Well, thank you so much, Michael, and thanks for the opportunity of speaking here this morning. Oh, okay. That wasn't it. I think you go down on this. Ah, okay. Push the down button. Gotcha. Okay. Thanks very much. I'm obviously technologically challenged. So these are my disclosures, probably relevant to this talk as I've been involved with the Euphrates trial, the Pauli-Mixon B column, although you'll kind of get my perspective on this as the talk evolves. I've also had some other academic and commercial activities that I should disclose. And here are the learning objectives for this talk. I want to basically review the concept of extracorporeal therapy. Maybe at the start, I would like to disclose my discomfort with the concept of blood purification. Both because this is not necessarily what we're trying to do. I'm not sure there is any such thing as pure blood. And perhaps more importantly, because when phrases like that come up, they sort of evoke in my mind images of eugenics and a number of things like that that are really quite unpalatable. I'm going to look at the spectrum of potential approaches and then finally look at the evidence for biologic efficacy and clinical benefit. So essentially the rationale here is that in many disease states, including sepsis, things are released into circulation. They could be bacteria, they could be bacterial products, they could be host-derived products that have been triggered by the infection itself. And that that may launch that soup of circulating factors contributes to the pathogenesis of sepsis. So if that hypothesis is supportable, that it's an excess of things in the bloodstream, then it would follow that either diluting them or removing them through extracorporeal means should improve clinical outcomes. And there are a number of different strategies, similar in concept but differing in technique, that have been developed to accomplish this role. Plasma exchange or plasmapheresis is probably the oldest of these. It's been used in other disease processes, such as Guillain-Barre syndrome. Pathogen removal is a distinctive concept, it's not really very much been written about that. Endotoxin absorption and cytokine removal are the areas where I think the interest is in critical care today. So plasma exchange, the principle is shown here. In a patient with a disease, you insert a line that allows blood to circulate through a pump, and in plasma exchange what happens is that pump then goes to a separator that divides the blood into the cellular components, which are returned to the body, the plasma components, which are removed, discarded, and replaced with some other solution. This may be albumin or a colloid, or it may be fresh frozen plasma. These are then returned to the body, and you can see that what happens is that you would dilute the patient's own plasma. Does it work? Well, in sepsis, the evidence is not really impressive. This was a study looking at the impact of plasma exchange on differing mediator levels, and you can see that one of the potential attractive features of this is you not only dilute or remove things that we classically think of being toxic, tumor necrosis factor, endotoxin, for example, but you may actually restore or prevent the deletion of things such as antithrombin or protein C, which are thought to be more beneficial. But I think if you look at the graphs here, you'll appreciate that even as a biochemical measure, this is a pretty modest clinical effect. And from the perspective of clinical effects, there's really not a lot of data here. This was a systematic review done by Emily Rimmer looking at a number, and I think the number is four, it's a very small number, of randomized trials. Fewer than 100 patients altogether show no benefit for plasma exchange and plasmapheresis in pediatrics, possibly benefit in adults, but the numbers are very, very small here, and I think the trials are subject to bias. And so I think it's pretty hard to draw any hard and fast conclusions on the basis of fewer than 100 patients in clinical trials. There is, interestingly, this study that came out during the COVID-19 pandemic, which actually suggested the possibility of a role for plasma exchange in patients with COVID-19. No effect on survival per se, but there was a statistically significant shorter duration of mechanical ventilation and a shorter ICU stay. Now bacterial killing is an interesting concept. There are a couple of small experimental studies that have looked at the idea that perhaps you could use these techniques to remove bacteria from the blood. This is a variant on that theme. These investigators developed a technique to use ozone, to pass the blood across an ozone device outside the body in hopes that it would reduce numbers of circulating bacteria. And what they showed in a swine model of infusion of live E. coli was that you did reduce the numbers of live E. coli, and you did have some effect on physiologic parameters. The big proviso here, beyond this being an animal study, is that the insult here is infused bacteria, and that is not a good model for the presence of microorganisms in the body that occur in clinical infection and sepsis. There's also been a similar study looking at trying to activate white blood cells outside the body. Again, some small biologic effects, but nothing very dramatic. So the adsorption systems are really the mainstay of treatment here. They involve using a filter with an adsorbent surface that selectively or non-selectively removes components of blood that passes through that filter. Non-selective ones use some kind of a polymer. This is, I think, the cytotex system that uses a polymer that's densely packed as beads into the filter. These remove hydrophobic molecules, which include a number of bacterial products, but also include certain host substances, such as bilirubin, myoglobin, and also some drugs, and it's actually been used in the context of removing drugs. Tegrelacor, I think, was the system that was studied. Now again, and this is going to be a recurring theme. The evidence for clinical efficacy here is very, very limited or non-existent. A number of very small trials showing essentially no effect. You'll see in the probably two most common, the ALTECO and the Cytoserb models, small numbers of patients in clinical trials and no clear signal. The one at the bottom is a Chinese system, and I'm not aware of any studies done outside of China. In contrast, there's many, many cohort studies, case studies, and that kind of stuff that are laudatory about this, but I think the absence of compelling data from clinical trials really should cause one to look at this more soberingly. Cytokine absorption has been looked at in COVID-19. Here was a paper published a couple of years ago in Lancet Respiratory Medicine, and you can see a striking effect here on survival with cytokine absorption. The problem, however, is not only that this was a very small study, but that the benefit is seen in the patients who didn't receive cytokine absorption. 76% of them survived versus 18%. This is what, 17 patients per group? I would not want to in any way claim that this is convincing evidence of harm, but it certainly underlines the fact that there's no compelling evidence of benefit for cytokine absorption in COVID-19. Finally, endotoxin absorption is a specific variant on this particular theme. It uses a column that's coated with polymyxin-coated beads, because polymyxin has a very high affinity for circulating bacterial endotoxin, and it will bind that endotoxin and selectively remove it from the circulation. So this is the polymyxin column that is the most widely used endotoxin-absorbing system. This has been studied, again, in a number of small trials and a few larger trials. There is, in some of the smaller trials, a suggestion of benefit, and perhaps one could say that if you look at it in select populations, as Dina Cruz did, looking in patients with intradermal infection, and she showed benefit. Perhaps that says that there is a signal here for a really promising therapy. In those trials that are at low risk of bias, that signal has not been seen in any definitive kind of way. So I was involved with the Euphrates trial, which looked at this column in patients with septic shock. We used a slightly more sophisticated design in that we focused on patients who had high circulating endotoxin levels using this particular assay, the endotoxin activity assay, which is based on an antibody that binds endotoxin that then primes the patient's own neutrophils so that they produce an augmented response when they're exposed to Zymazan. And so you have a three-tube assay, a negative control, a maximally stimulated control, and the patient's sample, and you can integrate the light generation over time correlates with the amount of endotoxin in the blood. So we're targeting endotoxin adsorption in patients who actually are endotoxemic. Here again are the results. This was published in JAMA. There was no overall benefit in the entire population. We were advised by the DSMB to focus on those with the most significant degree of organ dysfunction. That was patients with a MOD score of greater than nine. There was a 3% mortality difference, which favored the endotoxin absorption column. But again, it was nowhere near statistically significant. Possible reasons, small study. It looks like the total amount of endotoxin completely overwhelmed the capacity of the column, and maybe the dose was wrong. Maybe we should have been focusing on patients with lower endotoxin levels because of that. But there is an ongoing trial called the Tiger's Trial, which is trying to address some of these potential reasons why that study may have been disappointing. So I think, to me, the picture with extracorporeal removal of potential toxins is discouraging. First of all, when you do it nonspecifically, the removal of these things is indiscriminate. You may take out some injurious host proteins or bacterial proteins, but you may also remove some things that are actually beneficial to this integrated host response to a life-threatening infection. Generally speaking, with the exception of Euphrates, the treatment has not been guided by a diagnostic test, and the trials have tended to be small, low quality, and guided primarily by enthusiasm more than sober science. I want to just close with the concept here that we do have to be particularly careful in patients with, say, sepsis. These are data from the ACCESS trial. It was not an extracorporeal device, but a synthetic lipopolysaccharide molecule that bound toll-like receptor 4, but didn't activate it. So it's a specific endotoxin antagonist. Overall, in the study, there was no benefit with treatment with the antagonist. But you'll see here in patients who had gram-positive infections that there was actually a statistically significant increase in mortality in those patients who were treated with this substance or with this particular inhibitor. So it's not at all beyond imagination that we have evolved in conjunction with endotoxin to use it as part of our means of ramping up the immune system to deal with infectious challenges, and that, particularly in gram-positive infections, neutralizing endotoxin may be detrimental. Now, so I think it actually speaks to the future of extracorporeal remover. To me, it's probably not in patients with infection. They're the hardest patients to study because their response may be helping them in some way, as well as causing them harm. On the other hand, patients with multiple trauma, they don't need a response to infection. They also, some of them with severe multiple trauma, have elevated circulating levels of endotoxin after their injury. These are data showing a cohort of patients admitted with multiple trauma to the ICU. Particularly following abdominal trauma and particularly following surgery, by the third to fifth day, you see circulating endotoxin levels, and those patients who are endotoxemic have a worse outcome. So perhaps we should be focusing on endotoxemia, for example, in non-infectious contexts. Trauma, cirrhosis, congestive heart failure, renal failure, there's a large population of patients that potentially could be evaluated. So I wanted just to conclude by saying that my take on extracorporeal treatments right now is that any evidence of benefit is lacking or, at the very, very best, modest. Trials have also suggested potential for harm, and this has not been incorporated adequately into our thinking about this. And so that at the present time, I see the only role for these things in patients with sepsis is within the context of a randomized clinical trial. Many thanks for your attention this morning.

extracorporeal therapy

sepsis

plasma exchange

pathogen removal

cytokine absorption