Hello, my name is Jay Patel, and today I'm going to be discussing feeding the unstable ICU patient with an emphasis on paralysis, circulatory shock, prone positioning, and extracorporeal membrane oxygenation, or ECMO. Thank you to the Society of Critical Care Medicine and to the Program Planning Committee for the invitation to speak to you today. I have no disclosures related to the content of this talk. So let's consider a case. Let's say you have a 45-year-old individual with severe ARDS, secondary to SARS-CoV-2 pneumonia, who is intubated and sedated, and you are assuming care for this individual. So ask yourself, what is it that nutrition has the capacity to modify in this particular patient? Here are four conditions or interventions that may make you think twice, though, about your nutrition plan. And of course, nutrition therapy, like other therapy, should proceed after safety and tolerance are established. And so for the sake of this talk, I will describe them as mutually exclusive. But in reality, as many of you know, they are not and occur in combination. But we'll focus on the patients who are receiving a paralytic agent, again, those in circulatory shock, those who are prone positioned, and those who are undergoing ECMO. And we'll use our case as an anchor to answer each of these questions. So here's the conundrum. We have an unstable patient in the acute phase of critical illness, and let's say that's the first week of critical illness, where inflammation and catabolism are the rules. So on one hand, if we do not introduce any luminal nutrients, we risk loss of epithelial barrier function and gut dysbiosis, which serve to perpetuate gut dysfunction and then drive multiple organ dysfunction syndrome. These are observed during the early acute phase of critical illness. And downstream effects include escalating energy debt, acquired malnutrition, and these are probably more so observed after that early acute phase, but these certainly have lasting impacts on survivors of critical illness who have physical impairments. On the other hand, if we feed these so-called unstable patients, we risk GI intolerance, which can manifest in many ways, but concerns include vomiting with pneumonia. And the most grave consequences are non-occlusive mesenteric ischemia and non-occlusive bowel necrosis. I won't read these out loud, but I've listed how each of these unstable or perceived unstable settings can link to some of these complications. So let's go back to our case. Now the patient is paralyzed to enhance ventilator synchrony. So why is paralysis implicated in gut dysfunction? So if you recall, non-depolarizing neuromuscular blockers do not have a paralyzing effect on intestinal smooth muscle cells. And in fact, those intestinal smooth muscle cells are innervated by the autonomic fibers in the myenteric and submucosal plexuses. And the rationale for enteral nutrition intolerance in these patients is probably due to the underlying illness and concomitant use of narcotics. In fact, observational data suggests that greater severity of illness is associated with greater enteral nutrition intolerance. There's a paucity of data evaluating enteral nutrition in paralyzed patients. What I'm showing you here is contemporary literature, contemporary randomized control trials testing cis atricurium in patients with ARDS. These are the ACQUIRISIS and PETAL NETWORK ROSE studies that compare cis at to placebo in ARDS. But note, they did not report nutrition-related outcomes. In ROSE, one patient in each arm had ILEUS. Otherwise, no other major GI complications were reported. In 2003, Tammy and colleagues conducted a paracetamol gastric emptying study. In two fasting states, in sedated patients with and without neuromuscular blockade. Now, paracetamol is a semi-quantitative method for measuring gastric emptying and intestinal absorption. And the area under this paracetamol concentration curve is determined by the rate of gastric emptying and by the absorption capacity of the small intestine. And as you can see, the rate of emptying and absorption was similar in patients with neuromuscular blockade. Those are the open circles and without neuromuscular blockade, and those are the closed squares. And so gastric emptying and absorption was no different in sedated patients receiving paralytic and no paralytic. And so we can at least conclude that there doesn't seem to be a physiologic rationale for why paralyzed patients cannot be enterally fed. And observational data suggest it may be safe, but certainly efficacy data are lacking. So let's move on to circulatory shock. Recall our patient who was intubated, sedated, paralyzed, and now this individual is in septic shock and on a norepinephrine infusion. And so what is it about circulatory shock, in this case septic shock, that makes one think differently about the nutrition? If we look at a meta-analysis of randomized controlled trials comparing early to delayed or no EN on clinical outcomes, there's a clear benefit with early enteral nutrition. This is a force plot for mortality from 19 trials evaluating early to delayed EN. But if we zoom in a little bit, we can see that many of these trials did not report or did not include patients on vasopressors. In fact, only five trials reported patients who were receiving vasopressors. So when we think about the harms of nutrition and shock, the Neutroria 2 trial shed light on the consequences of introducing full dose EN into patients with circulatory shock. This was a large randomized multi-center trial out of France that randomized approximately 1,200 patients to receive early full dose enteral nutrition or early parenteral nutrition. And in those receiving parenteral nutrition, they did not receive enteral nutrition until they're off vasopressors for 24 hours and the lactate was under 2. Let's look at the results of the Neutroria 2 trial. Now off the bat, you can see that in this first row reporting 28-day mortality, there was no difference between early full dose enteral or early full dose parenteral nutrition on 28-day mortality. However, if we shift your eyes to the last four rows, you can see that there was a statistically significant difference in vomiting, diarrhea, bowel ischemia, and colonic pseudo-obstruction in those patients who received early full dose enteral nutrition. When we ask who were the patients that received early enteral nutrition, notice it was mostly septic shock patients on what some of you might consider a whopping dose of norepinephrine. Remember, these patients received full dose enteral nutrition, and thus it makes me wonder if these patients were primed for gut-related complications. And we examined other studies evaluating complications of early enteral nutrition in circulatory shock. And starting with retrospective data, we looked at non-occlusive mesenteric ischemia and bowel necrosis rates across retrospective studies. And we found that the incidence was 0.3% across retrospective data. We then moved on to prospective data from mostly surgical patients, and we found that non-occlusive mesenteric ischemia and bowel necrosis occurred mostly in those individuals who received NJ feeding. In any case, the non-occlusive bowel necrosis rate in this group of patients was also 0.3%. And then finally, we looked at randomized control trials that enrolled patients with shock and had an EN arm. And if we look at the seven contemporary RCTs that included those in shock, you can see only neutrorea 2 was the only one that evaluated those exclusively in circulatory shock. In any case, when we compiled the data, we can see that the incidence of non-occlusive bowel necrosis was, again, 0.3%. Then we looked at outcomes like gut blood flow when EN is introduced into a hyperperfused gut. And these trials seem to suggest that there was no increases in non-occlusive mesenteric ischemia or bowel necrosis. And in fact, there was improved gut blood flow. And in the Berger study, they tolerated enteral nutrition of more than 1,300 kilocalories per day. Not only did we look at tolerance, but we also looked at outcomes. And so these data show that tolerance was achieved in five of eight studies. In fact, in the Merchant study, enteral nutrition was tolerated up to a norepinephrine dose of 0.14 microns per kg per minute. And in the Flordella study, up to 0.32 microns per kg per minute. And number two, early EN was associated with improved outcomes in four of eight studies. And the outcomes that were looked at were mortality benefits in the COLLID and Rainier studies. So if we go back to Neutria 2 for a moment, the adverse GI effects may be related to EN dose. And therefore, more data are needed to compare different EN doses. If we hone in on the guidelines for just a minute, we can see that just the European Society for Clinical Nutrition guidelines suggest that in patients with septic shock receiving basal pressures or anatropes, no evidence-based answer can be proposed, as no interventional studies have been reported. And since we do not know if delivering enteral nutrition is feasible and safe in septic shock, we conducted a pilot feasibility study testing trophic dose enteral nutrition in septic shock, where we randomized 15 patients to early trophic EN while on a vasopressor, and 16 patients with septic shock on no EN while on a vasopressor. We found it feasible to deliver EN. And there was, as you can see, less vomiting in those who received early EN compared to late EN. Now, no patient developed intestinal ischemia, bowel obstruction, or a gastric residual volume of more than 500. And I'm hoping this pilot work will spawn additional trials in this space. But for now, I think we can conclude that low dose or trophic dose, 10 to 30 cc's an hour of an isoosmotic formula in mechanically ventilated patients with septic shock is safe and tolerated. And RCT-level efficacy data are still lacking. So let's go back to our case for just a moment. So now your patient is prone positioned for refractory hypoxemia. And so during the pandemic, a group outlined some key strategies for practical strategies for feeding patients who are in the prone position. Now one of them was to feed into the stomach. And that was because post-pyloric tube placement is just more cumbersome. The second one is to elevate the head of bed 10 to 25 degrees, which may decrease the risk of aspiration, facial edema, and intra-abdominal hypertension, which may then propagate some of the enteral feeding intolerance. And of course, use prokinetic agents as they may decrease the risk of aspirations. These recommendations were outlined in recommendations specifically for COVID patients, where at the time the goal was to cluster care, to minimize exposure, and preserve PPE. But here are seven studies of nutrition in prone positioning. And there's three takeaways from these studies. The first takeaway is that most of these are observational studies. There is one randomized control trial comparing gastric versus post-pyloric EN in 20 patients. That is the SAM study in the middle of the screen there. The second thing you notice is that the EN dose varied. In other words, there was heterogeneity in the intervention. So some studies like the Savio and Vander and Rainier studies, they advanced EN over two to four days, while others started at goal rates. And then number three, and importantly, there were no significant differences in adverse events reported between supine and prone positioning. And in the SAM study, the frequency of microaspiration, which was detected by pepsin in a tracheal aspirate, was not different between gastric and post-pyloric feeding. And so we can conclude here that EN during prone positioning is safe and tolerated. And this is based on mostly observational-level work. And again, RCT-level efficacy data are lacking. Okay, back to our patients. Recall that you have a middle-aged individual with severe ARDS secondary to SARS-CoV-2 pneumonia. You're intubated, sedated. And now your patient is on VV ECMO for refractory hypoxemia despite prone position. This is a timeline of ECMO. And this slide is to say that techniques, and perhaps as a result, mortality have improved with ECMO. So for example, in the CSER trial, survival surpassed 50%. And in the EOLIA trial of ARDS patients, survival surpassed 60%. Now the use of VV ECMO increased during the COVID-19 pandemic. And recent data from the Extracorporeal Life Support Organization Registry show that there were more than 15,000 COVID-19 patients on ECMO worldwide. And they stayed on it for a mean of 24 days. And hospital mortality approached 50%. Now, parenthetically, about a quarter of these patients were discharged home. And about 10% of these patients were transferred to rehab. Since many patients are on ECMO for the long haul, there's biologic plausibility that nutrition could minimize some key complications, including what I've outlined here in red squares, and include bone disease, muscle atrophy, and gut mucosal atrophy. But the approach to nutrition in ECMO patients is based on weak evidence. And in fact, it's often due to personal beliefs, local logistics, and perceived or real complications of nutrition during ECMO. So why does ECMO confer instability to the gut? Now, this is an oversimplification, but there's generally two reasons. The first is that there is an exposure of blood to the circuit, which activates the coagulation cascade. And the second reason in VA ECMO is that there's a loss of pulsatile blood flow, which impairs microcirculatory flow. And that can impair gastrointestinal blood flow. And both of these are risk factors for mesenteric ischemia. Now, as stated earlier, there's not a ton of data in this space. Here are 10 studies evaluating the safety, tolerance, and some associated efficacy of enteral nutrition in ECMO. There's four key things to note here, and I'll walk through these with you. Number one, the risk of non-occlusive mesenteric ischemia rate was generally low. But in four studies, it was not even reported. Now, in the PARCC study, notice the Apache 2 was high at nearly 60% in-hospital mortality prediction. And these patients may have had additional risk factors for enteral nutrition intolerance. Number two, many patients received what is considered full-dose EN. However, the time to reach full-dose EN varied. And so there's heterogeneity in how quickly patients achieved full-dose enteral nutrition. Number three, and this is self-evident, is that all of these are observational studies. And so we're marred by some confounding in these studies. And so we have to take it with a grain of salt. And then number four, as I've outlined it in asterisks, three studies reported outcomes data. And these were the FAERI-OBE, which is the largest study, and the PARCC study. In the largest study of over 1,700 patients in cardiogenic and obstructive shock on VA-ECMO, the OBE study, which is the third one from the bottom, they saw an associated improvement in 28-day mortality when patients on ECMO received early EN. But EN dose and time to reach goal, again, were not reported. So even though these are all observational studies, perhaps we can conclude with some degree of confidence that early EN during ECMO is feasible and safe. But again, efficacy data are lacking. And so in summary, when you have a patient who is paralyzed, observational data show EN is safe and tolerated. But again, RCT-level data are lacking. In patients with circulatory shock, observational and small RCT data show low-dose EN in shock is safe, tolerated, and associated with improved clinical outcomes. But large RCT data are still lacking. And in fact, the Neutrovia 3 trial, which is testing a little bit to a lot of enteral nutrition in patients with circulatory shock, should provide additional information on this question. Number three, in those patients who are prone-positioned, again, observational data show EN in prone position is safe and tolerated. But RCT-level efficacy data are lacking. And finally, in patients on ECMO, multiple observational data show EN during ECMO is safe, tolerated, and associated with clinical benefit. But again, RCT-level data in this space are also lacking. Thank you very much for your attention.

ICU patients

enteral nutrition

circulatory shock

prone positioning

ECMO