All right, well, thank you so much for the introduction and welcome everybody to this afternoon's session. So what I'll try to do is paint a broad picture for thinking about the microbiome in the context of critical illness, including drilling down a little bit on the microbe and microbial ecology, because I think this conceptual framework is actually really essential, to think about what we know, what we can perturb, what we think we may be able to perturb and sort of what are the approaches that will be needed in order to make those perturbations as we think about changing health in the context of critical illness. I have several disclosures that are noted on this slide, neither of which bear any relevance to anything that I'll talk about. So I think there's a few key questions to contemplate as we think about the microbiome and critical illness. The first is, does the microbiome actually cause critical illness? I think most of us could probably answer this pretty quickly. The second, does the microbiome modulate outcome in critical illness? I think that probably most of us have a suspicion of what the answer to this is, but the actual biologic proofs for it may be a little more challenging to come by. And then the third is, can the microbiome be harnessed in order to protect against critical illness or modify the outcome? I think this is actually probably the most difficult of these questions to address, and one that I would suspect may take us perhaps another few decades to actually wrestle with to the point that we have molecular determinants that allow for perturbation. So what I'd like to do is just frame an understanding of the microbiome in order to shed some light on how one may think about or frame the answering of these questions. So I think as everyone in this room probably knows, there has been a real wealth or an explosion of literature within the field over the past, say, two decades, and even in the past few years, looking at the microbiome in the context of health. This has probably been paved very early on in mechanistic studies in mice and has now been scaled to the level of humans and actually understanding, how do we begin to grapple with the complexity of the microbiome in humans? Understanding that the microbiome is not one microbiome, but it is actually multiple microbiomes that exhibit site specificity, that exhibit content specificity, and that exhibit extraordinary heterogeneity between individuals and even in a single individual as a function of time. So shown here are just a few relatively recent papers that I think, why I put these few on this slide, again, there are hundreds that one could have selected, is really to capture sort of what's the scale at which we're starting to think about this problem and trying to understand the interaction of the microbiome with disease. So the first paper on the left really talks about the scale of thousands of humans. So looking at microbiome correlates in composition to health and disease. The second is drilling down to the level of metabolic function and how simple things that we may do in the ICU, such as giving drug A or drug B, may interact differentially with the microbiome in a given individual. This is a more recent example of how a specific microbial signature may impact patients of interest to many of us. And then this is, there are many, many reviews in this space. This is one that I particularly liked because I like the concept of emergent host phenotypes. Sort of really capturing that this is a dynamic space. This is a moving space. And the microbiome adapts with us and in the context of us. And as such, this is truly an emergent property that we need to think about in the context of our patients. So I put this slide in because I really like this reminder. So this is the Global Burden of Diseases Study looking at mortality related to bacterial infection. This was compiled in 2019. What I like about looking at this list of organisms that cause lethal infection is the reminder that the clear majority of these are organisms that live with us. The majority of us have many, if not most of these organisms in our microbiome or may have these organisms in our microbiome at some point in time. So our microbiome is the source for the infections that we try to deal with at the bedside. So I wanna frame this from the microbial ecology perspective because I think this is a necessary foundation for thinking about whether the microbiome is perturbable and how it may be perturbed. So you can imagine that these three colored spheres are different microbes just for the sake of simplicity. And on the X-axis is time, and you can envision this sort of Y-axis is the assembly and how these change as a function of time. So for instance, here, you can see that the blue microbe will increase over time, the red microbe a little less so, the green microbe initially increases, faces a disturbance, and then may be restored within the ecologic context. I think this captures two relevant and in a very simplistic way attributes of the microbiome. One, these are complex and multi-organismal, and two, change over time is met out as a property of the microbiome as a whole, but also at the level of individual organisms. And I think this captures a little bit of the complexity that we have ahead of us. This slide, I wanna try to capture the notion of microbial competition because this is inherent to thinking constructively about the microbiome. Microbes intrinsically interact with each other in both antagonistic and sort of beneficial ways. There are processes of competitive exclusion that can eliminate a microbe from the microbiome. There are resource-based spatial allocations by which certain microbes can become dominant versus others non-dominant in the environment. And these resources are really met out in geographic space within the body. And I think what's important to remember about these competitive interactions that are ecologically driven is that many of these are known to be the result of even single gene level effects. So the concept of sort of a broad probiotic really comes into question when you think about what do we know about every single gene? And the possibility that gene A can impact outcome A. So I think the need for mechanistic clarity is really important here. So this schematic really shows that this is a complex adaptive system. The host adapts, the microbiome changes and adapts, and our host immune systems are distinct. And what I've listed here are several of the attributes of both the microbiome as well as the hosts that are key to contend with when we think about this dynamic interaction. What are the genomic constitutions of both the microbes and the hosts? How do microbes get transmitted? How do they establish the niche? How does the host adapt? What is the host state of immunologic competence? Each and every one of these will influence how the host and the microbiome interact, and therefore what the product of those is in terms of disease. So this scheme from John Alverde I really like because it sort of illustrates the incredible complexity of what we deal with in an ICU. Sort of all of these features at the outer edge of this sphere are different things that happen in an ICU context that can actually imprint or modulate the microbiome of the host. So how do we deal with this complexity? This diagram I included because it really, I think, presents a nice conceptual framework for the ICU patients. So what's on the top is microbiome diversity. What's on the bottom in a parallel fashion is risk of sepsis. So you can follow along the risk of sepsis with its sort of inflection points where the risk is increased, and see these visually as perhaps corresponding to a dysbiotic state, which is a word that's coming a little bit out of favor in the microbiome community. But you could imagine initiation of antibiotics may perturb the microbiome. That may recover and then the patient suffers respiratory failure and has an intubation event which may precipitate another alteration in the microbiome, and then be fine for a while and have yet another event, say a surgical intervention, that may then set the patient up for a septic outcome. So I think understanding the complexity of these interactions and what's happening to our patients over time will be necessary. So how do we grapple with this complexity? I'd propose that we need to both understand the composition of the microbiome as well as the function of the microbiome. And this is a difficult challenge when you think of the number of individual genes encoded by even a single microbe within the microbiome. So let's take a normal pathogen. So a typical bacteria like Staph aureus or like E. coli will have two to five million base pairs of DNA and thousands of coding sequences, all of which are under different regulatory control. This is the magnitude of complexity that we have to think about when we think about what is the functional outcome of the microbiome. And I think we again have to remember that this is a process of adaptation over time. So I want to show just two quick slides of data that illustrate two principles about the microbiome. This first slide is a relatively recent paper looking at a group of neonates that had bacteremia. And I'll just pick one of these. The slide overall shows the whole picture. Let's say Enterococcus faecalis bacteremia. So looking at Enterococcus in the stool microbiome immediately prior to Enterococcus faecalis bacteremia. What you can see is a bloom of the Enterococcus in the children who had bacteremia. And you can see this across six different pathogens. The bloom occurs immediately proximate to the time of bacteremia. This is an association of microbiome change with disease. We don't know in this context that Enterococcus bloom caused the Enterococcal bacteremia. It's an association. And this is an important attribute of understanding the microbiome. Second attribute is illustrated by this particular study, which I'll go over. So this is an animal study where one looks at percent survival from lethal influenza challenge. And what I've schematized here is that as the interferon response in the host goes down, the outcome of the host worsens. So you can see this host does well. These hosts do poorly, predicated on the interferon outcome. This group defined that desaminotyrosine, or a single microbial metabolite, correlated with this interferon response. And what they were able to do was to demonstrate that depending on the microbes that are in the microbiome, they could generate more or less of a desaminotyrosine response. So I'll draw your attention particularly to C. Orbiscendens in the microbiome actually generates a DAT response. And you can see that this nicely protects the animals. And they show that this is based on the interferon response. So this is a very different picture of a causative role for the microbiome and disease modification, where one can pinpoint a specific metabolite generated by the microbiome and map out its role in disease. And I think it's important for us to think about these two attributes, both association and causation, when we think about the microbiome and its ability to perturb critical illness states. And I think we have a little bit of a vexing problem because the complexity of the microbiome and the host immune response is already extraordinary. Superimpose on that critical illness, and it's induced changes to the microbiome, which becomes, in some contexts, the pathobiome and the host immune response. I think this becomes a vicious cycle. And I think this is a really challenging problem that we'll have to deal with as we contemplate perturbation of the microbiome. So just a few closing thoughts for how can we conceptualize shaping the future of microbiome-based intervention. I think key will be to remember there will not be a one-size-fits-all approach to solving this problem. Each host and its microbiome is unique genetically, constitutionally, and as a function of time. Early imprinting and other events in life will shape the microbiome. And this is not static. This is a constantly moving endpoint for each one of our patients. And I think we're really going to have to drill down to the level of genetics and understand microbial genetics in order to be able to get a handle on how we can perturb the microbiome in the context of critical illness. So a few simple conclusions. Understanding risk and designing prevention and therapy will require knowledge of the functional impact of the microbiome. Associations of the microbiome may be very informative for predictions based on disease risk and outcomes, et cetera. However, associations will not likely be suitable for therapy. They're not going to be interventional tools. We're really going to need to understand causation in order to yield meaningful intervention. And what I think is really exciting about this field at the time is that the tools for studying the microbiome are extraordinary, both at the level of experimental model systems as well as within humans. So I think the field is well-poised to be very enriched in this approach. So with that, I will conclude and move on with our next speakers.

microbiome

critical illness

host-pathogen interactions

correlation vs causation

genetic-level interventions