Thank you, Sophia, for the kind introduction. It's also... I have the unenviable task of following two great talks. Here are my disclosures. I have no competing financial interests. The objectives of my talk today are to really summarize host-pathogen interactions in critical illness, to highlight mechanisms of pathogen-induced immunosuppression, to provide an overview of this through the lens of existing critical illness phenotypes and endotypes as they pertain to sepsis, and then highlight current translational gaps in the field that impede the appropriate selection of immunomodulatory therapies. As many of you well know, the interaction between pathogen and host really dictates the immune response at the level of individual patients. You may have pathogen-specific factors, such as the type of pathogen, the virulence, and the pathogenic load. And on the other hand, you have the host-related factors, which is patient age, underlying genetics, comorbid conditions, and immunosuppressive medications that they may be on. So these factors play together to dictate the host response in any given patient. Pathogens are recognized by conserved receptors by the host immune system, and shown here are the various toll-like receptors, which can identify both intracellular pathogens as well as extracellular pathogens. And on the right here is the prototypical toll-like 4-receptor, which can act through MYD88-dependent pathways to trigger NF-kappa-B signaling and freely releasing your pro-inflammatory cytokines, or act in manners that are independent, resulting in interferon signaling. And the latter is really the driving mechanism of force in the host antiviral response. So upon recognition by the host, there's the initial triggering of the innate immune system, which is the first line of defense. So think of this as your blunt hammer that's really trying to curtail the pathogens. And the innate immune response then triggers the adaptive response. And this can here shown as T helper cells that are being differentiated into type 1 cells to fight intracellular bacteria, type 2 to fight against helminthic infections, type 17 to fight extracellular bacteria. So think of these as snipers that are dedicated to targeting specific pathogens. I think the one piece to remember here is that there's actually significant crosstalk between these two innate and adaptive arms, with the innate arm serving to trigger the adaptive arm, with the adaptive arm then feeding back to say, we don't want too much inflammation and trying to downregulate the blunt innate immune response. So it's certainly a two-way street. The slide is busy, but pathogens continuously seek to evade the host immune response. So they may develop cell surface changes that hide them from recognition. They may alter the pathogen receptor signaling. They may hijack mechanisms of intracellular gene expression pathways to really use them for the purposes of replicating themselves. And finally, they may trigger both intracellular and extracellular cascades of cell death with the idea of spewing out pathogen, replicating, and keeping the cycle ongoing. So where SARS-CoV-2 is particularly effective is, like SARS-CoV-1, it really internalizes itself through the human ACE2 receptor on epithelial cells. And this is abetted by serine proteases. But essentially, the short story here is that it's really effective at getting into cells before it's detected by the host immune response. Once it enters cells, it creates these replication organelles within which it continues to replicate and then develops nucleocaspid proteins that protect it from the antiviral innate immune response. So it's really great at evading the host immune response. So summarizing some of the mechanisms of pathogen-induced immunosuppression, we've alluded to various pathways of cell death. And this could be several, there are several ways, some of which are caspase-dependent, some of which are not. But really what this does is, as has been alluded to, is it really triggers damage-associated molecular patterns that are released from host cells, which propagate this immune response in a very dysfunctional manner. Second is critical illnesses associated with immune exhaustion, right? So shown here on the left is human leukocyte antigen expression by antigen-presenting cells. And an appropriate response would be T cell activation. But in critical illness, what's really happening is that, A, the antigen-presenting cells have low expression of HLA-DR. And then there's stimulation of negative pathways, including programmed cell death pathways, that results in T cell exhaustion. And finally, there's a metabolic component to this, is that acute inflammation results in a shift towards glycolysis. But prolonged, sustained inflammation has started to really downregulate all pathways of metabolic activity, resulting in immune exhaustion of cells. So taking a step back, how do these three major mechanisms sort of affect the various cell types across the innate and adaptive arms? As has been alluded to, there's really poor neutrophil function and predisposition to netosis. You have an increase in developing neutrophils and myeloid-derived suppressor cells, which themselves inhibit T cell proliferation. And then, as alluded to, monocytes are poor at presenting antigens appropriately. And the theme to take away is that there's, A, immunosuppressive cytokines, such as interleukin-10, that are being secreted, as well as an expansion of immunosuppressive cell types in critical illness. The shift in adaptive immune response is sort of similar. We've spoken about T cell, both lymphopenia and exhaustion, but really it's a shift from appropriate effective T cell function towards an enhanced regulatory T cell function. So the final theme to sort of think through is that, and this is especially true for viral infections, such as influenza and SARS-CoV-2, is that because they cause tissue trauma and they modify the host local inflammatory response, they're really a great at predisposing to secondary infections. Think of your methicillin-resistant staphylococcus aureus infections in patients with flu and then a variety of co-infections among patients, or secondary infections among patients with COVID. So this results in sort of, this is shown over and over again, but your classical, if the normal function of the immune system is to curtail pathogen and return to a state of homeostasis, and the classical paradigm of dysfunctional host response in critical illness is sort of either the sustained hyper-inflammation, or what's shown below is various mechanisms that drive a compensatory anti-inflammatory response, with the latter being thought to result in risk of nosocomial infections and chronic critical illness. So how is this different in COVID-19? And so mild cases of COVID-19, and it's thought that the triggering of the innate and adaptive arms happens relatively improperly, resulting in curtailing infection and resolution and return with relatively mild clinical symptoms. But I think it's really important to think through what's happening in severely critically ill COVID-19 patients, is that A, because the virus is able to evade the pathogen, you have much substantially higher numbers of viral load. You have a sustained activation of your innate immune system. And then you have really T-cell lymphopenia and T-cell exhaustion, which is never able to shut down the innate immune system. So the classical cytokine storm combined with a T-cell response, that is not appropriate. These coupled together really are thought to drive the poor outcomes in patients. So this is a slight deviation, but thinking of a majority of our critically ill patients with sepsis, biological heterogeneity among patients is such a major challenge to sort of say, who are the patients that will benefit from certain immunomodulatory therapies? So there's substantial progress been made within the last two decades to address this biological heterogeneity. And this is work by Carolyn Kalfi's group and Pratik Sinha, showing sort of these latent profile phenotypes of critical illness. What's shown in red is the hyper-inflammatory phenotype, and what's shown in blue is the hypo-inflammatory phenotype, and Dr. McCauley gave a great summary of all the progress that's been made in this realm. But really what this alludes to is an abnormal innate immune response, broadly speaking. So patients with a hyper-inflammatory phenotype have up-regulation of neutrophils and monocyte signatures relative to those with a hypo-inflammatory phenotype. So the broad reflection here is abnormal innate immune response. Jumping to sort of gene expression endotyping approaches, and this is work by Hector Wong and colleagues, really used a 100-gene panel to tease apart subclasses of pediatric septic shock patients with the genes really reflective of the glucocorticoid signaling and adaptive immune response. And endotype A is associated with relative repression of your adaptive immune response. And similar analogous results have been identified in adults, critically ill adults, and this is work by the Davenport groups. So sepsis response signature 1 and 2, and in adults, it's sepsis response signature 2 that has a repression of your adaptive immune response. So the broad, again, this is not a summation of all of the subclassification approaches, but here are just the kind of broad themes. So latent profile phenotypes reflect the innate immune system broadly, and then gene expression endotypes, at least a few of them which have been reproducible, have reflected adaptive immune response. So again, depending on which cells you focus on, even something which is relatively homogenous like COVID-19 can be teased apart into different immunophenotypes. This is work by Matthew et al., where they looked at sort of patterns of CD4, CD8 activation among COVID-19 patients and identified these three immunophenotypes with distinct outcomes. So where does this leave us? I think the real idea should be towards there's a shift towards integrated subclassification schemes to look at what is the balance between your innate and adaptive response. And this is work, really recent work, where they're looking at what's the overlap between patients with a hyper-inflammatory phenotype and the different gene expression endotypes. And the consistent theme that emerges is that patients with activation of the innate arm and then repression of the adaptive arm of the immune system tend to have the worst clinical outcomes. Again, this is an illustration showing that the patient on top really has an immunocompromised state, has overactivation of the innate immune response, and repression of the adaptive response. But then lies a spectrum of patients with disease states that go from one end of the spectrum to the other. I will end by sort of highlighting translational gaps and potential therapeutic opportunities that exist in the field. I think that choosing the right tools to profile the immune response among critically ill patients remains a challenge. This could vary from something as simple as neutrophil to lymphocyte ratios, things like flow cytometry to look at HLA-DR expression, profiling your myeloid-derived suppressor cell numbers. Things such as ex vivo stimulation of leukocytes using DNF-alpha or LE-spot assays. But really there's no one consensus way of profiling the immune system. Second, I think thinking through what is the interaction between host underlying genetics, pathogen, and microbiome is an area that's ripe for discovery. This is especially true in the realm of next-generation metagenomic sequencing, wherein we're able to really more comprehensively capture what pathogens, especially the more rare ones, that are not necessarily picked up in your blood cultures. A third aspect is really understanding compartmentalization of the immune response. We always presume that what's true in blood is reflective of what's true in the individual organ systems, but that assumption may or might not be true. So if you have a patient who is immunosuppressed in their blood, and then has an immunosuppressive phenotype in their organ, and there's concordance there, giving them immunostimulants may make sense. But let's say you have a patient who is immunosuppressed systemically, but is hyper-inflammed at the tissue level, it's unclear what the benefit of immunomodulation in those patients would be. So I think more to come. And finally, we're really at a very early stage of working towards precision medicine-based immunotherapy in critically ill patients. There needs to be consensus on the way we classify patients, the way we interpret and how we are informed of their fluxes in their immune response between the various arms. How do you bring this back to the bedside? How do you make point-of-care assays? And then to be able to embed these within the context of clinical trials. And finally, how do you bring this to not just the ivory towers of critical care, but are able to do this at scale? With that, I will end, and we should be open for questions. Thank you. Thank you very much.

host-pathogen interactions

sepsis

immune responses

precision medicine

immunotherapy