between nards, pards, and ards. And this title was given to me, so we're gonna lead with that. But that's fine. We can, we, I have some thoughts. But, I don't know how to advance. Scroll wheel? No, that's fine. That's fine. These are my disclosures. So ards was defined in 1967 for adults by adults, and pretty much all the key elements are there. This is the 1967 Lancet publication, which, I tell this story every time, and I'm gonna do it now because it's every time. They like submitted this paper to like New England and then JAMA and then Blue, and they got rejected each time, largely because of like considerations that like, there's no way that peep was actually working, OK. And so that, that just seemed ridiculous, because people thought that it would put you into heart failure, or like squeeze your box and, and, and be damaging. And so they didn't like the idea that there was a suggestion that peep could be helpful here. So then, after rejection by Blue, they actually upped it to Lancet, which had a higher impact factor. It still does. And they got it accepted. And so like, I use that as like motivation. Every time I get like a review, I'm like, that I completely disagree with, I'm all like, oh yeah, I'll show you. And it's the, it's the 67 Ashpaw publication that really motivated that. All the key elements are there. It's acute onset, it's hypoxemia, peep, that controversial peep from the get-go is a problem. Nothing has changed in the last 50 years. All right. This is the pediatric definition from version one, where, and the, excuse me, in version two now, where we have changed this to like, all patients less than 18 years of age without active perinatal lung disease should be diagnosed with PARDS using PALIC-2 criteria. Practitioners can use either PALIC-2 or neonatal definition. The unfortunately acronymed Montreux NARDS for neonates and can use either PALIC-2 or Berlin ARDS for young adults. And this was an ungraded remark that was placed after a definition statement. So let's talk about NARDS. So NARDS also excludes different perinatal diseases, OK. But it explicitly defines them a little bit. Probably because it was designed for neonates. And so, like, respiratory distress, distress syndrome, or like, or, and transient acute and newborn congenital anomalies as a primary current would not be considered NARDS. So, what are we trying to do here? We have three separate definitions. And I think we're all acknowledging that we're getting at the same physiology, or at least we're trying to. There's some exudative inflammatory pulmonary edema that happens from a direct or indirect insult. And we all seem to have settled on seven days as a operational definition. Which is probably the same way why we figured out that seven days of antibiotics is the right answer. Is like, we just kind of made it up. Because it sounds right. There are some notable differences, OK. The pediatric ARDS definitions, both last version and this version, allow unilateral as well as bilateral. Whereas NARDS and adult ARDS, like, still stick with bilateral. With kind of a nod to, like, the systemic nature of the, of the process that you're invoking. The etiologies, OK, are somewhat different. Meconium aspiration would be excluded from pediatric ARDS, but would be allowed for neonatal ARDS. And there's a different numbers of hypoxemia categories. You know, mild, moderate, severe, versus non-severe and severe that we have for, for pediatrics in the, in this version of the definition. But what's the point? What are we trying to do here, OK? Do we think these patients have a shared, this inflammatory lung injury across the age spectrum? Do we think it has a shared epidemiology? Do we think it has a shared physiology? Should we have shared managements? OK, and should we be thinking about this in terms of enrollment for trials to test out some of the questions for one, two, and three? So, in among the more overused slides in this particular SCCM 2023 is this one, OK, where there's been a lot of attention about latent class analysis for subphenotyping ARDS, including this, like, seminal work from Carolyn Kalfi, in which you have, like, latent class analysis performed on clinical and biochemical markers, in which you end up with two subphenotypes of ARDS defined by higher levels of hyperinflammatory biomarkers and lower levels of blood pressure protein C and bicarbonate in the hyperinflammatory subtype. And this was reproduced in a reanalysis of Bali, which found similar sort of signatures, OK, suggesting that hyper and hypoinflammatory ARDS may, in fact, be a thing. And so we've, we've done this, we've focused so much on this because of, like, stuff alluded to by Brian and stuff alluded to by Robbie is that ARDS is heterogeneous, and we're looking for ways to get past the limitations of our syndromic definitions, because our syndromic definitions have not changed in the last 55 years, and have led to many, many negative clinical trials. However, the, our favorite syndromic, our favorite subphenotyping strategy is not specific to ARDS, so this is a cohort study in which people use biomarkers to subclassify adult ARDS, much like the Carolyn Calfee paper did. And they, in fact, found two subphenotypes of ARDS, so they reproduced this. However, when they performed the exact same analysis on patients who did not meet criteria for ARDS, they still found two subphenotypes of acute respiratory failure, which I don't know what ARFA is, acute respiratory failure A. I'm sure it made sense at the time, but. And people who are looking at kidney phenotypes, OK, like, also were able to reproduce two subphenotypes of kidney injury, subphenotype A and B, OK, according to a combination of clinical variables and biomarkers, OK, suggesting that hyper and hypoinflammatory is not restricted to lung injury either. It is, in fact, possibly common to critical illness. And one of the things that people have talked about, like in that Nature Medicine review, which I now feel remiss for not having included in this talk, like, it's possible that we're moving past biomarker-based diagnoses, and that instead of a traditional PICU ward where you have abdominal sepsis with ARDS, or pneumonia ARDS, or polytrauma, or post-cardiac arrest, you actually have more biomarker-directed phenotypes, such as, like, angiopoietin 2 dysregulation and IL-1 dysregulation. Two of these patients, for example, have infections that necessitate antibiotics. Three of these patients have angiopoietin 2 dysregulation. OK, two of these patients are immunosuppressed and may benefit from some sort of immunostimulant. So things like that, where we're kind of getting at the idea that we have overlapping physiology, and that our clinical syndromic definitions have limitations that have been hurting us for purposes of clinical care and for trials. And this is not a unique concept. This is like, I mean, oncology came up with this a while ago, where you can have multiple disease process tested for in a common intervention. So the PD-1 ligand story is actually, like, a good one, where you can have different anatomic cancers, OK, but they all have, like, dysregulated PD-1 expression as part of their physiology, so you could use that as, like, a therapy to test in those cancers. So you could imagine that for ARDS as well, where you can conduct a basket trial with different nominal syndromic subtypes, like ARDS sepsis or cardiac arrest, with a common targeted intervention, let's just say endotheliopathy, which could be shared between those. So why am I talking about that in the NARDS, PARDS, and ARDS talk? And I will tell you, age is also a nonspecific component of this definition, and I think it's holding us back. So when does NARDS become PARDS becomes ARDS? It's in separate ICUs, they're in separate consent practices, but nobody's shown me that these are different disease processes, particularly given how we've defined them. So to tease a little bit of data about this, we just looked at it, and so asked a very simple question of what is the association between age and mortality in ARDS? And so we had been, any given adult cohort can't answer this question. Any given pediatric cohort cannot answer this question. Any given adult cohort has a median age of 50, and is very underrepresented in young adults, between 20 and 40 years of age. Any given pediatric cohort can only answer the question up to generally between year zero and 18, and is very underrepresented in both adolescents and young adults between the age of 20 and 30, although we all take care of that population, between five to 10% in many ICUs. But if you combine all of those together into a cohort of 5,000 adult ARDS, then you'll have a lot more of those 20 to 40-year-olds, so your confidence intervals don't look ridiculous. And when we did that, then you can actually see there's a nonlinear relationship of age, where there's a somewhat attenuated increase in mortality between the ages of zero and 40, and then it sharply takes up after 40. Sorry, I just thought you guys should know. So, looking at the median age of this room, this may not come as a shock to you, but it's still striking to see it on print. There's precedent for this, okay? Pediatric patients, excuse me, let me take that back. Young adults do better on pediatric chemo regimens, okay? So this is an example of age on the x-axis, and relative survival on the y-axis, and the gap in between the pediatric regimen, I don't know if you guys can see that at the top, but in the left half of the graph, between zero and 20, are patients who are, like the mortality rate for pediatric leukemia on pediatric regimens, and on the bottom is the adult's comparable mortality. There is no reason that an 18 and a 19-year-old should have a 20% difference in their leukemia mortality, other than the difference of being cared for in a pediatric versus adult institution. So, what can we do about this? Well, a little while ago, there was an RFA that was put out by NHLBI and NIGMS together for U01, which was, in theory, gonna fund five studies, who knows, or five of these, but who knows how many they'll end up funding, called the APS. So that was ARDS, pneumonia, and sepsis, and that RFA is a tacit acknowledgment that there's overlapping physiology, there's overlapping clinical, and there's overlapping clinical definitions and clinical aspects, as well as biomarkers, probably, and potentially outcomes that are shared between whether somebody meets ARDS criteria or not, whether somebody meets pneumonia criteria or not, or whether somebody meets sepsis criteria or not. Annoyingly, it is a biospecimen collection of 5,000 adults hospitalized with one of those conditions. So this is actually what it needed to read, Dr. Agarwal, if you're out there. Neonates children and adults across the lifespan, so this is what we need to do, okay? We actually need to prove this, we actually need to investigate this and study this, that children between the ages of eight days to 80 years, there's probably a better alliterative way to say that, but I really couldn't figure out a good way to include neonates in it, but between zero and 80, zero and 100, between zero and 100, we should actually just recruit patients with what we consider potentially having overlapping clinical-like physiology, and in this case, ARDS, but perhaps we should lighten that up, too, because a lot of this talk has been suggesting that critical illness syndromes have a lot of shared physiology, but specifically for ARDS definitions, okay? We should investigate this, we should investigate to see when NARDS becomes PARDS, becomes adult ARDS, and moving forward, this will help us define the epidemiology, potentially different inciting etiologies, clearly different comorbidity profiles, and potentially different concurrent organ failures which determine outcome, and if you define the molecular epidemiology across the lifespan, you can actually see how these subtypes shift between ages and between diagnoses, or at least between nominal diagnoses. You can clarify the organ failures and see which ones are bystanders or which ones are causal. You can clarify the role of endotheliopathy, which may be a shared organ failure between many of the syndromes we worry about, and you can characterize immune cell trajectory, which is completely understudied as a biomarker. It is starting to get a little bit more traction as that technology becomes more affordable, and we can finally design better trials and do what we all wanna do and profit for the kids. So what do you mean by better trials? And I should preface this, that a good portion of my career is firmly, firmly in this field of some thoughts on how everyone else is bad at research, so just take it with a grain of salt. But let's go back to this graph. You could envision that children under a certain age, particularly young infants and neonates, may have a pathophysiology with ARDS, or however you wanna define it, that may be particularly amenable to nitric oxide, like both between impaired lung growth, like a more reactive pulmonary vasculature, underdeveloped pulmonary capillaries, stuff like that. You can also imagine that between the ages of, say, five and 50, the negative effects of aggressive recruitment and PEEP may actually be better tolerated by healthy hearts. So it's possible that the negative high PEEP trials, and despite the fact that there have been seven negative high PEEP trials, we're still recommended to use higher PEEP in adults, which is crazy to me, but nevertheless, here we are. You could imagine that this is actually an age grouping in which this actually makes more sense, rather than subjecting 60-year-olds to recruitment maneuvers of 60 over 35, which they did until they had heart attacks and then they stopped. But they did it, which is crazy to me also. A lot of stuff's crazy to me, in retrospect. I'm gonna conclude here. Critical care syndromes are nonspecific, and there may be, and likely is, shared pathophysiology across syndromes. And I submit to you that there's likely shared pathophysiology across the lifespan. And all of this can lead us toward better trial design. With that, I'll take questions. But apparently, at the end. Thank you.

Acute Respiratory Distress Syndrome

ARDS

Age groups

NARDS

PARDS