Review question
The intention of this systematic review was to show whether specific enteral nutrition (EN) formulations have any beneficial or harmful effects in the treatment of patients with acute pancreatitis (AP), and whether possible advantages and disadvantages are associated with certain types of EN in comparison with others. Enteral nutrition consists of artificial complete nutrition in liquid form that is absorbed through the intestines.
Review authors conducted searches of available literature until August 2013 to look for studies comparing different types of EN formulations in the treatment of patients with AP. We included only randomised clinical trials in this review, as these studies, if designed and conducted properly, represent the highest methodological standard in clinical research.
Background
Acute pancreatitis is an inflammatory disease of the pancreas - a gland situated in the upper abdominal region that is involved in the process of digestion. The main causes of AP are gallstone disease and excessive alcohol intake. Various factors may activate pancreatic digestive enzymes inside the gland itself, causing tissue damage and extensive inflammation, possibly leading to further damage and resulting in failure of the blood circulatory system, kidneys and/or lungs, and eventually death.
Despite improvements, mortality associated with severe AP is not decreasing, and no specific treatment is available. EN has proved to be more effective than total parenteral nutrition (stopping oral intake with intravenous administration of nutrients) in reducing organ failure, infectious complications and mortality. EN is usually intended to avoid the stomach and is, therefore, given by a feeding tube inserted through the nose, throat and stomach into the middle part of the small intestine. Many types of EN formulations are available; however, no systematic review of evidence has assessed potential benefits or harms of certain formulations over others.
Study characteristics
We included 15 trials with 1376 participants. Two trials included more than two study groups comparing different EN formulations. Six trials compared immunonutrition (EN supplemented with substances potentially able to change the immune response) versus control (other EN, sham treatment (placebo) or no treatment), and six trials investigated EN enriched with probiotics (live bacteria or yeasts that replace or add to helpful bacteria in the gastrointestinal tract). Two trials researched the use of semi-elemental formulations, which are types of EN in which nutrients are broken down to smaller particles. Two trials studied fibre-enriched EN, which may stimulate the growth of intestinal micro-organisms. Only one trial compared immunonutrition enriched with probiotics and fibres versus control.
Key results
Immunonutrition compared with control showed reduction in all-cause mortality. However, when only specific types of EN were compared, this could not be confirmed. Available evidence does not support the effectiveness of probiotics in AP. One trial that made this comparison reported a higher rate of serious adverse events, and consequently more occurrences of organ failure and higher mortality rate. When this trial was excluded, results showed a decrease in mortality, organ dysfunction and pancreatic infectious complications, but with evidence of low to very low quality. Fibre-enriched formulations had a beneficial effect on decreasing local non-infectious complications and shortening hospitalisation. No effects were confirmed for semi-elemental formulations and immunonutrition enriched with probiotics and fibres. These results are inconclusive because of the paucity of data. Comparison of any kind of EN versus no intervention revealed a beneficial effect on all-cause mortality. Overall, EN was associated with a rather small number of mild adverse events (most often nausea, vomiting, bloating, diarrhoea, pain relapse and higher serological concentrations of sodium) not requiring cessation of tube feeding. We cannot be certain that EN is safe in this population because the quality of evidence for adverse event outcomes is low.
Quality of the evidence
All included trials have been assessed as having high risk of bias, most often because they did not provide enough information for adequate assessment of certain study design characteristics, but also because some clear flaws were noted in the way they were designed and carried out. The quality of the evidence throughout this review is considered to be low to very low primarily because of the relatively small numbers of study participants and events included. Study results may reflect systematic and random errors.
We found evidence of low or very low quality for the effects of immunonutrition on efficacy and safety outcomes. The role of supplementation of enteral nutrition with potential immunomodulatory agents remains in question, and further research is required in this area. Studies assessing probiotics yielded inconsistent and almost contrary results, especially regarding safety and adverse events, and their findings do not support the routine use of EN enriched with probiotics in routine clinical practice. However, further research should be carried out to try to determine the potential efficacy or harms of probiotics. Lack of trials reporting on other types of EN assessed and lack of firm evidence regarding their effects suggest that additional randomised clinical trials are needed. The quality of evidence for the effects of any kind of EN on mortality was low, and further studies are likely to have an impact on the finding of improved survival with EN versus no nutritional support. Evidence remains insufficient to support the use of a specific EN formulation.
Acute pancreatitis is a common and potentially lethal disease with increasing incidence. Severe cases are characterised by high mortality, and despite improvements in intensive care management, no specific treatment relevantly improves clinical outcomes of the disease. Meta-analyses suggest that enteral nutrition is more effective than conventional treatment consisting of discontinuation of oral intake with use of total parenteral nutrition. However, no systematic review has compared different enteral nutrition formulations for the treatment of patients with acute pancreatitis.
To assess the beneficial and harmful effects of different enteral nutrition formulations in patients with acute pancreatitis.
We searched the Cochrane Upper Gastrointestinal and Pancreatic Diseases Group Specialised Register of Clinical Trials, the Cochrane Central Register of Controlled Trials (CENTRAL) (2013, Issue 7), MEDLINE (from inception to 20 August 2013), EMBASE (from inception to 2013, week 33) and Science Citation Index–Expanded (from 1990 to August 2013); we conducted full-text searches and applied no restrictions by language or publication status.
We considered randomised clinical trials assessing enteral nutrition in patients with acute pancreatitis. We allowed concomitant interventions if they were received equally by all treatment groups within a trial.
Two review authors independently assessed trials for inclusion and extracted data. We performed the analysis using Review Manager 5 (Review Manager 2013) and both fixed-effect and random-effects models. We expressed results as risk ratios (RRs) for dichotomous data, and as mean differences (MDs) for continuous data, both with 95% confidence intervals (CIs). Analysis was based on an intention-to-treat principle.
We included 15 trials (1376 participants) in this review. We downgraded the quality of evidence for many of our outcomes on the basis of high risk of bias. Low-quality evidence suggests that immunonutrition decreases all-cause mortality (RR 0.49, 95% CI 0.29 to 0.80). The effect of immunonutrition on other outcomes from a subset of the included trials was uncertain. Subgrouping trials by type of enteral nutrition did not explain any variation in effect. We found mainly very low-quality evidence for the effects of probiotics on the main outcomes. One eligible trial in this comparison reported a higher rate of serious adverse events leading to increased organ failure and mortality due to low numbers of events and low risk of bias. When we excluded this study as a post hoc sensitivity analysis, risks of mortality (RR 0.30, 95% CI 0.10 to 0.84), organ failure (RR 0.74, 95% CI 0.59 to 0.92) and local septic complications (RR 0.40, 95% CI 0.22 to 0.72) were lower with probiotics. In one trial assessing immunonutrition with probiotics and fibres, no deaths occurred, but hospital stay was shorter with immunonutrition (MD -5.20 days, 95% CI -8.73 to -1.67). No deaths were reported following semi-elemental enteral nutrition (EN), and the effect on length of hospital stay was small (MD 0.30 days, 95% CI -0.82 to 1.42). Fibre-enriched formulations reduced the number of other local complications (RR 0.52, 95% CI 0.32 to 0.87) and length of hospital stay (MD -9.28 days, 95% CI -13.21 to -5.35) but did not significantly affect all-cause mortality (RR 0.23, 95% CI 0.03 to 1.84) and other outcomes. Very low-quality evidence from the subgroup of trials comparing EN versus no intervention showed a decrease in all-cause mortality with EN (RR 0.50, 95% CI 0.29 to 0.86).