Okay, welcome everyone, and thanks to the organizers for inviting me to speak today. So here are my disclosures, three disclosures at the top not relevant to this talk, and then two disclosures at the bottom that are relevant. The first, I'm the PI of a study that I'm going to be briefly talking about called the SPARC study, and co-chaired another clinical trial called the ASTER study, which I'll briefly mention. So here was my charge for the talk today, the speaker will explore use of various immune restorative therapies and anti-inflammatory therapies, including evaluating the current level of knowledge for each's use. So I took a bit of speaker liberty here to craft this in a way I thought could get across and hopefully yield some important information in about 12 minutes. So this is how my talk is going to be broken down. I'm going to briefly talk about the past, the abridged version of drug trials and sepsis, then move on to current concepts of sepsis, and mention a couple of thoughts that are my own about the challenges with the definition as they apply to clinical trials. And then I chose to focus on molecular targets and pathways that are further along in clinical trials, either in sepsis or in other conditions that might be relevant to sepsis. And you can see these four pathways or molecules below, interferon gamma, IL-7, the TY-2 angiopoietin pathway, and sulfury hemoglobin. And then lastly, end with a positive note that I really do think that the very near future really looks promising for sepsis therapeutic trials. So I suspect over the past decade or 15 years at SCCM, there have been a lot of talks about why there have been a number of so many failed clinical trials in sepsis. This is a really nice summary by Dr. Kavilian and Singer, and I think this quote summarizes it nicely. We analyze the reasons for this failure, including focusing on including appropriate patients and the use of irrelevant animal models. So there's a lot of people in this room interested in sepsis, and whether you're a health outcomes researcher or a bench researcher working with animal models, I think it's our charge to really think about this carefully as we study sepsis in every one of our studies. So this is the current sepsis-3 definition, life-threatening organ dysfunction caused by dysregulated host response to infection, which is a broad definition. And there are a number of review papers in the literature trying to capture maybe some of the complex underlying biology of this condition. This is a nice schema from Dr. Pietro and Singer as well, which I think drills down a bit on what we've learned, that infection can induce the release of damage-associated molecular patterns, or tissue damage can lead to damage-associated molecular patterns that cause this robust inflammatory or immune response, now recognized that there are multiple facets of this that we can describe, metabolic, neural, hormonal, bioenergetic problems, and these can coexist with this immunosuppressive state. A condition is associated with injury to a variety of cell types, endothelial, epithelial, and a coagulation cascade. And in the host, there's a stress response that can either become adaptive with a metabolic favorable response or a failed response, leading to organ injury and failure or recovery. So what about the sepsis-3 definition in relation to clinical trials and novel targets? Well, there's some good things about the current definition. It can be defined from clinical and hemodynamic and readily available laboratory data. It's inclusive. It identifies a group of patients at high risk of mortality, and it's easy to identify these cases if you're looking at clinical retrospective data sets. The bad part, if you're thinking about molecular targets, is that it's syndromic. It's not biological-based. It ignores the kinetics of injury and resolution, which, as we know, may change very rapidly. And it also ignores the heterogeneity and interrelatedness of pathogen and host factors. So here are just a few areas that we could go down the pathway and start talking about potential targets here, and I tried to split these out into different groups here. Inflammation or immunosuppression, endothelial pathways, cell-based therapies, epithelial targets, neural or metabolic. I chose to focus on the first two categories, and I'm going to talk about interferon gamma IL-7, ANG2 and TY2, and cell-free hemoglobin, just for time's sake here. So interferon gamma has been of interest to the sepsis community really since the late 90s, and sepsis deaths, patients that die of sepsis, if you look at their cellular profiles by flow cytometry, you can see downregulation of monocyte, major histocompatibility to complex receptors. And administration to interferon gamma, which is produced from T cells, can restore this monocyte function and can stimulate responses and microbial clearance. On the right side, you'll see a nice schema of a natural killer cell releasing TNF, sorry, interferon gamma, and leading to maturation of monocytes. This is a bit in controversy, though. There's some data suggesting the opposite of this is true, that interferon gamma elevation is actually more associated and potentially causative of secondary infections, and that blockade of interferon gamma signaling actually improves survival. So this is an interesting area where there's debates on both sides about what might happen here. Searching clinicaltrials.gov, there's two ongoing clinical trials that are small, both phase two trials, looking at recombinant human interfering gamma in patients with sepsis and infection. The first one is an international open-label study enrolling 200 patients, randomizing patients with candidemia to active drug or usual care. Primary outcome here is time-to-first negative culture. And then the second study that I have listed here, which has a great trial name called IGNORANT, is a phase three placebo-controlled multicenter European trial looking at administering interferon gamma after ventilator-associated pneumonia is diagnosed, and the primary outcome here is ventilator-free days after the patients are enrolled in the trial. Moving on to our next target, IL-7, which a lot of folks know about this molecule. IL-7 is a factor that is required for lymphocyte growth and survival and differentiation. You can see on the right-hand side, IL-7 interacting with receptor on lymphocytes and through downstream JAK-STAT signaling induces a cascade that both inhibits lymphocyte apoptosis and also promotes lymphocyte proliferation. And there is a recombinant form of human IL-7 that is available. There have been two very small clinical trials looking at IL-7 in sepsis patients. By Dr. Hotchkiss and colleagues led these trials. So the IRIS-7 trial looked at patients with septic shock and organ failure with persistently elevated SOFA and lymphopenia. This was a phase 2b placebo-controlled trial. And recombinant human IL-7 was given intramuscularly once or twice weekly for four weeks. There were 27 patients enrolled. The primary outcome was looking at lymphocyte counts. The drug was well-tolerated and lymphocyte counts increased in the study. There was a follow-up study to this that was just published this year in Annals of Intensive Care, which took the same compound and administered the drug intravenously. It was a randomized placebo-controlled trial. This study, unfortunately, was terminated early due to safety events, enrolled 21 patients out of a planned 40. The drug did increase lymphocyte counts, which was the primary outcome. But there were three patients that had adverse reaction to the drug infusion with fever and respiratory failure. And the DSMB ended this study early. In a pre-planned analysis of drug levels, the study authors found extremely high drug levels here of this compound. And so I'm not entirely sure what the current state of development of this drug is, but it may be that the IV infusion is not appropriate. The next pathway I wanted to mention was the endothelium. I think the endothelium is just such a critical target in sepsis. And I wanted to focus on this angiopoietin type 2 pathway. So there have been a number of observational studies evaluating this protein angiopoietin 2. And elevated ANG2 levels correlate with disease severity and worsened outcomes both in sepsis and ARDS. This is a schema looking at these cell-cell junctions of endothelial cells here and the type 2 and ANG1 under normal conditions create a tight barrier here. Under an inflammatory state, ANG2 is released into this space and impairs this binding leading to leakiness of these endothelial layer here. VEGF, which is also involved in this pathway, has been a really targetable receptor and has been beneficial in ocular disorders. And there's been development now of augmenting this pathway for a number of ocular disorders. So this pathway can be manipulated through a variety of approaches, one by taking advantage of the agonist approach of ANG1 here or using these antibodies that may block the activity of ANG2, or lastly, targeting this co-receptor that's involved with this interaction. To my knowledge, there's no trials of these compounds in sepsis underway, but I think the biology is interesting and important. The last pathway I wanted to mention is cell-free hemoglobin. So we know from a number of observational studies that intravascular hemolysis occurs frequently in sepsis and cell-free hemoglobin is elevated and it directly causes endothelial injury through a variety of mechanisms. So here's the ruptured erythrocyte and hemoglobin at the endothelial space here can cause damage by nitric oxide scavenging, oxidative stress, and endothelial barrier dysfunction. This facet of cell-free hemoglobin being potent in the sepsis state is also compounded by oxidation of iron on the hemoglobin moiety here. So there's a number of ways pharmacologically you could think about targeting this. So you could try to prevent red blood cell hemolysis through nitric oxide donors. You could use hemoglobin scavenging approaches with haptoglobin or hemopexin. And there have been a number of preclinical trials in mice and canine pneumonia that hemoglobin scavengers improve outcomes. And then lastly, acetaminophen actually turns out to be a really potent iron reductant and has been proposed as a potential therapy here. Under that premise, the NHLBI PETAL network actually just finished a phase 2b trial entitled Acetaminophen for Sepsis, Targeted Therapy to Enhance Recovery. The trial enrolled 447 patients with septic shock or sepsis and lung injury to IV acetaminophen or placebo for five days. Primary outcome is days alive and free of organ support. And there's a pre-specified subgroup analysis looking at stratification of the outcomes by cell-free hemoglobin level. So this study enrollment is completed this past spring and results are pending and data cleaning is underway. Results should be out very soon. So where are we going in sepsis? I wanted to point to one or two really fascinating studies here. This one by Chris Seymour at UPMC, a retrospective study including 60,000 patients with sepsis using machine learning approaches that were agnostic of what was going on. These patients could be clustered into four groups which were labeled alpha, beta, gamma, and delta here. It turns out that site of infection, type of infection, Apache, etc. didn't classify patients in the same way as this machine learning approach. And these groups, when in a post hoc analysis looked at sort of classical inflammatory cytokines, they had differing biological signatures. There are other studies underway, one of which that we're involved with. This grew out of the ICE by COVID trial and is an observational study using rapid point of care biomarker assays to classify patients with acute hypoxic respiratory failure or ARDS rapidly at the point of enrollment into subphenotypes using the methods that have been defined by Carolyn Kalfi and her colleagues. And we think this might lead to a really nice platform where we could use biomarker based randomization in the very near future. These are just some data from this trial. This is our enrollment in this observational study just a couple of days ago. We've enrolled 171 of a target of 250. We can phenotype these patients 1.8 hours from the time of signing informed consent. And our cohort includes patients both with unilateral infiltrates with acute hypoxic respiratory failure and bilateral infiltrates. Interestingly, we're finding hyperinflammatory phenotype in both the unilateral infiltrate and the bilateral infiltrate patients. And then lastly here, I just wanted to mention the APS consortium funded by NIGMS and NHLBI, consortium-wide study of ICU patients with acute cardiovascular and respiratory failure. There are six clinical specific projects that are going to enroll 4,000 to 5,000 patients and collect biospecimens and really rigorously phenotype these patients. So this is really going to provide a wealth of data for the sepsis field to hopefully understand the syndrome better. So in summary, preclinical studies and rigorous clinical cohorts have identified key targets in sepsis that may be modifiable. I've talked about a couple here. I personally feel that the animal models will remain relevant. We need to be careful with them and understand what we're studying. It's critical that we incorporate human specimens and we characterize these properly. Newer data techniques, such as AI technologies, may yield new subgroups that we're not aware of at the bedside. And these rapid point-of-care biomarker assays are really coming to clinical use, I think, very soon and will help us define the biology and target it at the point of care. The APS study, I think, is going to help us. And then lastly, I wanted to just make one point that the COVID experience, where we really use novel trial designs, platform, and adaptive clinical trials, should really be embraced to expedite efficiency and discovery. I think we're really entering a new phase here. So with that, thank you very much.

sepsis therapies

immune restoration

molecular targets

machine learning

novel trial designs