARDS due to pneumonia, this does not work. So there is a distinction that we need to make here before extrapolating evidence across illnesses. And I kind of, as an example of that point that I made, I just want to highlight this paper. This is from Majuri and colleagues published in ICM recently, where they kind of took patients with community-acquired pneumonia, but requiring critical ICU admission. They gave methylprednisol on 40 milligrams per day, and the usual care was placebo, and the primary outcome was 60-day mortality. As you can see here, there is no difference. And even in patients on mechanical ventilation, there is no difference. If you've got a pneumonia patient who is on a mechanical ventilator, the diagnosis probably there is ARDS. It is not just some pneumonia. And I think that distinction has to be kept when you think about steroids in this context. And this is an example of one of the many guidelines. Again, if you blink, you'll see another guideline on the topic. So this is the ERS-ESICM-ECMID-ALLAT guidelines for community-acquired pneumonia, where they suggest that in patients with severe CAP, the use of corticosteroids in shock is useful. And it's a conditional recommendation with a low-quality evidence. So the evidence base that we are pinning some of these interventions on is really poor. But given the outcome benefit that you see with community-acquired pneumonia that is not requiring, that is not in severe ARDS or moderate ARDS, that is a population that might be potentially of value here. And that evidence doesn't account for some of the other viral diseases. So that kind of wraps up my pneumonia piece for the PRO debate. So very little evidence, but there is some population where it may be useful. Let's think about ARDS. And I said all the way through, we are at risk of extrapolating evidence from community-acquired pneumonia into ARDS. And we should refrain from that for the points that I've already made, but I'm just going to kind of reinforce with some more points, too. And the reason why there is such kind of interest is because probably, aside from the COVID-19-related ARDS trials and the guidelines, it is this paper. This is the DEXA-ARDS paper published by Isis Villar and colleagues. This is in Lancet Respiratory Medicine. Stopped early. Mortality was not the primary outcome. Dexamethasone dose is way more than what you would use in community-acquired pneumonia or in COVID-19. It is 20 milligram dose for the first five days and then 10 milligram dose for the next five days. So 10 days of high dose dexamethasone. That is not what you would give for pneumonia patients. So it's a very different trial. And in that trial, there was a mortality benefit, even though mortality was not the primary outcome. And there is also a primary outcome ventilator-free days benefit. The trial was stopped early. And we know from previous literature, if you have a trial that gets stopped early for efficacy signal, it's seldom true when you do a large pragmatic effectiveness trial, especially the signal of this magnitude, which is 15% reduction in mortality. Nobody's going to see it. If we were to see it in ARDS and steroids, we would have seen it by now. There is something about extrapolating evidence. And this is another scenario where if you start to extrapolate evidence, we might actually do more harm than good. And this is a systematic review of non-COVID-19 ARDS by Bram Roschweig and colleagues, where they highlight that there is a significant treatment effect of improving outcomes with corticosteroid therapy. There are a few things here that you may recognize. ANAND 2006 was a septic shock study. It was not an ARDS trial. It's a subgroup of the ARDS trial extrapolated as ARDS here. There are trials where the sample size is lower than the annual number of ARDS patients you would see in your unit, and you would treat personally in your unit. So, some of these is very low-quality evidence that we're trying to extrapolate from. And the last point about risk of extrapolation to ARDS is this wonderful paper, which is the Cape Cod trial in the New England Journal of Medicine, where the authors looked at commuter cornemonia and ARDS and gave some hydrocortisone and showed that it improves outcome. And I want to kind of focus in, focus your mind again, the distinction between severe CAP and ARDS. The mortality of patients with commuter cornemonia in this trial was around 12 percent. That is 28-day mortality, a 90-day mortality of around 14 percent. I don't know about your patients. My ARDS patients have got a mortality of at least two or three times that. So, this is extra. We can't extrapolate this evidence to treat ARDS patients with corticosteroids. And the guideline that I'm going to refer to for ARDS is this American Thoracic Society guideline published in the Blue Journal a few weeks ago, where the authors identified 19 RCTs, including 2,790 patients. I just want to kind of go back one slide, a couple of slides, because the, sorry. Can we go back a few slides, please? One more. One more. Previous slide. Thank you. So, 19 ARDS RCTs are not there. This is a systematic review of non-COVID ARDS RCTs, and we have at best six. So, I think, again, even within guidelines, there is extrapolation that is going on for ARDS. So, to kind of come back to this 19 ARDS RCTs with 2,790 patients, we don't have that, but there is a broader population that the authors have included in the trial. The pooled analysis includes around 17 studies, and they highlight a risk ratio for around 0.8. And they do conclude that actually it's a conditional recommendation. There are so many unknowns within the ARDS literature. So, to kind of do the pro, it may sound like a con debate, but actually, I want to kind of highlight that we have to be careful when using corticosteroids, partly because they are drugs with potent immunomodulation effects, and there are so many uncertainties. And currently, I think, in terms of severe community-acquired pneumonia, probably there is an uncertain mortality benefit, and we don't know what actually is the right dose or the duration of therapy or the right drug. And the recommendation is similar for ARDS. There is something going on where there is probably a signal, but we need to be careful when extrapolating evidence. So, to wrap up my pro part, there are so many systematic reviews, meta-analysis, and guidelines that I was confused at the end of reading how best to give you a pro debate. I hope that I've given you a balanced idea of there may be some patients where this might work. And I want you to kind of take one message out of this talk from me about the pro part, which is separate out severe community-acquired pneumonia of whatever measure you use from patients with community-acquired pneumonia and ARDS. They are different things, and they need different interventions. With that, I'll thank you very much.

corticosteroids

community-acquired pneumonia

acute respiratory distress syndrome

hydrocortisone

treatment guidelines

mortality benefits