CASE

A 60-year-old male with esophageal cancer was admitted after a witnessed aspiration, leading to pneumonia with acute hypoxemic respiratory failure and septic shock. Prior to admission, he had poor oral intake, resulting in 10 lb (4.5 kg) of weight loss over 2 months. Current body mass index is 19 kg/m2. As part of his hospital course, he received 3 L of intravenous fluids and his blood pressure was 120/50 while receiving norephinephrine 0.1 mcg/kg/min.

DISCUSSION

This case presentation and panel discussion took place during the virtual course, Comprehensive Nutritional Therapy: Tactical Approaches in 2021 (Part 1, March 19, 2021; Part 2, March 20, 2021), which was organized by the ASPEN Physician Engagement Committee and preceded the ASPEN 2021 Nutrition Science & Practice Conference. The moderator for this session was Jayshil J. Patel, MD, and the panelists were Yaseen M. Arabi, MD; John Alverdy, MD; Stephen A. McClave, MD; and Zudin Puthucheary, PhD.

Protein supplementation in a patient who is critically ill

Dr McClave: The controversy regarding protein supplementation continues. In the United States, there is a feeling that there is no limit to the utilization of protein, and it is key for us to provide adequate protein to overcome anabolic resistance. Data from Europe, as Dr Puthucheary presented, reveal that there is a price to pay for providing high amounts of protein. So we still do not know the right amount of protein to provide in patients who are critically ill. Are there any markers that we can use, such as the urea-to-creatinine ratio, that allows us to monitor tolerance?

Dr Puthucheary: The urea-to-creatinine ratio has been around for a very long time especially in mobilization and muscle physiology circles. However, we have never used it in a clinical setting to monitor tolerance of protein intake. There is also a history in medicine of jumping before the evidence in critical care and doing things that are not helpful or actually harmful to the patient. We have seen this especially in the early treatment of coronavirus disease 2019 (COVID-19). So I would say that we need to conduct trials to see if the urea-to-creatinine ratio is safe and an efficacious monitor of protein loading.

Dr Arabi: I think that the benefit of protein may very well be time dependent. So outcomes could be different if you provide high amounts of protein in the first few days of critical illness vs during the recovery phase. We are now conducting a randomized control trial of 2500 patients in which patients are randomized starting on day 5 to receive higher protein (REPLENISH trial, Replacing Protein Via Enteral Nutrition in Critically Ill Patients, NCT04475666). The enrolled patients in this trial are patients who are mechanically ventilated on intensive care unit (ICU) day 4; therefore, these patients are likely to have a relatively long stay in the ICU. Patients are randomized to start receiving an extra 1.2 g/kg/day of protein enterally for the duration of their ICU stay. Hopefully, the findings will provide more insight regarding the clinical benefit of giving moderate vs high protein during the recovery phase. The EFFORT Trial,1 which is also evaluating higher vs standard protein, is addressing a different question; by starting high protein from ICU day 1.

The other point that we are learning from our experience with the COVID-19 pandemic is that, in nutrition, we need a more efficient way of answering clinical questions. It is taking a really long time to answer one question at a time. With COVID-19, adaptive platform clinical trials have been efficient in answering multiple questions at the same time. So we need to learn from this model and think outside of the box in designing our nutrition clinical trials.

Dr McClave: This is a very interesting point. Many components of nutrition research—such as having an accurate measure of energy needs through indirect calorimetry, having a Nutrition Risk in the Critially Ill (NUTRIC) score to identify disease severity, the amount of feedings, how many calories, and how much protein—so far have not seemed to change clinical outcomes. A completely different approach is to state that it is not those things that in themselves matter, but their impact on other factors such as the microbiome or their impact on immune signaling. So what is the most important component of nutrition that is going to change outcome?

Dr Alverdy: As we evaluate these discussion points, I think it is important for us to start by qualifying key terms. As an example, when we bring up the term “gut barrier,” what are we referring to? By “gut barrier,” are you referring to selective permeability to certain macromolecules? If so, which ones? We often neglect to discuss that level of detail. Similarly, we bring up immune function but fail to discuss the details. Are we referring to the impact on macrophages, T cells, or some other immune cell? It is the same with the term “enteral nutrition.” We treat the gut as just a tube, a conduit to which we can throw in some stuff and it gets into the bloodstream. However, much happens to the nutrients, the digestive response, and the microbiome once foodstuffs enter the gut conduit. It strikes me that we simplify our terms too often.

We have learned in pharmacology that there is a first pass effect with drugs in which medicines are absorbed and undergo processing by the gut followed by the liver prior to release into the bloodstream. Well, under this framework, the first pass for oral nutrients is the microbiome. The second pass is of those metabolites generated by the microbiome, such as butyrate, indols, tryptophans, etc, that undergo further metabolism by the mucosa before they get into the portal system. In fact, when we measure butyrate in the portal vein, you can hardly measure it because it is a volatile fatty acid. So the key questions to ask include how are nutrients processed through the various layers before they enter the bloodstream? What role does the microbiome and mucosa play in this response? How does passage through these layers alter immune cells? What nutrients actually make it intact to the liver and how do they affect or are affected by Kupfer cells in the liver?

These are key questions that we must answer. I would also push further and say that it is key for us to try and understand not only the differences between the two groups (intervention and the control) in a trial but also the within-group differences. Let's take the example of enteral nutrition. In situations where demand exceeds supply in terms of oxygen and blood flow, if we push enteral nutrition too far, we are going to have a problem. Yet if we examine the clinical data, how often does this situation actually arise? Let's assume 5% of the time. Then the question is what happened the other 95% of the time? We regularly compare between-group data and ignore within-group data. I call this the “between-group comparison fallacy.” For example, has anyone determined the within-group difference that occurs when you give a 50% Lethal Dose (LD-50) of a toxin or infectious agent? Generally, treatment is given to one group and compared with vehicle in the other. If the mortality rate, for example, between-group A drops from 50% to 25% and the P value is <.05, the analysis stops. No one is interested in what governs why, following an LD-50 dose of an infectious agent, half the mice live and the other half die. We are currently studying this and our hypothesis is that it is the metabolites that the microbiome produces that determines the “within-group” outcome in such models.

What is so exciting and interesting about the field of microbiome sciences is the high variability in the microbiome when it collapses in structure and function and then recovers following a major perturbation. For example, recent studies demonstrate that matched human volunteers given the same antibiotics respond in unique ways to the antibiotics; some rapidly recover their microbiome whereas other do not. Now, because these volunteers are the same age, have the same medical comorbidities, and the same baseline microbiome, you would suspect that their microbiome is going to dive down and recover similarly. However, there is much variability in this response. Therefore, the extent to which one's microbiome is affected by a perturbation, its ability to recover and the essential metabolites it produces that drive the immune response needs further study. Similarly, the extent to which that individual's visceral splanchnic circulation can deliver oxygen and the extent to which enteral nutrition is demanding oxygen is also going to play a role in outcomes with enteral nutrition. Sorting out these issues may be important for us to understand why randomizing patients to diet A or B and examining outcome alone will not yield the expected results. It is now going to be important to account for changes in the microbiome that occur within a treated group of patient to define the variability in outcome that often blurs our expected results.

Bedside evaluation of malnutrition and implications for nutrition optimization 2 Dr Patel: I agree. We bring up the fact that pre-existing malnutrition has significant impact on outcomes. However, we do not have the best means to identify pre-existing malnutrition. We also state that the patient's stay in the ICU, especially if prolonged, leads to development of malnutrition and sarcopenia. We use these concepts of pre-existing and acquired malnutrition to stratify patients into whether they should receive a higher protein dose. However, Dr Alverdy, what you are bringing up is identification of a metabolic profile for the patient that allows us to discern whether they will even tolerate nutrition including incorporation of those macronutrients for skeletal muscle formation. How do you envision this happening at the bedside?

Dr Alverdy: I will just answer your question very simply and say that we cannot just measure who is there but also what are they doing. Metabolomics is the next hot thing for expelled gas chromatography to measure both what is coming out of the lung and what is in the stool. These measurements have to be done serially—give the drug, measure the response, give the drug, and measure the response.

Examining the evidence base for nutrition in patients who are critically ill 3 Dr McClave: Let me turn to the use of fiber. In the previous guidelines, perhaps the risk of fiber is overstated. We are now seeing a trend in giving fiber, going from refined formulas to whole-food formulas. We are also seeing more consideration for the use of probiotics or fecal microbiota transplantation. Are we moving in the right direction and are these safe for the patient?

Dr Puthucheary: I think there is this emotional wish to do something nutritionally. But by doing this without the scientific support, you do not know if, despite your good intentions, you are harming patients.

Dr Patel: We call that a “commission bias.” It is a natural tendency to want to help and therefore we throw everything at our patients without the full knowledge of whether that intervention is going to help or not.

Dr Puthucheary: And the problem is that this commission bias is actually a hindrance in two ways. The first is in terms of the clinical care in which we do not know if we are causing harm. The second is in terms of clinical trials, in which we then have to overcome this bias. As an example, if we are making someone ketogenic, we often state that we have to give them protein. But the question is why do we have to give them protein? The answer often is that we just do and we end up in a circular argument with a lack of scientific support that leads to harm.

Dr Alverdy: For me, the instinct to act is the main problem. We have high-resolution analytic tools today, sequencing, and gas chromatography of expelled gases, metabolomics, etc. We need to have a rational top-down approach to determine the precise nutrition needs of patients. We cannot simply empirically develop an enteral nutrition product that contains fiber, probiotics, and/or prebiotics and call it the Holy Grail. The gut during critical illness needs both seed (microbes) and soil (nutrients). We can just make it up. What is exciting is that the workflow, analytics, and technology exist to get us there, we just need to be patient and work in a rational top-down assembly method to develop personalized nutrition.

Dr Arabi: We see the impact of this inherent bias when conducting clinical trials pertaining to protein as well. Many colleagues are in the pro–high-protein camp and do not favor providing less protein. Others are in the low-protein camp, and have difficulty providing higher protein doses. We have to overcome these biases in order to generate reliable evidence. In many regards, we need to go back to the basics. We are seeing this with formulas such as whole-food formulas. This has been practiced in low- and middle-income countries for a long time and are now being developed and utilized more in high-income countries. However, we need to make sure that these formulas are tested properly.

Branch chain amino acids and mitochondrial function 4 Dr Patel: For patients in acute liver failure, hyperammonemia can sometimes be attenuated through pharmacological interventions such as lactulose and/or rifaximin therapy. In Europe, branch chain amino acids have been utilized as a means to pull that ammonia into the muscle for deamination. Are there any data that you are aware of utilizing branch chain amino acids to enhance or “rev up” the mitochondria in this situation?

Dr Puthucheary: I believe the L-ornithine-L-aspartate (LOLA) trial came the closest to evaluating this but did not look at the level of the metabolic function.2 Srinivasan Dasarathy, MD from Cleveland Clinic, focuses on ammonia and raises an important point that we forget about ammonia in nonliver failure and the toxic role it has for muscles and for recovery.

Dr Patel: It is an excellent point, but we have historically been taught not to measure ammonia because of the false negative in measurement.

Dr Puthucheary: Plasma ammonia has no relationship with muscle or tissue ammonia. There are groups that are working on trying to see if plasma metabolomics can help give us greater insight into ammonia metabolism. But until we have that data, plasma ammonia levels are not that useful.

Future of critical care nutrition research 5 Dr Patel: Can I ask where do you think we are headed with critical care nutrition? What do you think are the one or two most important questions that need to be answered to help satisfy what we do at the bedside or at least propel us a bit at the bedside?

Dr Alverdy: One thing that needs to happen is that we develop a critical evaluation of the formulations we use now and determine how they can be supplemented so that when we feed the patient this formula, their microbiome over the course of critical illness resembles or reassembles in both community structure and function toward that of a healthy microbiome. Right now we are giving X vs Y and looking at crude outcomes such as C-Reactive Protein (CRP), Interleukin 6 (IL-6), and days on ventilator. But these are typically fiber-free, sterile, chemically refined formulas and this cannot be the right approach because our gut has evolved and spent its life consuming regular unprocessed food. Although we may understand that our patients need more substantive unprocessed food, we also know that during critical illness, their gastrointestinal (GI) tract is not going to tolerate meat, potatoes, and peas while they are on vasopressin and norepinephrine.

The point I am trying to make is that we cannot simply vary formulation, feed, and measure crude outcomes. We should be feeding and working in conjunction with Big Pharma to say, “Here are the end points that we think you need to focus on,” and food passes through many levels of interactions (microbiome, mucosa, submucosa, lymph nodes, and liver) and measuring plasma metabolites may be inadequate. Instead, the question, in my opinion, is how is this formula assembling and building a healthy microbiome through the course of illness, especially when the patient is in the burn center, going to the operating room, getting a blood transfusion, or getting antibiotics? How is this formula driving the resilience of the microbiome in the same way as it has over millions of years when unprocessed food is consumed?

Dr Arabi: I would say a priority would be to examine some of our current practices to evaluate if they are effective and more importantly to ensure that they are safe. In nutrition, we may be doing certain things that are not useful or lead to wastage of resources and energy and are even harmful to the patient. So, I think we need to look at our clinical practices in an objective manner in proper study designs.

Dr Puthucheary: I am going to say that the biggest thing we can do is to push big nutrition societies to put an end to the demagogues. Individuals who stand up and say “I believe this and therefore you should do this.” We need to deconstruct this pattern and focus on saying whether there is evidence or not. If you want to give an opinion, you should qualify that statement by clearly stating that there is no evidence behind the statement, however you still want to do this. This will allow us to develop a very clear list of what the unanswered questions are.

The second major change is to stop creating a separation between protein and energy, which our patients do not do. None of us are going to have dinner tonight and say, “I'm just going to eat protein,” or “I'm just having calories tonight.” We need to look at an integrated system and we need to understand the signs. We need to bring back facts and not opinions into nutrition.

Fiber in patients who are critically ill 6

Dr McClave: Let's discuss fiber for a second. There are 10–12 randomized controlled trials of fiber and critical illness that say fiber has benefits and if you wait until the recovery period to add fiber, it may be too late.

Dr Alverdy: Let's take this point and deconstruct why we do not give fiber. We recently wrote a review article that should be coming out soon addressing this exact point. If we look back to the 1960s, we developed products such as vivonex for space flight because they had not figured out how to get astronauts to defecate in space yet. There is a legacy effect that drives us to continue to use chemically defined enteral nutrition because it is well tolerated and the patients do not defecate. Although this formula may be convenient for nursing care, is easier to provide, is sterile, and comes in a can, these reasons alone do not justify why we do not give fiber. However, if we think about it, we begin to realize that we really do not have a good scientific reason for not providing it.

Dr McClave: This is a great point, Dr Alverdy. As we deconstruct our decision making regarding fiber in critical illness, there are certain reports such as the Dutch PROPATRIA study in which they reported a cast of fiber in the small bowel with ischemia.3 Perhaps this report led to panic and irrational fear of fiber. There are four other cases in literature with similar findings and perhaps those impacted practice significantly. Dutch PROPATRIA study was in 2008, and so we now have 12–13 years of fear that is probably overly exaggerated.

Nutrition therapy in COVID-19 7 Dr McClave : Let's discuss COVID-19 because this is such as unique disease process. These patients are so hypermetabolic for such a long time. Does COVID-19 change any of our paradigms about the way we are delivering nutrition therapy now? Dr Arabi: I think there are several aspects of COVID-19 that may make this patient cohort a unique subgroup; some aspects are related to the disease itself and others to the way we treat these patients. The disease and especially how it involves the GI tract is yet to be fully understood. Additionally, these patients are often treated with noninvasive ventilation for a prolonged period of time, which, in some patients, is associated with limited feeds. The use of corticosteroids, the use of IL-6 receptor antagonists, and other immune modulators may also play a role in how patients with COVID-19 interact with nutrition. So, I think there are enough reasons to justify evaluating this group of patients in clinical trials because to evaluate whether they may require a different approach to nutrition compared with the rest of patients who are critically ill. 8 Dr McClave: We recently did a small survey to find out how nutrition was actually being delivered and we found that one of the overlying prejudices is the fear of aspiration. We have spent two decades dealing with the fear that feeding anything into the GI tract would lead to aspiration pneumonia and rates of pneumonia would go up. And we have data from randomized control trials showing that feeding through the enteral route leads to lower rates of pneumonia. In the COVID-19 setting, these patients have such low marginal pulmonary reserve that we are even more afraid of aspiration. Is that a legitimate fear and should we back off of enteral feeding and modify our nutrition strategy?

Dr Arabi: Typically, this is not a major concern with most patients with COVID-19 who can protect their airway, especially if the amount of enteral nutrition provided is moderate. These patients take a long time to recover and so we cannot deprive them from feeding for a prolonged time period. However, some caution is needed to avoid giving aggressive feeding during the prone position.

Fecal microbiotal transplantation in patients who are critically ill 9 Dr Patel: Should we be providing a “healthy” fecal transplant as part of a protocol in caring for critically ill patients and what are the pros and cons of this strategy?

Dr Alverdy: We should not. I have reviewed a few manuscripts that are case reports showing improvement in patient outcomes; however they are very anecdotal. There has also been a case report noting transmission of a multidrug resistant Eschericheria Coli that resulted in the death of a patient. This is what I can empiricism ad absurdum. We cannot simply “make it up as we go.” We need to be more precise and prescriptive about proving the seeds and soil that are needed during critical illness.

In the group I work with, our feeling is that once we understand what we need, we can organically home grow it by providing the seeds as well as the substrate (soil and nutrition) as part of microbiome maintenance therapy that keeps the microbiome at the level that we want through the course of whatever level of critical illness exists. If the patient has COVID-19, the patient is getting dexamethasone, the patient is on an IL-6 inhibitor, or patient is going back to the operating room for a liver transplant, whatever the insult, we need to know what perturbations to the microbiome are occurring, what key immune-stimulating metabolites are missing, and how to provide them. However, just giving fecal transplantation is a big mistake in my opinion.

Dr McClave: The counter to that is use of Fecal Microbiota Transplantation (FMT) in Clostridium difficile. We use it frequently and the change in toxicity it produces is dramatic. The patients are essentially different the next morning. Providing probiotics is also like putting one plant in the middle of a bare yard when you are trying to replenish the garden. So, we may not be ready for FMT right now, but is it a direction that will eventually go down?

Dr Alverdy: There are over 100 start-up companies that are trying to come up with what I refer to as the perfect “crapsule.” These companies have raised incredible amounts of money in the range of $20 million–$100 million. In our mouse model of peritonitis or hepatectomy for example, if we give a fecal transplant from normal untreated healthy mouse, we can drive survival to 100% . Yet when we give that same regimen to mice with pancreatitis (created by injecting sodium deoxycholate in the pancreatic duct) mortality is increased to 100%. This is perhaps because pancreatitis causes more gut permeability or more systemic inflammation and the microbes you add actually worsen things. So, there is a lot for us to learn, and there is not going to be a one-size-fits-all fecal transplant to use on critically ill patients.

Role of exclusive parenteral nutrition in patients who are critically ill 10 Dr Patel: The European Society for Clinical Nutrition and Metabolism (ESPEN) guidelines are more supportive of parenteral nutrition at an earlier time point in the critical illness process compared with the ASPEN guidelines. As pointed out, ESPEN guidelines allow for parenteral nutrition to be provided as early as 3 days after inability to feed the intestine. Why the difference? Do you feel this use of early parenteral nutrition is supported by evidence?

Dr McClave: I think ESPEN has been softening on that stance. They used to say that by 24–48 h, if you are not meeting requirements, you should start bringing in parenteral nutrition. They have now pushed that back to about a week in their latest guidelines. So I think our guidelines are closer in agreement.

Dr Arabi: I think in principle we should follow what is the best available evidence based on randomized controlled trials and as has been mentioned, 1 week is the best available evidence at the present time. So unless there is a different direction, I think that is the way to go.

Dr Patel: I have also shifted my practice to a “less is more” strategy during the acute phase of critical illness. My goal in the early acute phase is to focus on preserving gut epithelial barrier function and thus, I am not thinking about parenteral nutrition at this phase. The caveat is obviously someone who comes in with pre-existing malnutrition and we just cannot use their GI tract. There may be a role for early exclusive parenteral nutrition in these individuals. Anecdotally, some COVID-19 patients may fall in this category.

CONFLICT OF INTERESTS

Manpreet S. Mundi has research grants from Fresenius Kabi, Nestlé, VectivBio, and Real Food Blends and is on advisory board of Fresenius Kabi and Baxter. Jeffrey I. Mechanick received honoraria for lectures and program development by Abbott Nutrition. Jayshil J. Patel is a consultant for Baxter. Yaseen M. Arabi is a principal investigator on the investigator-initiated REPLENISH trial (Replacing Protein Via Enteral Nutrition in Critically Ill Patients, NCT04475666). John Alverdy is supported in part by National Institutes of Health RO1 NIGMS-5R01GM062344-20. John Alverdy is founder and Chief Research Officer of Covira Surgical. Zudin Puthucheary received funding from Fresenius Kabi, Faraday Pharmaceuticals, Nestlé, Baxter, Nutricia, Orion, Lyric Pharmaceuticals, GlaxoSmithKline, and Vitaflo. Stephen A. McClave is a part of the Speaker's Bureau for Nestlé, Abbott and is on the Advisory Board for Avanos, Baxter.

FUNDING INFORMATION

The content of this article was presented during the virtual course, Comprehensive Nutritional Therapy: Tactical Approaches in 2021 (Part 1, March 19, 2021; Part 2, March 20, 2021), which was organized by the ASPEN Physician Engagement Committee and preceded the ASPEN 2021 Nutrition Science & Practice Conference. The author(s) received a modest monetary honorarium. The conference recordings were posted to the ASPEN eLearning Center https://aspen.digitellinc.com/aspen/store/6/index/6.

AUTHOR CONTRIBUTIONS

Manpreet Mundi, Jayshil J. Patel, and Jeffrey I. Mechanick contributed to the conception and design of the manuscript. Manpreet Mundi contributed to the acquisition of data. Manpreet Mundi drafted the manuscript. Jayshil J. Patel, Jeffrey I. Mechanick, Yaseen M. Arabi, John Alverdy, Zudin Puthucheary, and Stephen A. McClave critically revised the manuscript. All authors gave final approval and agree to be accountable for all aspects of work ensuring integrity and accuracy.