Controversies in Short Bowel Syndrome

Short bowel syndrome (SBS) is defined as reduced small bowel length that leads to intestinal failure. Intestinal failure (IF) is defined as a condition caused by inadequate intestinal absorption of nutrients, water, or electrolytes, resulting in the inability to maintain hydration and provide sufficient nutrition to support health, growth, and development, needing at least partial parenteral nutrition (PN) for a minimum of 90 days. IF includes motility problems as well. SBS usually is the consequence of intestinal resection or atresia; however, there are limited reports of congenital short bowel. Because patients with IF may not always have reduced bowel length, the two terms are not synonymous. The approaches to care, however, are very similar and are termed intestinal rehabilitation . Patients with SBS generally require specialized nutritional support: first PN, followed by PN plus enteral nutrition (EN), then EN, and ultimately transition to all feedings by mouth, as possible.

This chapter attempts to deal with some of the leading controversies of the present decade affecting the management of this challenging population of patients by neonatologists, pediatric surgeons, pediatric gastroenterologists, nurses, dietitians, therapists, and others.

Controversy 1: Lipid Minimization versus Lipid Modification

Before the development of parenteral lipid formulations 50 years ago, patients had essential fatty acid deficiency, hepatic steatosis, and hyperglycemia. This formulation was composed of soybean oil emulsified in an egg yolk-derived phospholipid layer to simulate enteral fat absorption.

Toxic Effects of Lipids

Although numerous factors play a role in the development of IF-associated liver disease (IFALD), attention has been focused on the contribution of parenteral lipid emulsions to the pathogenesis of IF. The use of parenteral lipids has been associated with development of severe and often life-threatening liver disease. A higher incidence of cholestasis and liver fibrosis has been reported in patients receiving high doses of parenteral lipids (>2 g/kg/day). Lipid overload syndrome has been reported in infants receiving parenteral lipids at doses of >4 g/kg/day, manifesting as coagulopathy, elevated liver enzymes, hepatosplenomegaly, and thrombocytopenia.

The etiology of the potential toxic effect of parenteral lipids remains unclear. Evidence points toward phytosterols, which are plant-derived sterols similar in structure to cholesterol. Phytosterols cause a reduction of bile flow in animal models. Furthermore, phytosterol levels appear to be elevated in children with cholestasis, although it is not clear whether this elevation is the cause or the result of liver disease. Other studies implicate a role of the ω-6 fatty acids, the major components of plant-derived lipid preparations, such as the soybean-based product, Intralipid. The ω-6 fatty acids are generally proinflammatory, and experts speculate that these fatty acids promote hepatic inflammation and injury. A key mechanism of injury has been established in animal models, in which stigmasterol, one of the major sterols in soybean oil emulsion, inhibits farnesoid X receptor (FXR) target genes. FXR is the hepatocyte nuclear receptor for bile acids and mediates cytoprotection by suppression of bile acid uptake, reduction of bile acid synthesis, and enhancement of bile acid efflux through the bile salt excretory protein.

Clinical Approaches to Minimizing Lipid Toxicity

Two evolving lipid management strategies appear to play a role in minimization of IFALD. Lipid minimization is one such approach that may result in prevention of liver disease. Although IFALD has been reported without use of intravenous (IV) lipid, IFALD is more likely to be associated with clinical and histologic complications in individuals receiving IV fat emulsions at doses >1 g/kg/day. Cober and Teitelbaum examined the effects of a lipid minimization strategy in a neonatal intensive care unit population. In their preliminary report, neonates with cholestasis receiving 3 g/kg or more of IV lipid were enrolled in a lipid minimization protocol to receive 1 g/kg of IV lipid twice weekly. A significant negative trend in bilirubin was observed in the lipid reduction group compared with a similar recent historical cohort. One fourth (8 of 31) of these neonates developed biochemical evidence of mild essential fatty acid deficiency (triene-to-tetraene ratio of 0.2-0.5), but all the infants responded to increased lipid administration while maintaining lipid minimization strategies. In a small retrospective study, Colomb et al. also reported normalization of serum bilirubin with the temporary reduction or elimination of IV fat emulsion. Rollins et al., in a prospective trial of surgical infants given 1 g/kg parenteral lipid versus 3 g/kg, also found a slower rate of rise of direct bilirubin. No infant in this study became deficient in essential fatty acids. Whether this beneficial effect of lipid reduction results from reduction of phytosterols (or another component of parenteral lipid preparations) remains unclear. With this strategy of lipid reduction, a higher glucose infusion rate is required for adequate calorie provision in neonates.

A control cohort study by Sanchez et al. compared surgical neonates with use of lipid restricted to 1 g/kg/day and an earlier cohort of infants with use of lipid 3 g/kg/day, and they found a significant reduction of cholestasis 22% versus 43% ( P = 0.002). Calkins et al., however, found no difference when 1 g/kg/day lipid restriction was used.

Most clinicians that adopt lipid restriction use 1 g/kg/day of lipid; the author of this chapter has not seen essential fatty acid deficiency at this dose. There has been concern that there could be some neurodevelopmental sequelae even at this level of lipid restriction. A 2-year study by Ong et al. on neurodevelopment and growth compared infants on 1 g/kg lipid versus the traditional 3 g/kg dose. There was no difference in neurodevelopment and growth outcomes except for a higher 12-month cognitive scaled score in the 1 g/kg group. For background information, in a study on neurodevelopmental and cognitive outcomes in children, 80% of infants and children with IF scored within normal limits, with risk factors for problems being prematurity and repeat operations. However, in the National Institute of Child Health and Human Development retrospective cohort analysis, major surgery in very low–birth weight infants was associated with a 50% increased risk of death or neurodevelopmental impairment.

Another strategy involves use of fish oil lipid emulsions in the management of IFALD. Fish oil or ω-3 fatty acid–based parenteral lipid infusions, such as Omegaven, may have beneficial effects on IFALD. Proponents of Omegaven cite the potential detrimental effects of conventional plant-derived IV lipid emulsions mentioned previously. Gura et al. first reported improvement in two infants with cholestasis after they were given Omegaven. Improvement was subsequently seen in 18 infants with SBS when they were switched to Omegaven after they developed cholestasis compared with 21 historical cohorts receiving soybean oil. Use of fish oil in the latter study was not associated with clinical evidence of essential fatty acid deficiency, hypertriglyceridemia, coagulopathy, infections, or growth delay.

Interpretation of this study is complicated by the fact that the infants received the IV ω-3 fatty acid emulsion at doses lower than those of conventional plant-based lipid preparations. Improvement in this fish oil cohort with liver disease may therefore be the result of lipid reduction as well. Many studies have now demonstrated improvement in infants with cholestasis with use of Omegaven, a 10% lipid solution administered over 12 hours at 1 g/kg/day. In some infants, cholestasis is not resolved despite administration of Omegaven therapy. The effects of Omegaven on liver histology are also unclear. Two reports described infants with persistent portal fibrosis on liver biopsy despite resolution in cholestasis after treatment with ω-3 fatty acid lipid.

Nevertheless, use of Omegaven in doses of 1 g/kg/day appears to improve biochemical disease in most cases. It is still not approved by the U.S. Food and Drug Administration (FDA) and can only be obtained after approval of applications to the FDA, a veterinary application to the U.S. Department of Agriculture, the Institutional Review Board, and an order to the company in Hamburg, Germany. The medication costs $50 to $100 per day per child, and unless special permission is granted for use as an experimental drug, the cost of the drug cannot be billed to the patient, and the institution has to bear the cost.

The strategies of lipid minimization and use of ω-3 fish oil preparations do suggest some role of conventional parenteral lipid preparation in IFALD. Nevertheless, several other nutrients have been implicated over the years in PN-associated liver injury, including amino acids, iron, choline deficiency, and endotoxin. IFALD is likely to have a multifactorial etiology, and IFALD and cholestasis were noted historically before the introduction of parenteral lipids. SMOF, a new lipid combining soy, medium-chain triglyceride (MCT) oil, olive, and fish, has recently been approved for use in adults. It has been shown to be safe and well tolerated in preterm infants and currently is being studied in the United States to obtain FDA approval for pediatrics. It has been hoped that with this formulation, higher doses of IV lipid could be given without the side effect of cholestasis. Recent results of the NEON (Nutritional Evaluation and Optimisation in Neonates) trial, however, have indicated that use of SMOF, at a dose of 3 g/kg, in neonates did not reduce intrahepatic lipid accumulation. Further studies investigating the role of ω-3 fatty acids and lipid reduction are necessary to clarify the optimal lipid strategy.

Controversy 2: Is Septicemia in Short Bowel Syndrome Caused by Bacterial Translocation or by Suboptimal Central Line Care?

Parenteral Nutrition and Epithelial Integrity

The long-term use of central indwelling catheters for the administration of PN places infants at an increased risk for bacteremia and sepsis. Translocation of enteric bacteria through the bowel wall and into the bloodstream, generally referred to as bacterial translocation (BT), may also play a role in septicemia. Clinically, it is not always possible to determine the source of infection (direct catheter contamination versus BT) because the isolation of enteric organisms from central indwelling catheters is not proof of BT. In addition, PN fluids are conducive to microbial growth because of their nutritional components. Nevertheless, there is no direct evidence in humans that PN promotes bacterial overgrowth, impairs neutrophil function, or causes villus atrophy. However, in studies in piglets, a large-animal model that is more similar to humans than rodent models, the switch from enteral nutrition to PN produces atrophy of the villi and a decrease in crypt cell proliferation rate, along with increased crypt and villus cell apoptosis.

The beneficial trophic effect of feeding has been carefully studied in piglets exposed to an isocaloric diet containing 0%, 10%, 20%, 40% 60%, 80%, or 100% of total nutrient intake enterally, with the rest given parenterally. In this model, Burrin et al. demonstrated that intestinal mucosal mass increased when a minimum of 40% of calories were given enterally. Between 60% to 80% of calories were required by the enteral route before normal villus height, brush border enzyme (e.g., disaccharidase) activity, mucosal blood flow, and trophic hormone levels were noted.

Microbial Ingress: Mechanisms and Clinical Evidence

The role of the intestinal tract as a central organ in systemic infections and multiorgan failure was proposed more than 20 years ago. Major mechanisms promoting bacterial translocation include intestinal bacterial overgrowth, deficiencies in host immune defenses, increased intestinal permeability, and damage to the intestinal mucosal barrier. BT has been identified in several diseases in humans (demonstrated by positive mesenteric lymph node cultures). These include burns, intestinal transplantation, hemorrhagic pancreatitis, malignancy, cardiopulmonary bypass, and obstructive jaundice.

Septic complications are linked to the carriage of abnormal microorganisms. Bacterial growth in the bowel is controlled by several mechanisms, including gastric acidity, pancreatic enzyme activity, enterocyte turnover, normal peristaltic activity in the small intestine, and the presence of an ileocecal valve. These factors can be altered in individuals with SBS. Bowel dilation and reduced peristalsis may develop as adaptive mechanisms to improve enteral absorption, but these factors may also promote bacterial overgrowth by reducing the bowel’s ability to expel microorganisms. Moreover, the intestinal endotoxin pool may increase in infants without an ileocecal valve. Increased endotoxin level has been shown to impair liver function affecting the body’s bactericidal activity. Endotoxins, such as lipopolysaccharide, also stimulate Kupffer cells (liver macrophages) to produce increased amounts of inflammatory mediators.

The mucosal barrier (tight junctions) of the intestinal epithelium in patients with SBS is overall intact, even though BT to mesenteric lymph nodes is markedly increased in patients with SBS. The immune response to intestinal bacteria also appears to be altered in SBS. The reasons for this are unclear, but atrophy of gut-associated lymphoid tissue may play some role. Alterations in lymphocyte function may also play a role in susceptibility to translocation. In a mouse model, elimination of enteral feeding resulted in changes in intraepithelial lymphocyte (IEL) profile, with reduced numbers of circulating CD4 ⁺ , CD8 ⁻ helper cells, CD4 ⁺ , CD8 ⁺ cytotoxic T cells, CD8αβ ⁺ thymus-dependent, and CD8 ⁺ , CD44 ⁺ mature IELs. There were also major changes in gut cytokine profile. Therefore bacterial overgrowth and impaired mucosal immunity are factors that place patients with SBS at risk for BT.

Bacterial Profiling

A link between bacterial overgrowth (of aerobic gram-negative bacilli) and septicemia in infants has been shown in some studies. In these studies, the authors found that the episodes of septicemia occurred after the acquisition of specific bacteria. Thus the carriage of abnormal flora increased the risk for septicemia and sepsis, whereas the incidence of septicemia in the infants with normal flora was significantly lower. Potentially pathogenic microorganisms (PPMs) included Klebsiella , Proteus , Morganella , Enterobacter , Citrobacter , Serratia , Acinetobacter species, Pseudomonas aeruginosa, and Candida albicans . The types of bacteria that translocate are mainly aerobic (gram-positive and gram-negative) bacteria. A recent retrospective study reported that infants <1 year of age with IF had an incidence of sepsis of 68%, with most episodes caused by gram-positive organisms, especially Staphylococcus (60%) and Enterococcus (18%). The most common gram-negative isolates were Klebsiella (13 %), Enterobacter (11%), Escherichia coli (10%), and Pseudomonas (5%).

Many clinicians believe that septicemia is not caused by abnormal carriage of microbiota but, instead, by factors related to central venous line care. There could be either a scenario in which PPM colonization of the gut is followed by entry through the skin or one in which PPM colonization of the gut is followed by translocation.

Central line–associated bloodstream infections (CLABSIs) are a frequent and challenging complication occurring in infants with IF. CLABSIs, along with necrotizing enterocolitis (NEC), have been seen to negatively affect somatic growth in infants as both are inflammatory conditions. Preventive measures in pediatrics now include ethanol lock prophylaxis, which has been shown to be well tolerated and to decrease CLABSIs. From the practical standpoint, for neonatology versus pediatrics, most infants will not be able to tolerate a window if the infant is on lipid restriction or is receiving Omegaven because the glucose infusion rate (GIR) becomes too high to provide sufficient calories unless the infants are able to absorb significant enteral feedings. The GIR becomes even higher with time off PN for ethanol instillation. The infants at greatest risk for CLABSIs are often the ones who cannot be enterally fed. Many institutions are using ethanol locks, but not yet for infants under 6 months of age.

What else can we do to address this problem? One commonly accepted approach is to initiate enteral feedings in infants with SBS. Enteral feeding appears to be the single most important factor in restoring gut-related immunity, reducing the incidence of infection, improving intestinal permeability, and enhancing macrophage function.

Controversy 3: What Is Wrong with the Intestinal Microbiota in Short Bowel Syndrome?

Abnormal Microbial Colonization

16S ribosomal RNA (rRNA) gene sequencing enables analysis of the entire microbial community within a sample. This technique has enabled researchers to describe the abnormal colonization of patients with SBS compared with healthy controls.

Lilja et al. analyzed stool from 11 children with SBS, age 1.5 to 7 years. All but one had been premature. The comparison group comprised seven healthy siblings, 2 to 13 years of age. Five were still on PN, all did not have an ileocecal valve, and four were being treated for small bowel bacterial overgrowth (SBBO). The children still on total parenteral nutrition (TPN) were consuming a lactose-free hydrolyzed protein formula and age-appropriate solid foods, with the only modification being a reduction in disaccharide content. The Shannon Diversity Index (SDI) was used to illustrate the variability in diversity. Of the patients with SBS, those still on TPN had the lowest SDI score. Enterobacteriaceae was the predominant taxonomic family in four of five PN-dependent children, and one had a relative abundance of Lactobacillaceae , with Enterobacteriaceae being the second most predominant. These families totally dominated the microbiota of the PN-dependent infants. Overall, Enterobacteriaceae was evident in relatively high amounts in 6 of 11 patients with SBS. Only one of the patients with SBS reached the same SDI score as that of the controls. In the other children with SBS, all off PN, there was more diverse microbiota with more uniform distribution of taxonomic families.

Severe gut dysbiosis has been associated with poor growth in children with IF. A unique gut microbiota signature deficient in Firmicutes (anti-inflammatory Clostridia and Lactobacillus spp.) was observed in children with SBS with poor growth compared with those with good growth by Piper et al. The functional output of bacteria, such as short-chain fatty acids (SCFAs), are now recognized not only as a preferred gut fuel for the colon but also an important source of calories, a stimulant of vascular flow and motility, and a means to increase sodium absorption.

Davidovics et al. showed a clear difference between infants and children with SBS (ages 4 months to 4 years) and healthy controls. They observed relative abundance of species from the phyla Proteobacteria and the class Gammoproteobacteria as well as the class Bacilla, with Escherichia Shigella and Streptococcus being the most notable. Patients with SBS were subcategorized into those experiencing diarrhea and those not experiencing diarrhea. Lactobacillus was in greater abundance in those with diarrhea. All of the SBS group had been treated for SBBO within the previous 6 months, and seven out of nine had been treated with metronidazole.

In some studies, a relative abundance of Gammaproteobacteria was observed in the stools of premature infants who developed necrotizing enterocolitis. A thorough discussion of the pathogenesis of NEC and the relationship to microbiota has been provided by Neu and Pammi.

Lactobacillus has been observed in abundance in adults with SBS as well. The microbiota of these patients with SBS was found to be dominated by lactobacilli, with a subdominant presence and poor diversity of Clostridium species (primarily Clostridium leptum and Clostridium coccoides ) and poor diversity of Bacteroidetes. One species, Lactobacillus mucosae, was detected in the mucosa and feces of patients with SBS but was not found in the stool of any of eight normal volunteers.

Studies primarily in adults showed a marked reduction in fecal concentration of SCFAs, the primary anions of normal stool. The highly acidic stools had lactate concentrations that were found to run as high as 60 mmol/L, in contrast to levels <1 mmol/L in normal volunteers’ stools. In addition, an enormous osmotic gap (the difference between Na ⁺ plus K ⁺ minus Cl ⁻ in stool) was found to be produced by severe malabsorption of osmotically active particles, especially carbohydrates.

Stool cultures indicated a major shift from anaerobes (with virtually undetectable levels of Bacteroides and Clostridium ) to high levels of aerotolerant enterobacteria. At the lower pH levels seen in children with SBS (5.0-5.5), Bacteroides species were uncultivatable from stools of patients with a short gut, whereas lactobacilli predominated. In one study of patients with SBS, the total population of bifidobacteria plus lactobacilli added up to 91% of the total microbial population, whereas the populations of Bacteroides species, Enterobacteriaceae, and Clostridium species were 100-fold to 1000-fold less abundant.

Metabolic Impact of the Altered Microbiota in Short Bowel Syndrome

The importance of the microbiota in patients with SBS is related not only to the tendency of some (Enterobacteriaceae) to invade systemically but also to the powerful metabolic capacity of the microbial community that is able to supply up to 1000 kcal/day. Lactobacilli are of major importance for two reasons. One is that they have a unique tendency to produce lactic acid (both L-lactate and D-lactate, the latter of which is not produced by human tissues). The second reason is the relationship with vitamin B ₁₂ .

A serious complication of adults and children with SBS is the development of D-lactic acidosis, a condition associated with confusion, speech disturbances, a severe metabolic acidosis (with increased anion gap), and sometimes shock. The underlying pathophysiology has been described by Halperin and Kamel. Both D-lactic acid and L-lactic acid are the products of this rapid bacterial metabolism in the colon, which is contingent on the predominant bacterial population. If there is insufficient exposure time, the bacteria have insufficient time to metabolize D- or L-lactic acid to final products, such as acetic acid. Thus acetic acid predominates in normal individuals, whereas lactic acid is a major anion in individuals with SBS. Lactic acid, when fully oxidized, per mole of adenosine triphosphate, yields 70% more hydrogen ions compared with acetic acid. This may help explain the propensity of children with SBS to develop severe perianal dermatitis as well as acidosis.

One other metabolic consequence of the Lactobacillus -dominated flora is vitamin B ₁₂ deficiency. Certain lactobacilli require vitamin B ₁₂ for growth and therefore compete with the human host for its uptake. A 14-year-old with SBS with an intact ileum and an intact ileocecal sphincter developed severe macrocytic anemia and generalized fatigue. He was found to have overwhelming overgrowth with bifidobacteria and lactobacilli and vitamin B ₁₂ deficiency.

It is very rare for patients with SBS to develop bacteremia caused by lactobacilli or bifidobacteria despite their predominance in the gut lumen. Although reports of Lactobacillus species in central blood cultures have been published, reports of bifidobacteremia are not available. “Lactobacillemia” has been reported to occur, but is not common, and in most cases was associated with probiotic administration. In some large series, the three most important infectious agents associated with septicemia in patients with SBS (in order of frequency) were gram-positive cocci, especially coagulase-negative staphylococci, gram-negative rods, and fungi, especially C. albicans. However, clinicians in Houston, Texas, have noted that infections with enteric organisms greatly outnumber infections with skin organisms (e.g., staphylococci, Candida ). During a 1-year period, at the University of Texas, Houston, babies who had SBS and were receiving PN developed 23 central line infections with enteric organisms compared with 14 infections with skin organisms (62% versus 38%). Of note, 17 of 23 episodes were associated with blood cultures positive for gram-negative rods. Similarly, Weber found that 81% of episodes of SBS-associated sepsis were associated with gram-negative rods in enterally fed children.

Controversy 4: Should Small Bowel Bacterial Overgrowth Prophylaxis Be Given?

The ileocecal sphincter provides a mechanical barrier to bacterial migration into the small intestine but also assists in regulating the exit of fluid and nutrients into the colon. Loss of the ileocecal sphincter can lead to small bowel bacterial overgrowth, a condition associated with diarrhea and fat and vitamin (B ₁₂ ) malabsorption, both resulting from bile salt deconjugation, along with fluid loss, abdominal cramps, and liver injury. Furthermore, bacterial overgrowth of the small intestine in human babies with SBS is associated with proximal intestinal inflammation. Whether it is the cause or the effect, children with SBS have nonspecific immune system activation. In one study, there were elevated concentrations of soluble tumor necrosis factor (TNF) receptor-II and interleukin (IL)-6 in urine and serum in patients with SBS and increased TNF-αcompared with healthy controls. Infants with SBBO are at higher risk for bloodstream infections and have higher levels of fecal calprotectin. Cole et al. saw an inverse relationship between percentage of enteral nutrition calories and levels of proinflammatory cytokines, TNF-α, IL-6, IL-8, and IL-ß.

The use of antimotility agents, such as loperamide, is not always recommended in children with SBS because slower transit may exacerbate SBBO. After bacterial overgrowth has been confirmed by intestinal microbiota analysis (duodenal) or hydrogen breath testing, SBBO treatment could be considered. Goulet and Ruemmele recommended the use of intermittent antimicrobial therapy based on oral metronidazole, either alone or in association with trimethoprim-sulfamethoxazole. However, anaerobic organisms are depleted in children with SBS, and aerobic gram-negative rod coverage with medications, such as amoxicillin-clavulanate or ciprofloxacin, may be preferred. The response can be determined by assessing clinical improvement and feeding tolerance. Broad-spectrum antibiotics should be used cautiously, given the risk for emergence of multiresistant strains of bacteria and the effects on colonic bacterial flora. There are no prospective trials comparing the outcome in infants with SBS treated with prophylactic antibiotics versus placebo. The optimal duration and schedule of cyclic antibiotic prophylaxis for preventing SBBO is not standardized across intestinal rehabilitation centers.

The counterpoint to consider when treating SBBO is that reports in the literature have suggested that SBBO treatment may not help and could even make things worse. In the study by Piper et al., five children with SBS treated for SBBO had poorer growth and a lower quantity of a subset of bacteria producing SCFAs, the A _1C that stimulates the gut to make anti-inflammatory cytokines, such as IL-10 and colonic regulatory T cells. These cells can dampen intestinal inflammation. Deficiencies in SCFA-producing bacteria can lead to more inflammation, worse absorption and, with that, poorer growth.

Galloway et al. evaluated anti-flagellin (FLiC) and anti-lipopolysaccharide (LPS) immunoglobulins in infants and children with SBS, about half of them receiving cycled antibiotics for SBBO treatment. Antibodies against LPS and (FLiC) were statistically elevated at baseline in those who received a prophylactic regimen compared with those not on a regimen. In theory, during prophylactic treatment against the gram-positive anaerobes that deconjugate bile acids and are responsible for malabsorption, the genera of bacteria in SBBO are suppressed. There is then the likelihood that the remaining gram-negative aerobes proliferate and consequently translocate across the mucosal barrier into the portal circulation. This increased rate of translocation could lead to more activation of inflammatory pathways and a greater number of bloodstream infections.

Probiotics are microorganisms, which, when given orally in adequate quantities, have health-promoting properties. One might wonder whether prophylaxis against SBBO with probiotics to “out-compete” the enteric flora that tend to produce SBBO would be beneficial. However, because of the reports, albeit rare, of probiotic organisms in the circulating blood of patients with SBS and central lines, currently the recommended treatment of patients with SBS is with probiotics. With further research, we may find out specific strains deficient in infants and children with SBS and gather more information about very specific targeted and effective therapy.

Controversy 5: What Is the Optimal Way to Feed?

Enteral nutrition is the key factor for initiating and maintaining the adaptation of the intestine. Intestinal adaptation, a process by which the remaining intestine increases its ability to absorb nutrients, is highly individualized, encompassing structural and functional changes. Food in the intestinal lumen works directly by providing energy and protein for the developing mucosa and indirectly by stimulating gastrointestinal (GI) hormones that regulate pancreatic, gastric, and intestinal functions.

There is very little literature regarding how infants with SBS should be fed. There are two published studies, one in infants and one in adults, illustrating the benefit of continuous feeding in individuals with short or damaged gut. Parker et al. found benefit in absorption with continuous feedings compared with bolus intermittent feeds in infants. There was noted difference in weight gain (168 g ± 16 g/72 h versus −171 ± 26 g/72 h), and absorption of fat (22 ± 2.0 g/ to 13 ± 0.8 g/24 h), nitrogen (1.7 ± 0.2 to −0.63 ± 0.2 g/24 h), calcium (145 ± 4 to −63 ± 20 mg/24 h), zinc (1.3 ± 0.2 to −0.57 ± 0.2 mg/24 h), and copper (0.21 ± 0.02 to −0.09 ± 0.03 mg/24 h). In adults, Joly et al. found improved absorption using continuous tube feedings (compared with oral feeds) was seen in protein (72% ± 13% versus 57% ± 15%), lipids (69% ± 25% versus 41% ± 27%), and energy (82% ± 12% versus 65% ± 16%). This method can achieve the greatest delivery of nutrients by constant saturation of carrier proteins. On the basis of this information, continuous feedings is recommended as part of the enteral regimen to enhance absorption.

Others feel that bolus feedings are preferable to mimic the gastric filling and emptying in normal feeding. Although bolus feeds are more physiologic, they can present intermittent high, osmotic loads to the intestines, leading to osmotic diarrhea. Bolus feeding has one beneficial physiologic feature: improved gallbladder emptying. Gallstones are a known complication of SBS. For premature infants without gut issues, there are reports supporting a metabolic advantage to continuous feedings. Others have described no difference in growth and macronutrient retention.

Because of the need for normal oral motor development, it would seem important to adopt a combination of both methods, with continuous feeding for enhanced absorption and small-volume bolus feeding to develop normal feeding development. Infants with SBS often develop an oral aversion. Interventions such as as “sham” feedings may allow for earlier oral feeding and may be crucial for the development of normal feeding skills. The skills of nurses, occupational therapists, and speech pathologists are essential to maximize oral motor therapy for these infants.

The addition of solid spoon-fed food at a developmentally appropriate age, usually 4 to 6 months corrected age, is recommended. The Cincinnati Children’s Hospital Intestinal Rehabilitation group has developed recommendations for introduction of solid foods and a daily food guide; the ASPEN’s Pediatric Intestinal Failure group also has developed handouts on fluids and electrolytes and foods for clinicians and for families of infants with SBS.

The composition of the diet can affect how the intestine adapts to feeding. The ideal formula to promote intestinal adaption and to wean the infant from parental nutrition has not been determined. The options that exist for infants are maternal or donor breast milk, premature infant formula with intact proteins, whole-protein formulas, partial hydrolysates (hydrolyzed until the taste changes), completely hydrolyzed formula, and amino acid formulas.

Breast milk is the recommended feeding for all infants, with fortification used if the infant is premature. The use of breast milk in SBS appears to be advantageous, with a retrospective review of infants with SBS showing the benefit of breast milk feedings over an amino acid formula. This benefit may be related to an increase in secretory immunoglobulin A and other immune factors in the breast milk; glutamine/glutamate; and/or growth factors (e.g., epidermal growth factor and transforming growth factor-α). All these factors are important for adaptation and are highly abundant in human milk.

If breast milk is unavailable, donor milk can be used. There are differences between mother’s milk and donor milk, not only because of the lack of specific immune protection from the mother’s exposure that can be conveyed to her infant and differences in the stage of lactation and the age of the baby but also because of the effects of Holder pasteurization.

Based on experience and the findings of small studies, when human milk is not available, use of amino acid–based (elemental) formulas has demonstrated feeding tolerance, which helps increase the chances of weaning the infant off PN. The paper by the Pediatric Intestinal Failure Consortium reported that human milk was given to 19% of infants, and that 20 different formulas were used as the initial diet and 40 different formulas used overall. Clearly, in the period 2000 to 2004, a lot of variation was observed. A recent paper duplicated the same survey, and there is now greater consensus, with more emphasis on the use of human milk.

Although studies have been performed on the choice of formula components, their findings are slightly misleading because those studies were performed in older infants or children. A key factor may be the age at which the products were tried. The intestine changes over time, not only in the premature and term infants but during the time after any insult to the intestine or any resection performed. An algorithm for the best enteral feedings depending on these components may be the best approach to determine what to feed, when, and under which circumstances.

Whole-protein formulas or hydrolyzed formulas provide either full proteins or dipeptides/tripeptides and are thought to confer benefit in terms of enhanced adaptation with optimal paracrine stimulation. The more complex diets may enhance the levels of luminal growth factors, which can influence mucosal growth. Animal studies have shown that a more complex diet is associated with increased signs of adaptation (functional and morphologic) compared with an elemental diet. Hydrolyzed protein formulas have not been shown to be superior to whole-protein formulas when tested at age 4 months.

Although studies indicate benefit of luminal feeding, the optimal diet has yet to be determined. Some studies have demonstrated that early dietary advancement, complex formulas, and the addition of solids (in older patients with SBS) has no adverse effect on adaptation. In some cases of SBS, there is an increased risk for developing colitis, presumably related to increased intestinal permeability, allowing for the development of allergic sensitization. In such cases, an amino acid formula may be beneficial to avoid “SBS colitis.” The author’s center encourages mothers to pump milk, with mother’s milk being the first choice and donor milk the second choice.

The composition of fat in the diet can also affect growth and adaptation. MCTs are often employed to help improve fat absorption and are particularly useful in patients with bile acid or pancreatic insufficiency. The addition of dietary long-chain triglycerides (LCTs; microlipids) was shown to slow gut motility, reduce ostomy output, and perhaps contribute to feeding advancement and weight gain. Formulas with LCTs promote enterocyte proliferation and mucosal adaptation better than those with a high MCT content. However, in patients with significant ileal resection, there may be decreased bile acid formation and therefore impairment of the ability to absorb LCTs. In summary, there is no general consensus, but the author would prefer the use of mother’s milk, with donor human milk as a backup, followed by use of an amino acid formula and then transition over time to a more complex formula and diet as the infant’s clinical condition permits, always keeping in mind the ultimate goal of weaning from PN. Some clinicians do not see any advantage in the use of amino acid or casein hydrolysate formulas and prefer the use of premature infant formulas, even though it slows down enteral advancement. A key concept is that even term enteral formulas have more minerals available compared with those that can be put into PN because of precipitation concerns. Whether the infant can absorb these enteral nutrients is not known, although there is some evidence that calcium absorption continues to improve over time.

In some fortunate cases, the surgeon may have had time to put in a mucous fistula to allow enteral access to the remainder of the gut. Once the distal bowel is healed, the refeeding of the stoma output to the mucous fistula can contribute to the nutrition of the infant, thus decreasing dependence on PN, and sometimes allows complete weaning from PN. It also can make the next surgery easier by preventing mismatch in diameter when the distal bowel has not been used and is very small. There are now numerous reports documenting the advantage of this method that has been applied for years in the author’s center.

Controversy 6: What to Monitor and When?

A useful test for sodium sufficiency in infants and children with SBS is the spot urine sodium assay. Sodium depletion is common in infants without the colon in continuity. Growth can be affected when the urine sodium is <30 mmol/L. This is an easy test that can be used to determine improvement in growth; notable depletion in the urine value will occur long before the serum value indicates a problem. An algorithm is depicted in Fig. 8.1 for management of care.