Chapter 3.7
Pancreatitis and nutrition
Mary Phillips
Royal Surrey County Hospital, Guildford, UK
3.7.1 Acute pancreatitis
Acute pancreatitis is an inflammatory condition which may be mild or severe. In mild cases there are usually few or no long-term effects, and patients do not routinely require any nutritional intervention. However, in severe acute pancreatitis (SAP) there is a risk of developing systemic issues such as adult respiratory distress syndrome (ARDS) and later localised complications including pancreatic necrosis and pseudocysts [1]. Patients with severe disease often have protracted hospital admissions and require intensive nutritional support.
3.7.2 Causes of pancreatitis
The causes of acute pancreatitis are diverse and are summarised in Box 3.7.1; it is estimated that 80% of cases are caused by gallstones or alcohol.
3.7.3 Severity of acute pancreatitis
Acute pancreatitis is mild in 75% of cases but in the remaining 25%, the disease is classified as severe [2]. Severity is predicted by using a scoring system such as the Ranson Score, Atlanta Classification, Acute Physiology and Chronic Health Evaluation (APACHE) II score or the modified Glasgow or Imrie Criteria (Box 3.7.2). In the latter, a score of 3 or more within 48 h of admission is predictive of SAP. A serum C-reactive protein (CRP) concentration above 150 mg/L may be independently indicative of a severe attack, but this may not become apparent until 48–72 h into the disease process [2].
It is widely accepted that patients predicted to develop SAP require intensive nutritional support. Nutritional assessment, with severity scoring, should occur early in the disease process.
Mild acute pancreatitis
Patients presenting with mild pancreatitis are typically nil by mouth until their pain settles, and they are slowly weaned onto a normal diet over a period of 3–5 days. If the disease aetiology is gallstone related, they will have a laparoscopic cholecystectomy once they are recovered. In others, the cause of the pancreatitis will have to be investigated and treated accordingly.
The European Society of Parenteral and Enteral Nutrition (ESPEN) concluded that there was no evidence that enteral nutrition (EN) administered within the first week of disease onset provided any benefit to patients with mild pancreatitis [3].
Much debate remains over the use of low-fat diets in gallstone pancreatitis, with a paucity of evidence on which to base practice. Where the patient has not had a sphincterotomy and is waiting for a cholecystectomy, it would seem prudent to recommend a low-fat diet until surgery. However, where a sphincterotomy or cholecystectomy has been performed, there appears to be no indication for dietary fat restriction.
Severe acute pancreatitis
Patients with SAP may develop widespread complications including ileus, nausea, vomiting, pain [4], diarrhoea, steatorrhoea (due to exocrine failure), hyperglycaemia (due to endocrine failure), ascites, portal hypertension resulting in gastric varices, and/or the formation of fistulae, pseudocysts and abscesses, all of which affect tolerance of EN and parenteral nutrition (PN).
Patients have increased energy requirements, poor oral intake and reduced nutrient absorption which, in combination with repeated periods of being ‘nil by mouth’ for investigations and procedures, results in rapid deterioration of nutritional status.
Poor nutritional status results in reduced immune function, which impairs ability to resist nosocomial infections, and these often complicate the disease pathway [4]. Furthermore, higher mortality has been demonstrated in patients with SAP who had persistently negative nitrogen balance, compared to those in positive balance [5].
3.7.4 Enteral nutrition
It is well established that the enteral route is the feeding route of choice for patients with SAP. A meta-analysis comparing EN and PN which identified six randomised controlled trials (RCT) involving 263 patients [6] concluded that there were reductions in the incidence of infections, surgical interventions and length of stay in the EN groups compared to the PN groups. However, there was no difference in mortality or non-infective complications.
Early EN reduces GI atrophy and prevents the loss of villi [7]. Gastrointestinal atrophy is associated with the generation of cytokines and other inflammatory mediators, hypoglycaemia, worsening antioxidant stress and the systemic inflammatory response syndrome [8,9]. Additionally, EN is thought to reduce intestinal permeability. High intestinal permeability permits bacterial translocation and endotoxaemia, and this is hypothesised to be the source of many infectious complications of SAP [10].
Studies examining the effect of EN compared to PN on inflammatory markers did not report any benefit, but were limited by small sample size, unequal disease aetiology and severity in each arm, and with an average length of stay of 10 days, were probably not representative of SAP [11].
The priority in nutrition support has moved to establishing an enteral route early in the treatment pathway. A recent systematic review concluded that establishing EN within 48 h of admission was associated with improvements in mortality, infectious complications and multiorgan failure rates [12].
Nasogastric versus nasojejunal feeding
Increasingly, evidence supports early EN in SAP and research has moved to consider the specifics of enteral tube placement. Studies have shown that jejunal feeding at distances of 40–60 cm beyond the ligament of Treitz prevents stimulation of pancreatic enzymes to the same degree as PN, and also stimulates secretion of plasma glucagon-like peptide 1 and plasma peptide YY which inhibit pancreatic function [13–15]. Whilst postpyloric feeding is beneficial in reducing nausea, duodenal feeding continues to stimulate cholecystokinin and therefore pancreatic function [13,15].
Initial studies examining the use of nasogastric (NG) feeding in SAP suggest this is fairly well tolerated and safe, with 23 of 26 patients tolerating full rate feeding within 36 h [16]. These results led to two randomised controlled studies comparing NG and NJ feeding in SAP. In the first study involving 50 patients, NJ feeding tubes were inserted into the proximal jejunum, although the exact position in relation to the ligament of Treitz was not specified [17]. The authors reported one incidence of cardiac arrest during NJ tube placement in one subject. Although the patient made a full recovery, this serves as a reminder that endoscopic placement of NJ feeding tubes is not without risk. The second study examined data on 31 patients randomised to receive NG or NJ feeds and detailed increased overall mortality in the NG group, although this was not statistically significant [18]. However, the methodology stated that the tube was positioned in the third part of the duodenum, resulting in placement proximal to the ligament of Treitz, and thus excluding the study as a true comparison of NG and NJ feeding.
Despite their limitations, these two studies on 81 patients have been the primary citations in two reviews which conclude that NG feeding may be a safe and effective alternative to NJ feeding in SAP [19,20]. Larger, adequately powered studies are required to confirm the effectiveness of NG feeding in clinical practice and in the interim, NJ feeding remains the enteral route of choice.
Feed formulation
A number of studies comparing EN and PN have used standard polymeric feeds, leading to the hypothesis that these feeds may be well tolerated in this patient group. A small RCT carried out in France compared 15 patients receiving a peptide feed with a standard polymeric feed, and concluded that whilst both feeds were tolerated (in terms of patient-reported abdominal pain, diarrhoea, bloating, steatorrhoea and 24-h stool tests quantifying creatorrhoea, steatorrhoea, stool weight and frequency), weight loss and length of stay were both lower in the peptide feeding group (P=0.01 and P=0.006, respectively), suggesting that peptide feeds are associated with better outcomes [21].
Larger studies are required to examine the use of polymeric feeds, whilst peptide feeds are well established as the feed of choice for feeding beyond the ligament of Treitz [3,22]. Further work is required to establish the efficacy of alternative feed types in gastric and duodenal feeding in SAP.
Probiotics
Initial investigations using probiotics in SAP appeared promising. with Lactobacillus plantarum associated with a reduction in disease severity and an improvement in clinical outcome. However, this study compared L. plantarum-supplemented EN with PN and as such, the improvement in clinical outcome may be attributable to the use of the enteral route [23].
In a large, multicentre RCT in 296 patients with SAP receiving EN via an NJ tube, a significantly higher incidence of bowel ischaemia was demonstrated in the patients randomised to the probiotic group, with eight patients in the intervention arm dying as a result [24]. There was no difference in infective complications and a significant increase in mortality in the probiotic group. It must be noted that this trial used a novel probiotic that had not undergone extensive animal or human safety testing; in addition, those patients in the intervention arm were experiencing a more severe attack of pancreatitis at the time of randomisation [25]. A number of other trials were abandoned when these results were published, and although the editors of The Lancet issued an ‘expression of concern’ regarding this trial (published March 2010), this appears to be related to study design and reporting procedures, rather than the data collected. At present, there seems to be no reason to doubt the study’s findings and as such, probiotics are not recommended in patients with SAP.
3.7.5 Parenteral nutrition
Parenteral nutrition carries a higher risk of sepsis than EN, but with improving multidisciplinary management the risks previously attributed to overfeeding and catheter-related sepsis are falling [26,27]. However, PN does not carry the same benefit to outcome in acute pancreatitis and where PN is indicated it should be used alongside EN wherever possible [28].
Care should be taken with the use of PN in the case of triglyceride-induced pancreatitis where it is necessary to use a lipid-free bag if triglyceride concentrations are in excess of 12 mmol/L [28].
In all patients with pancreatitis it is advisable to monitor triglyceride concentrations weekly while PN is ongoing. Additionally, where the aetiology of pancreatitis is unclear, it is important to check triglyceride concentrations prior to commencing PN.