In patients requiring nutritional support who have a normally- or near-normally-functioning gastrointestinal tract enteral nutrition is preferred to parenteral nutrition(Reference Silk, Payne-James, Grimble and Silk1). Parenteral nutrition may often be associated with increased complications and, in critically-ill patients, an increased mortality(Reference Heyland, MacDonald, Keefe and Drover2).
Enteral nutrition is not, however, without problems in some patients. Problems occur, for example, in patients with gastric atony (Table 1) receiving nasogastric feeding. Gastro-oesophageal regurgitation and pulmonary aspiration of the diet is a common complication of nasogastric feeding in patients who are critically ill and ventilated, leading to the development of pneumonia(Reference Heyland, Drover, Macdonald, Novak and Lam3, Reference Heyland, Dhaliwal, Drovere, Gramlich and Dodek4). This route of feeding is poorly tolerated in patients with neuromuscular disorders of the gut(Reference Silk5).
The categorisation of intragastric enteral-feeding complications has pointed clinicians to developing techniques of post-pyloric enteral feeding(Reference Rees, Payne-James, King and Silk6). One of the most common problems with this technique, apart from difficulties of tube placement(Reference Payne-James, Payne-James, Grimble and Silk7), has been that only approximately one-third of positioned tubes remain in situ for the full course of nasoduodenal feeding; the remainder being either inadvertently removed or recoiling back into the stomach as a consequence of retroperistalsis(Reference Rees, Payne-James, King and Silk6). Most recently, therefore, techniques of nasojejunal tube feeding have been developed to overcome these problems and the indications for jejunal enteral feeding defined (Table 2). Long, usually single-lumen, feeding tubes are sited using standard upper gastrointestinal endoscopy-assisted techniques; radiologically using newly-developed guide wires (the author's preferred method of tube placement) or with the assistance of a newly-developed nasoendoscopic technique(Reference O'Keefe, Foody and Gill8). In some patients who go on to require longer-term jejunal feeding nasojejunal tubes can be replaced by directly-sited jejunal tubes, e.g. a needle-catheter jejunostomy tube sited at mini laparotomy or a percutaneously endoscopically-sited tube(Reference Payne-James, Payne-James, Grimble and Silk7). One of the issues that arises when considering nasojenunal and other techniques of jejunal enteral nutrition is what type of enteral diet should be employed, i.e. polymeric or predigested? Clearly, the answer to this question will be dependent on what influence intra-jejunally-administered enteral nutrition will have on the pancreatic secretory response. The present article examines this evidence.
Food-induced pancreatic secretion
The stimulatory effect of feeding on pancreatic exocrine function has been divided into three phases: the cephalic phase; the gastric phase; the intestinal phase. Following the cephalic phase there is strong vagal stimulation when food enters the mouth, is chewed and swallowed, and further vagal stimulation when food enters the stomach. The movement of food plus gastric secretions through the pylorus into the duodenum results in the maximal stimulation of pancreatic secretion mediated by humoral secretion of cholecystokinin and secretin; the combined stimulation being fortified by cholinergic excitation(Reference O'Keefe9).
Enteral feeding-stimulated pancreatic secretin
The stimulatory effect of polymeric and ‘elemental’ enteral diets on pancreatic secretion has been examined with diets either being administered orally or infused at the site of the ampulla of Vater with intestinal contents being aspirated 200 mm distally(Reference O'Keefe, Lee, Anderson, Gennings, Abou-Assi, Clore, Heuman and Chey10). The secretory response to oral ingestion of the polymeric diet was found to be similar to that of the infused diet. However, pancreatic secretion was found to be reduced by 50% in response to ingestion of a predigested elemental diet (l-amino acid N source, 6% energy only derived from fat). It is clear from these basic physiological studies that patients receiving post-pyloric intraduodenal feeding should receive a polymeric enteral diet, the pancreatic secretory response being entirely adequate to result in assimilation of nutrients from the diet.
Inhibition of pancreatic secretion
In 1984 two research groups (one of which was the author's group) independently described an important humorally-mediated braking mechanism whereby jejunal transit and motility is inhibited by the infusion of nutrients in the terminal ileum(Reference Spiller, Trotman, Higgins, Ghatei, Grimble, Lee, Bloom, Misiewicz and Silk11, Reference Read, McFarlane and Kinsman12). This mechanism was termed the ‘ileal brake’ and the hormones enteroglucagon, neurotensin and peptide YY were thought to mediate the braking effects. Later studies have also implicated glucagon-like peptide-1(Reference Van Citters and Lin13). Importantly, later studies of the ileal brake have shown that stimulated pancreatic secretion can be inhibited by the infusion of unabsorbed nutrients into the ileum(Reference Keller, Runzi, Goebell and Layer14, Reference Layer and Host15) and that the inhibitory action is mediated via the release of peptide YY and glucagon-like peptide-1(Reference Wen, Phillips, Sarr, Kost and Holst16). These very important studies led to an investigation of the effect of the site of jejunal feeding on stimulated pancreatic secretion(Reference Vu, van der Veek, Frolich, Souverijn, Biemond, Lamers and Masclee17). Normal subjects received a polymeric enteral diet infused either at the ligament of trietz (proximal jejunum) or 600 mm distally (mid jejunum). Proximal jejunal feeding was found to provide a strong stimulus for lipase, trypsin, amylase and bilirubin secretions. In contrast, feeding at 600 mm down the jejunum was found to have no significant effects on basal secretion rates(Reference Wen, Phillips, Sarr, Kost and Holst16). Subsequently, it has been confirmed in healthy volunteers that there is no significant pancreatic secretory response to the infusion of a predigested diet at distances of 400–1200 mm from the ligament of trietz(Reference Kaushik, Pietraszewski, Hoist and O'Keefe18). This lack of stimulatory exocrine pancreatic response has been shown to occur in association with elevated peptide YY and glucagon-like peptide-1 responses(Reference Kaushik, Pietraszewski, Hoist and O'Keefe18), indicating that the ileal brake is activated during intra-jejunal nutrient infusion. The same research group has developed a technique for measuring trypsin synthetic rates(Reference O'Keefe, Lee, Li, Zhou, Stoll and Dang19) and using this technique have shown that inhibition of pancreatic secretion is the result of inhibition of trypsin synthesis(Reference Zhou, Kiswardy and Kaushik20).
Predigested and polymeric diets during jejunal enteral feeding
For the reasons discussed, post-pyloric duodenal feeding techniques are not the optimum means of overcoming the problems of complications that occur with intragastric enteral nutrition. Properly-sited access beyond the ligament of trietz is the way forward in those difficult cases; certainly, initially and in the short term nasojejunal enteral-tube feeding is the technique of choice. In patients with normal small intestinal motility nasojejunal tubes, once they are correctly positioned, will be moved in a caudad direction by the phase III migrating motor complex(Reference Silk, Payne-James, Grimble and Silk1). It must be assumed therefore that in most patients the diet will be infused approximately 300–400 mm distal to the ligament of trietz. Infusion of polymeric diet at this site will activate the ileal brake, causing release of peptide YY and glucagon-like peptide-1. In turn this response will result in an inhibition of exocrine pancreatic enzyme synthesis and secretion, which is likely to have a most deleterious effect on the digestion and absorption of nutrients from the polymeric diet. Although the evidence points to the fact that the intrajejunal infusion of predigested diets also fails to elicit a pancreatic exocrine secretory response(Reference Kaushik, Pietraszewski, Hoist and O'Keefe18), near optimum absorption of nutrients can occur from suitably-formulated diets in the absence of pancreatic enzymes(Reference Silk, Payne-James, Grimble and Silk1).
On the basis of the evidence discussed it is concluded that properly-formulated predigested enteral feeds are the diets of choice for patients undergoing post-ligament of trietz intrajejunal enteral feeding.
The author declares no conflict of interest.