The gut is a complex organ that performs a variety of functions, some of which are vital for ultimate survival of critically ill patients (eg, barrier function, immune competence, and metabolic regulation). Unfortunately, gut dysfunction in critically injured patients is poorly characterized and routine monitoring of gut function is crude. Currently, the best parameter of gut function is tolerance to enteral nutrition (see Chapter 60). For several reasons, this is an attractive parameter to monitor and potentially modulate. First, tolerance to enteral nutrition requires integrative gut functioning (eg, secretion, digestion, motility, and absorption). Second, locally administered nutrients may improve perfusion and optimize the recovery of other vital gut functions (eg, motility, barrier function, mucosal immunity). Third, tolerance correlates with patient outcome and improving tolerance will likely improve patient outcome. Fourth, refined therapeutic interventions to improve enteral nutrition tolerance will lessen the need to use parenteral nutrition and decrease its associated complications (see Chapter 60).
Gut dysfunction certainly contributors to enteral feeding intolerance. Thus, a brief overview of the pathogenesis of each of these dysfunctions and how they are monitored clinically will be reviewed to provide the rationale for proposed therapeutic strategies to improve tolerance to enteral nutrition.
Gastroesophageal reflux is an important contributing factor to aspiration of enteral feedings, which is a common cause of pneumonia in ICU patients. Reflux will occur whenever the pressure difference between the stomach and esophagus is great enough to overcome the resistance offered by the lower esophageal sphincter. Increases in gastric pressure can be due to distention with fluids and failure of the stomach to relax to accommodate fluid. Decreases in resistance at the lower esophageal sphincter can be due to relaxation of the sphincter muscle in response to many stimuli including mediators released during injury and resuscitation. Additional contributing factors include (1) forced supine position, (2) the presence of a nasoenteric tube, (3) hyperglycemia, and (4) morphine.
Commonly used clinical monitors include laboratory testing for presence of glucose in tracheal secretions or by observing blue food dye, which has been added to the enteral formula in tracheal aspirates. Detection of glucose lacks specificity. False-positive results can occur with high serum glucose levels or presence of blood in tracheal secretions. The use of blue food dye is poorly standardized and lacks sensitivity. More importantly, however, several reports document absorption of blue food dye in critically ill patients that is associated with death. This is presumably due to a toxic effect that blue food dye has on mitochondrial function. A recent consensus conference recommended that both these techniques be abandoned.53 Unfortunately, there are no simple monitors of GER other than observing for vomiting or regurgitation, which are not very sensitive. The head of the bed should be elevated 30° to 45° to decrease the risk that when GER occurs that it is less likely to result in pulmonary aspiration (see Chapter 57). Gastric residual volumes (GRVs) should be monitored with the presumption that a distended stomach will lead to higher volume GER (see Chapter 66).
Gastroparesis and Duodenogastric Reflux
Gastroparesis is common in ICU patients and predisposes to increased duodenogastric reflux (a potential contributing factor for gastric alkalinization) and GER (a contributing factor for aspiration). The mechanisms responsible for gastroparesis in critical illness have not been well studied. Potential factors include (1) medications (eg, morphine, dopamine), (2) sepsis mediators (eg, nitric oxide), (3) hyperglycemia, and (4) increased intracranial pressure.
The common clinical monitors for gastroparesis are intermittent measurement of GRVs when feeding into the stomach or measurement of continuous suction nasogastric tube output when feeding postpyloric. The practice of using GRV is poorly standardized and is a major obstacle to advancing the rate of enteral nutrition.54 GRVs appear to correlate poorly with gastric function and GRVs less than 200 cc generally are well tolerated. GRVs of 200–500 cc should prompt careful clinical assessment and the initiation of a prokinetic agent. With GRVs more than 500 cc, enteral nutrition traditionally was stopped. After clinical assessment excludes small bowel ileus or obstruction, placement of a postligament of Treitz feeding tube should be considered (see Chapter 60). However, a study by Reignier et al, in JAMA should that not monitoring GVR and allowing patients to advance to goal tube feed rates for adequate nutritional supplementation was not inferior to checking GVR in preventing ventilator associated pneumonia.55
Although not well studied in trauma specifically, critically ill patients are known to have a high incidence of gastroduodenal reflux. In a study of antral, duodenal, and proximal jejunal motility, Tournadre et al demonstrated that postoperative gastroparesis occurs after major abdominal surgery and is associated with discoordinated duodenal contractions of which 20% migrated in a retrograde fashion.56 Heyland et al administered radio-labeled enteral formulas through a standard postpyloric nasoenteric feeding tube in ventilated ICU patients and documented an 80% rate of radio isotope label reflux into the stomach, 25% reflux rate into the esophagus, and a 4% reflux rate into the lung.57 Finally, Wilmer et al reported monitoring bile reflux in the esophagus of ventilated ICU patients using a fiberoptic spectrophotometer that detects and quantifies bilirubin concentration.58 Endoscopy was performed and documented erosive esophagitis in half of the patients of which 15% had pathologic acid reflux and 100% had pathologic bile reflux. These studies provide convincing evidence that duodenogastric reflux is a common event in ICU patients.
The stomach, through secretion of hydrochloric acid, normally has a pH below 4.0. This acid environment has been correlated with the relatively low bacterial counts found in the stomach. Reviews of several studies have shown that alkalinization of the stomach through the use of antacids, H2 antagonists, and proton pump inhibitors results in gastric colonization by bacteria not normally found in the stomach; and several, but not all, studies have shown that gastric colonization predispose patients to ventilator-associated pneumonia,59 and can increase the risk of community-acquired Clostridium difficile–associated disease.60
Several animal studies conducted by our group have shown that both lipopolysaccharide administration and mesenteric ischemia/reperfusion result in the gastric accumulation of an alkaline fluid.61 This most likely results from a decrease in gastric acid secretion with continued gastric bicarbonate secretion and the reflux of duodenal contents into the stomach. Even more recently, we have reported that the pH of gastric contents in trauma patients also is elevated, possibly due to similar events.62 Thus, even without the administration of antacids or inhibitors of acid secretion, gastric alkalinization and bacterial colonization of the stomach are likely in this group of patients. When this is combined with the gastroparesis often seen in these patients (see above), it is easy to envision the stomach becoming a major source of bacteria for ventilator-assisted pneumonia and perhaps translocation to other organs.
Impaired Mucosal Perfusion
Shock results in disproportionate splanchnic vasoconstriction. The gut mucosa appears to be especially vulnerable to injury during hypoperfusion. The arterioles and venules in the small bowel mucosal villi form “hairpin loops.” This anatomic arrangement improves absorptive function, but it also permits a countercurrent exchange of oxygen from the arterioles to the venules in the proximal villus. Under hypoperfused conditions, a proximal “steal” of oxygen is believed to reduce the Po2 at the tip of the villi to zero. The gut mucosa is further injured during reperfusion by reactive oxygen metabolites and recruitment of activated neutrophils (see Chapter 61). This mucosal injury, however, appears to quickly repair. Mucosal blood flow does not always, though, return to baseline with resuscitation and this is in part due to defective vasorelaxation. The gut mucosa is also vulnerable to recurrent episodes of hypoperfusion from ACS, sepsis, and use of vasoactive drugs. Whether recurrent hypoperfusion results in additional ischemia/reperfusion injury is not known, but it is reasonable to assume that hypoperfusion would decrease gut nutrient absorption and render the patient more susceptible to NOBN.
Monitoring gastric mucosal perfusion in the clinical setting can be done by gastric tonometry. With hypoperfusion, intramucosal CO2 increases due to insufficient clearance of CO2 produced by aerobic metabolism or due to buffering of extra hydrogen ions produced in anaerobic metabolism. As intramucosal CO2 accumulates, it diffuses into the lumen of the stomach. The tonometer measures the CO2 that equilibrates in a saline filled balloon (newer monitor uses air-filled balloon) that sits in the stomach. This is the regional CO2 tension (PrCO2) and is assumed to equal the intramucosal CO2 tension. Using this measured PrCO2 and assuming that arterial bicarbonate equals intramucosal bicarbonate, the intramucosal pH (pHi) is calculated by using the Henderson–Hasselbalch equation. Numerous studies have documented that a persistently low pHi (or high PrCO2 level) despite effective systemic resuscitation predicts adverse outcomes and that attempts to resuscitate to correct a low pHi do not favorably influence mortality.63 Unfortunately, alternative resuscitation strategies have not been able to increase pHi to improve outcome and thus this monitoring tool is in search of a novel application (see Chapter 55).
Impaired Intestinal Transit
Laboratory models of shock, bowel manipulation, and sepsis demonstrate that small bowel transit is impaired.64 This impairment in turn is expressed as a decrease in the number and/or force of contractions, or as an abnormal pattern of contractions. Although the results in animal models are convincing, surprisingly, clinical studies indicate that small bowel motility and transit are more often than not well preserved after major elective and emergency laparotomies.65 This observation coupled with the observation that small bowel absorption of simple nutrients is relatively intact provided the rationale for early jejunal feeding (see Chapter 60).
Clinical studies have documented that the majority of critically ill patients tolerate early jejunal feeding.66 In a recent study, severely injured patients had jejunal manometers and feeding tubes placed at secondary laparotomy.65 Surprisingly, 50% had fasting patterns of motility that included components of the normal migrating motor complexes (MMCs). These patients tolerated advancements of enteral nutrition without problems. The other 50% who did not have fasting MMCs did not tolerate early advancement of enteral nutrition. Of note, none of the patients converted to a normal-fed pattern of motility once they reached full-dose enteral feeding. This could be due to infusion of caloric loads insufficient to bring about conversion. On the other hand, the failure to develop fed activity, a pattern of motility promoting mixing and absorption, might explain why diarrhea is a common problem in this patient group.
Although manometry can be used to monitor motility, it is not practical. Unfortunately, simpler indicators of motility such as the presence of bowel sounds or the passing of flatus are unreliable. Other minimally invasive methods to monitor transit are needed. Contrast studies through the feeding tubes are relatively simple, but not validated.
Impaired Gut Absorptive Capacity
Small bowel absorption of glucose and amino acids is depressed after trauma and sepsis. Multiple factors have been identified including (a) cytosolic calcium overload, (b) ATP depletion, (c) diminished brush border enzyme activity, (d) decreased carrier activity, (e) decreased absorptive epithelial surface area, and (f) hypoalbuminemia. In an animal study, intestinal ischemia/reperfusion caused significant mucosal injury and significant depletion of mucosal ATP.52 When this was combined with exposure of the bowel to a nonmetabolizable nutrient, the damage and ATP depletion were more severe and the absorption of glucose was impaired. In contrast, exposure of the bowel to metabolizable nutrients (such as glucose) preserved ATP levels, protected against mucosal injury, and improved gut absorptive capacity (GAC).67
The clinical significance of these observations remains unclear because most patients tolerate enteral nutrition when delivered into the small bowel. However, decreased GAC may be a cause for diarrhea and may explain why patients commonly experience diarrhea with reinstitution of enteral nutrition after prolonged bowel rest. Unfortunately, there are no easily performed clinical monitors for GAC. d-Xylose absorption is clinically available but is used most frequently to diagnose chronic malabsorption.
Diarrhea may be indicative of depressed GAC, but there are other causes for diarrhea in the critically ill including impaired transit, bacterial overgrowth (eg, presence of C. difficile), contaminated enteral formulas, abnormal colonic responses to enteral nutrition (eg, ascending colon secretion rather than absorption, or impaired distal colon motor activity), and administration of drugs, which contain sorbitol (eg, medication elixirs) or magnesium (eg, antacids).
Increased Gut Permeability
Enhanced paracellular permeability represents a type of barrier dysfunction that allows increased passage between viable cells and may induce an inflammatory cascade. The major components of the epithelial barrier are tight junctions, which bind cells together and serve as the gateways to the underlying paracellular spaces. The integrity of the tight junctions is modulated by the actin cytoskeleton. Under conditions of ATP depletion, such as would occur during shock, disruption of the actin cytoskeleton with consequent opening of the tight junctions and loss of the integrity of the permeability barrier can occur.52
Intestinal barrier dysfunction has been suggested as a means by which inflammatory cytokines can lead to the SIRS and MOF. Additionally, increased intestinal permeability has been documented in high-risk patients after burns, sepsis, and shock, in most but not all studies.68
Decreased Gut Mucosal Immunity
During periods of intestinal disuse, critically injured patients are subject to a reduction in gut mucosal immunity. Lack of enteral stimulation (such as during starvation or with use of parenteral nutrition) quickly leads to lack of immunologic protection by mucosal-associated lymphoid tissue (MALT), which normally provides protection for both the gastrointestinal (GI) and respiratory tracts against microbial flora and infectious pathogens. Kudsk has demonstrated a link between intestinal IgA, intestinal cytokine production, and the vascular endothelium of the GI tract. With enteral stimulation, IL-4 and IL-10 production from the lamina propria of the small intestine stimulate production of IgA on the mucosal surface and inhibit intracellular adhesion molecule-1 (ICAM-1) of the vascular endothelium and subsequent neutrophil-associated inflammation and injury.69
Increased Gut Colonization
With progressive gut dysfunction, the normally sterile upper GI tract becomes colonized with organisms that become pathogens in nosocomial infections. Gastric alkalinization, paralytic ileus, loss of colonization resistance due to broad-spectrum antibiotics, and decreased local gut immunity have all been proposed mechanisms by which the upper GI tract becomes colonized. In an effort to decrease the incidence of infectious complications, selective digestive decontamination (SDD) has been proposed. SDD generally consists of topical nonabsorbed antibiotics along with a short course of parenteral antibiotics. There have been numerous clinical trials and at least six meta-analyses addressing this practice. Most studies have examined the incidence of ventilator-associated pneumonias and mortality and, in general, have demonstrated a decrease in both. Despite rather impressive results of these trials, most of which have been performed in Europe, intensivists in the United States have generally avoided use of SDD due to concerns of inducing antimicrobial resistance.70
Aggressive fluid resuscitation can result in significant bowel edema and altered intestinal function. Both in the laboratory and clinically, significant bowel edema following resuscitation can lead to elevated intra-abdominal pressure and potentially to ACS. The precise mechanisms by which bowel edema adversely effects bowel function, however, have not been fully elucidated. Laboratory investigations by Moore-Olufemi et al have shown that bowel edema alone (not associated with gut ischemia/reperfusion or hemorrhagic shock) is associated with impaired intestinal transit and this can be reversed with enteral feeding.71 A number of clinical studies, even though not in resuscitated trauma patients, have demonstrated the benefit of fluid restriction on postoperative bowel function and complications.72