• -

  To understand the physiopathology of the gastrointestinal tract in the critical patient with hemodynamic instability and the potential benefits of enteral nutrition (EN).

• -

  To review the available evidence on the benefits and risks of EN in patients of this kind.

• -

  To summarize the recommendations of the main clinical practice guides.

• -

  To offer practical ideas and the point of view of a group dedicated to research in this field.

The physiology of the splanchnic circulation is complex, but has been known for some years, and a basic understanding of its characteristics is essential in order to correctly interpret the alterations that occur in critical patients with hemodynamic instability. Under resting and normal conditions, 20–25% of cardiac output is located in the splanchnic circulation. In turn, the metabolic activity in this region accounts for 30% of overall body oxygen consumption. When eating, the blood flow in this zone can double thanks to a phenomenon known as postprandial hyperemic response. This response is basically mediated by local factors, giving rise to notorious splanchnic vasodilatation.1,2 The anatomical configuration of the intestinal microvascularization is complex, with arterial and venous plexuses at mucosal and submucosal level, and in the muscularis propria. This system is able to redistribute blood flow in the case of decreased intravascular volume within the systemic circulation, allowing precise regulation of splanchnic blood flow (Fig. 1). In the intestinal villi, the arterial and venous vessels run parallel to each other, though with flow in opposite directions. This anatomical distribution allows the direct passage of low-molecular weight molecules (such as oxygen) from the arterial to the venous circulation, conforming a “countercurrent” exchange mechanism. Although the latter is not relevant under physiological conditions, in the presence of hypoperfusion it generates a decreasing tissue oxygen pressure gradient from the base to the tip of the villus (Fig. 2). As a result, the tip of the intestinal villus is particularly sensitive to tissue hypoxia.

Figure 1.

Intestinal microvascularization (i). Schematic representation of the anatomical distribution of the intestinal microvascularization. Regulation can be made at several levels (precapillary, capillary and postcapillary), thus allowing blood flow redistribution.

(0.15MB).

Figure reproduced from Folkow,3 with permission.

Figure 2.

Intestinal microvascularization (ii). Schematic representation of the anatomical distribution of the intestinal microvascularization (artery, vein and lymphatic vessel) at villus level. The passage of low-molecular weight molecules (dots) and the countercurrent mechanism are shown. The triangle at left represents the oxygen concentration gradient from villus base to tip.

(0.15MB).

Shock is characterized by blood flow redistribution with vasoconstriction at splanchnic circulatory level and in peripheral tissues, in an attempt to maintain crucial brain and coronary perfusion.4 This can give rise to an imbalance in the oxygen supply/demand ratio at intestinal level, with resulting ischemia. This situation in turn leads to depletion of the cellular ATP reserves, with rupture of the tight junctions between the epithelial cells. Consequently, the solute concentration gradients between the apical and basolateral compartments are lost, producing intracellular edema, necrosis and apoptosis. The end result of this alteration of the intestinal microcirculation is rupture of the intestinal epithelial barrier, favoring bacterial translocation phenomena. Such persistent alteration of the microcirculation in shock patients5 can perpetuate the proinflammatory response and favor progression toward multiorgan dysfunction syndrome, which in turn results in increased splanchnic hypoperfusion – thereby completing a vicious circle that can increase the probability of a fatal outcome (Fig. 3). In addition, in this complex scenario, different patient interventions such as vasoactive and inotropic support,6 mechanical ventilation or the use of hemodynamic support systems, act upon the splanchnic circulation, with consequences that are hard to predict. Considering the above, nutritional support in these patients becomes a genuine challenge. From a conceptual and physiopathological point of view, enteral nutrition (EN) administered to the critical patient with hemodynamic instability would induce an increase in the need for oxygen at intestinal level and in splanchnic blood flow due to vasodilatation at this level. In this respect, predominance of an increased oxygen demand would predispose to intestinal ischemia, while an increase in splanchnic perfusion could counter the microcirculatory alteration found in patients in shock.2,7

Figure 3.

Schematic representation of the relationship between shock and splanchnic hypoperfusion. cytok: cytokines; MODS: multiorgan dysfunction syndrome.

(0.18MB).

In the last 15 years experimental models in animals and also human clinical studies have produced evidence of a beneficial effect of EN in the critical patient with hemodynamic instability.

The most feared complication of EN support in the hemodynamically unstable patient is intestinal ischemia, particularly of a non-occlusive nature, produced as a consequence of increased oxygen consumption in a proportion not adjusted to the increase in splanchnic blood flow, with consequent alteration of the supply/demand ratio at intestinal mucosa level. This is an infrequently described complication, though the associated mortality rate is high: between 11 and 27% in the postoperative period of heart surgery.21 Little is known of the incidence of intestinal ischemia. A retrospective study based on a registry of 4311 critical trauma patients22 reported an incidence of 0.3%, with an associated mortality rate of 44%. The condition was most often detected in the second week of admission to the ICU. In an attempt to identify a “risk profile” for the development of intestinal ischemia, the authors evaluated a series of clinical and laboratory test parameters, and came to the conclusion that none of them was sufficiently specific. Furthermore, some parameters, such as abdominal bloating, were of late onset. Thirty-one percent of the cases had significant abdominal trauma. The study did not specify the precise moment in which EN was started: the authors only indicated that EN began after the initial resuscitation phase. Likewise, no mention was made of whether EN was administered according to an established protocol in all cases, or of the hemodynamic situation of the patients at the time of starting EN support.

In recent years there have been reports of non-occlusive intestinal ischemia after EN, particularly following the use of postpyloric feeding tubes and with the utilization of nutritional formulas rich in non-fermentable fiber.23 However, in most of these studies EN was not provided on an early basis, and important details are missing, such as the hemodynamic situation of the patients at the start of nutritional support or the way in EN was provided, and daily monitoring of the clinical condition of the patients.24

Daily and careful monitoring of the possible alarm signs of intestinal ischemia in these high risk patients is of crucial importance. Such signs can be of a clinical, laboratory or radiological nature (Table 2). None of them are sufficiently sensitive or specific to establish a firm diagnosis, though their presence should cause us to consider temporary suspension of EN support until this serious complication has been ruled out. Clinical signs such as increased gastric residue, a rise in intraabdominal pressure to over 15mmHg (particularly when associated to recent oliguria), or sudden worsening of the hemodynamic situation of the patient, should be regarded as possible indicators of intestinal ischemia. Laboratory test measurements such as lactacidemia or leukocytosis can be useful, though they tend to manifest late in the context of intestinal ischemia. Regarding the use of imaging techniques, the presence of dilated and thickened bowel loops on the abdominal X-rays, intestinal pneumatosis, or the presence of air in the portal vein or peritoneal cavity, should cause us to suspect this complication. Techniques such as Doppler ultrasound, magnetic resonance angiography and arteriography are very useful for the diagnosis of occlusive intestinal ischemia, but their role in non-occlusive intestinal ischemia–which is more prevalent in critical patients with hemodynamic instability–is more limited.25 The introduction of new multi detector computed tomography systems may help confirm the diagnosis.26

Mancl and Muzevich27 have recently published a retrospective analysis of 259 critical patients requiring vasoactive drugs and the concomitant administration of EN. Although positive conclusions were drawn regarding the tolerance of EN on the part of the patients, the authors described three cases of mesenteric ischemia/intestinal perforation. This frequency (0.9%) is considered to be similar to that of spontaneous intestinal perforation described in the critical patient receiving EN. The description of these three cases may be useful for knowing the clinical characteristics associated to the development of non-occlusive mesenteric ischemia or intestinal necrosis associated to EN. The study indicates that a specific EN protocol was not applied in all the Units; instead, multidisciplinary management was provided in accordance with the current clinical practice guides. This may have resulted in variability among different professionals in starting nutritional support and in monitoring its efficacy and safety. In fact, substantial differences are noted regarding the way in which EN was provided among the different patients. Likewise, no definition is provided of the different complications or monitored clinical parameters, or of the frequency with which they were monitored. On the other hand, in the Discussion section of their article, the authors pointed out the need to implement a specific EN protocol in their center. As data requiring special consideration, it should be mentioned that in the first described case of non-occlusive intestinal ischemia, full-dose EN was started and was not interrupted upon detecting abdominal bloating. In this regard, it should be remembered that abdominal bloating in these patients is an alarm sign, particularly when accompanied by an intraabdominal pressure of over 15mmHg. This should have led to the suspension of EN. Furthermore, given the previously described high risk circumstances, EN support, in concordance with the previously cited prospective studies,16,17 should be started on a gradual and stepwise basis – thereby allowing closer monitoring of patient tolerance, and its immediate interruption if any alarm sign is detected. Likewise of note in the second described clinical case is the fact that nutritional support was not interrupted despite sudden and significant worsening of the hemodynamic situation of the patient following good initial tolerance of EN.

In conclusion, the diagnosis of non-occlusive intestinal ischemia is complex and must be based on the combination of strong clinical suspicion and the laboratory test results and imaging study findings.28

The current clinical practice guides are not sufficiently clear in addressing the start of EN in critical patients with hemodynamic instability. The guides published by the American Society for Parenteral and Enteral Nutrition,29 and those published by the European Society of Parenteral and Enteral Nutrition,30 state that EN should not be started in hemodynamically unstable patients until complete resuscitation has been achieved and/or the patients are hemodynamically stable. These recommendations are based on the results of non-randomized clinical studies, studies with historic controls, case series and expert opinions. Furthermore, the definition of hemodynamic stability is not addressed in detail –a fact that reduces the practical applicability of the recommendations. The Canadian guides31 in turn recommend the early introduction of EN in the critical patient, with reference to the study published by Khalid et al. in 2010,18 but do not clearly specify the approach to hemodynamically unstable patients.

The guides published by the Spanish Society of Intensive Care and Coronary Units (SEMICYUC) and by the Spanish Society of Parenteral and Enteral Nutrition move one step further in their recommendations compared with the rest of the guides. In effect, they recommend the start of EN in the septic patient32 following resuscitation or, at least, once a state of “stable shock” has been reached, with adequate perfusion pressure (stabilized vasoactive drug doses, stabilized and/or decreasing metabolic acidosis and lactate levels, and a mean arterial pressure of ≥60mmHg). These guides moreover indicate that close monitoring for signs of intestinal intolerance is required. Regarding cardiac patients,33 they describe that when cardiac function is greatly impaired, the introduction of EN is possible, though because of caution due to the risk of intestinal ischemia, it is usually postponed beyond 24–48h. In both cases, the recommendations are based mainly on expert opinions.

Apart from the main clinical practice guides described above, it should be mentioned that the review published by Turza et al.34 in 2009 explored a practical approach to EN support in patients of this kind, following four steps: evaluation of the case history and nutritional antecedents of the patient; evaluation of the current physiological situation of the patient; organization of the logistics for starting EN; and clinical monitoring once EN has been started. Likewise, a very recent review has examined EN support in patients with extracorporeal membrane oxygenation.35

Based on the experience gained from their different studies, the authors of the present review consider it interesting to describe the way in which EN support was started in their post-cardiac surgery patients, together with the monitoring of efficacy and safety, with a view to offering readers some orientation in this respect. The EN protocol of the Department of Intensive Care Medicine of our center was followed, and was required to be known by the medical, nursing and auxiliary personnel. The mentioned protocol detailed the way to start EN, as well as daily monitoring of the different complications associated to EN. The readers are invited to consult the critical care nutritional interventional algorithms published by the SEMICYUC for further information.36

The authors declare that they have no conflicts of interest. The instructions for authors and ethical responsibilities have been taken into account.