Thank you for the kind introduction. It's lovely to join everyone today. Thank you for your interest in this session. So I'm going to try to, I think, expand on some of the themes that were raised by my colleagues, and I think hopefully be a little bit more provocative about the real nature of the host-pathogen interaction, and try to at least form a basis for thought and consideration about what do we actually understand about the raw nature of that interaction? What do we understand about the pathogen and its interface with the host? So I think a lot of the field has focused on the host response as a means to think about targeting sepsis and critical illness and infection in that context. But I would argue that probably dialing back and really taking a deeper dive on the host-pathogen interaction is really going to be critical. So my charge is to focus on the host-pathogen interaction, and I'm going to frame this as a conceptual framework for the future, because as I'll mention in a few minutes, I don't think we actually have much in our armamentarium right now that does this. I have a few disclosures that are noted here. Neither of these will be specifically relevant to anything that I'll talk about today. So when I think about the host-pathogen interaction, the simplistic diagram, I think, to me is very informative, because it highlights some of the extraordinary complexity in how this balance is struck. So on the one hand, if you think about the pathogen or the microbe, most of the pathogens that we encounter in our ICU are actually bugs that every single one of us has in our bodies right now. Pathogens are not creeping in through the ICU door. They're coming in with the patient inside their body. And this is a real challenge, because it highlights that we need to think about and understand microbial niche establishment. We need to understand the factors that govern, what bugs do I have in my colon relative to the bugs that you have in your colon, and does that predispose us differentially, even in the setting of the same operative procedure, to a different outcome? What are the attributes of pathogen acquisition and transmission? Do we contribute to pathogen transmission or pathogenic traits in our ICUs by the things that we do every day? How do different organisms regulate their virulence factors? This is highly regulated, even between two strains that are distinct within the same organism. Two different E. coli will interact with their environment in a completely different fashion, sometimes based on two or three gene regulatory elements. We're not thinking about that problem as we're considering what is E. coli doing in the context of our patients. And then when you look at the host, we all know now that the host is full of a microbiome, which we have a very limited understanding of in terms of dynamic behavior. Each one of us has a unique immunologic competence based on our own genome and our encoding regulatory elements that govern our immune responses, which I think we're not really tapping into. And this governs virulence factor susceptibility. I may be susceptible to an E. coli that one of you may not be. And then I think we haven't, again, addressed this issue of how does this map out onto specific tissues within the host? So when you think about the sort of pathogen and what it brings to the table, this is just a diagrammatic illustration that can, you know, it's a particular organism, but this can be said for any microbe. The virulence properties of the microbe are multiple. Most bacteria have at least 100 different factors that can directly injure our bodies and our cells. This complexity alone, even if you just start there, is a two-log problem, which is a real challenge. So on this backdrop, I'll ask the question, what's in our current toolbox? What have we done so far that actually really targets the host-pathogen interaction? Vaccines, antimicrobials, and infection prevention. On the one hand, when you're in a session talking about novel approaches, these seem, frankly, quite uninteresting. But I'd like to argue these are really incredible as a foundation to think about how to move the needle from a public health perspective, even in our ICUs. So I'll ask the question, what can we learn from history? I really love these graphs. These were pulled together by a broad consortium that was funded by Gavi and the Gates Foundation to address, what have vaccines done for us? How have vaccines changed the outcome of some of the diseases that we actually really care about? If you look on the left hand of the y-axis, there's deaths averted. If you look at the right hand of the y-axis, it's per fully vaccinated. And if you look across all of these vaccine-preventable diseases, what you appreciate is millions and millions of lives are saved by prevention. And what's interesting about this graph, you all can recognize the pathogens that are on here. Many of these are viral pathogens. Several of them are bacterial pathogens. Most of the organisms that kill patients in the ICU right now are not vaccine-preventable. And they are organisms that live within our own microbiome. This to me represents the single largest opportunity to generate a meaningful population-level impact in terms of how we think about modifying outcomes of sepsis is actually preventing. So if you look at these numbers, 50 million deaths averted just from this relatively small array of vaccines. This is all obviously pre-COVID data garnered from 2000 to 2019. So this raises the question, which I think forms a nice basis for consideration, why do vaccines actually work? What they do, which we all appreciate, is that they change immunologic competence of the host. And they garner an advantage for the host in targeting the host-pathogen interaction. So immunologic competence on the part of the host dependent on the vaccine will modify the microbial niche. It will modify competition within that niche. Certainly modifies the ability of a microbe to become pathogenic and to rapidly escalate its pathogenic properties by neutralization of those pathogenic attributes. So this, I think, to me, forms a nice foundation for thinking about the real essence of targeting the host microbe or the host-pathogen interaction. So I really enjoy this conceptual framework that I'll share with you. And in a minute, I'll sort of share the permutations of this model. So this is a model that was put forth by Arturo Casadoval in 2003. And what it really gets at is, I think, a lot of what we talk about with the nature of this host response. And I really like this conceptual framework. Because if you notice on the sort of y-axis or the left side, you'll see host benefit, which is really not what we're talking about in the context of life-threatening infection. If you look at the top, this shows host damage. And if you look at the bottom, the host response being weak or strong. And obviously, we're talking about much more granular level of sort of stratification than weak versus strong. But I like this conceptual model. Because you can see that the damage response can be linked to the host response. So this is the initial conceptual framework. And then in its embodiment, what he argues is that different pathogens interact with the host in different ways that are classified as these various classes. And I'll just pick out two of these as exemplars. So in this context, class one, a weak host response may be associated with significant damage to the host. However, a so-called strong host response may be associated with benefit or protection of the host. This is a very different model or outcome of the host-microbe interaction, for instance, in class five, where a weak response may be associated with the host doing OK. And a strong host response is actually associated with damage. And I really like this conceptual framework. Because it allows us to appreciate, again, in a pretty simplistic way, why response differences occur between individuals and based on the pathogen of interest. And he extends these, even in 2003, to argue that there are pathogens we have not encountered yet that may have very distinct phenotypes. So in this context, irrespective of the host response, the pathogen's ability to elicit damage is extraordinary. In this context, just the right host response may actually be beneficial, whereas too weak or too strong could only be harmful. So I think that this makes me ask the question, do we understand enough about the host-pathogen interaction to actually identify strategic targets? And I'll extend that a little further to say, to deploy strategic targets on an individual basis. So this then frames the sort of broader thought process of, why is it that every hour matters? So if you look at this framework of the pre-disease state, what we're really looking at in our ICUs is these early manifestations of infection. And we initiate therapy once we're able to detect those manifestations of infection. And quite often, especially when we talk about life-threatening septic insults, the patient is exhibiting severe disease manifestations sometimes at the time we're starting therapy. So this matters because pathogen multiplication is extraordinary. It's a two-log difference in the number of bacteria that can be cultured in a matter of two hours. And I think this really raises in my mind the level of challenge we have when we aren't thinking about the nature of the pathogen that incites the insult. The pathogens can elicit an extraordinary amount of injury before we even recognize that the patient is infected. So what this sort of reactive response leads to is a lack of predictability about our outcomes, sometimes poor outcomes. And I think given that our best armaments for managing these infections is empiric antimicrobial therapy as potentiation of resistance. So can we better leverage this pre-infection or early infection window of time? And this gets at a lot of the concepts that my colleagues very nicely raised, which is, how do we take data that would actually inform us of risk? And I would argue that the scale to which we need to look at this data really needs to be able to manage clinical information about the patient, genomic information about the patient, the microbiome, and an understanding of molecular disease mechanisms. Clearly this will require predictive analytic modeling. This will require the ability to take large data and use machine learning to be able to leverage all of this data across the host and the pathogen in order to define groups to study specific targets. So ideally, as we've talked about, stratification of the host based on the nature of the host and the pathogen interaction I think is the future. Certainly this sort of subtyping or the ability to define what segregates one individual from another will be key ultimately to targeting interventions when we have the ability to define what those interventions should be, which I think really is where the future will lie. And importantly, we'll have to be able to understand risk as a function of time, which is a point that Vanessa made. Looking at a static point in time is not going to reflect appropriately the very dynamic nature of the host-microbe interaction, which may be different 60 minutes from now, not even just five days from now. So how do we integrate this knowledge over time to recognize that risk for one patient may have a curve or a sort of a picture that looks like this, whereas for another patient that risk may be confined to a certain time window and very different from the patient in the next room. So I think these multiple variables integrated over time to me is the real hurdle to being able to define how would we think about molecularly targeted strategies when we have those strategies available. So what are the scope of potential targeted interventions that we may see in the next few decades? I think it's possible that there may be approaches to genetically driven resistance modification among microbes. There may be the ability to discern what's a favorable microbiome in an individual and what's unfavorable and use genetically targeted probiotic therapies, which is a very, very far cry from where we're at right now in the field. Understanding likelihood for infection by a specific microorganism may enable real-time immunotherapy or vaccination targeting to that pathogen. Identifying the at-risk host immune response, as Vanessa had mentioned, and being able to utilize specific designer biologics to target those responses. Understanding genetic susceptibility to a given pathogen. What discriminates me from you if we're both in an ICU in terms of our susceptibility? And then appreciation of the time frame in which the highest risk for a specific individual occurs, which may not be early in the ICU course. It may be in the middle of the course. It may be as they're being discharged. So what I'll propose is what I see as sort of how can we garner the most traction in this field when we're faced with a number of clinical trials that have sort of left us scratching our heads as to what's the next direction forward. I would argue that it's incumbent upon us to really think about prevention. Support of vaccination and the development of vaccines against organisms, such as our escape pathogens and those organisms that kill patients in our ICUs every day must be an important focus of our efforts. This on a population level has a lot more potential, I think, than highly targeted therapies to be met at the bedside of a single patient through personalized medicine approaches. As many of you know, I think vaccination was raised as quite interesting during COVID. But when you look at vaccine development for many of the pathogens that kill our patients, there's not really a pipeline. Risk mitigation. I think this is something that we all think about every day. How do we limit antibiotic exposure? How do we not potentiate antimicrobial resistance, which changes every 20 minutes for some organisms? How do we appreciate the benefits of the microbiome? Are we wise to be wiping our patients down every day with chlorhexidine? I'm not sure. What that argues is that getting rid of the pathogens is predominant over modification of the endogenous and potentially beneficial microbiome. I think we have to ask these questions a little differently. Understanding risk is a function of medical care. How can we modulate our approach to procedures to change risk? How do we modify or think about transmission dynamics within our ICUs? What about feeding? Transmission of the microbiome every single day occurs in the gut that's fed or not fed. How does this matter for our patients in this broader context of modulating the microbiome? And then as we think about the future and think about our ability to leverage enhanced technologies as they come forth, biologics and designer approaches, how do we really sort of generate the infrastructure to understand the patient as an individual and utilize sort of high-level machine learning and analytics to really capture data and enable us to generate both personalized and precision medicine, to me, I think is probably the big question that we all have to address in the coming years. So with that, I will conclude, and I think we'll probably start the discussion session. Thank you.

host-pathogen interactions

sepsis

ICU

vaccine development

personalized medicine