Introduction

The effective management of intestinal failure remains a major challenge. Fortunately nutritional, medical and surgical advances in conjunction with the evolution of specialized multidisciplinary intestinal rehabilitation programs have been associated with significant improvements in patient outcomes.

Over the past forty years the mortality of pediatric short bowel syndrome has decreased from almost 50 [1] to approximately 10% [2]. This area of pediatric surgery continues to evolve rapidly.

Pediatric intestinal failure may be defined as intrinsic bowel disease resulting in an inability to sustain growth, hydration or electrolyte homeostasis. Classically short bowel syndrome (SBS) is considered a subset of intestinal failure that is the consequence of actual small bowel loss or resection. Mucosal enteropathies (e.g. microvillus inclusion disease, tufting enteropathy) and motility disorders (e.g. chronic intestinal pseudoobstruction) are other etiologies of intestinal failure that do not involve bowel loss [3].

No strict convention exists in pediatric patients as to what length or proportion of residual small intestine constitutes SBS. There is, however, no question that the length of remaining small intestine is highly correlated with the ability to wean from parenteral nutrition (PN) dependence. By custom, small intestinal length is measured along the antimesenteric border with no tension applied to the bowel. In the absence of any surgical bowel lengthening and tapering procedures, historical data would suggest that 35 cm of remnant neonatal small bowel is associated with a 50% probability of weaning from parenteral nutrition [4]. Variability is high and patients with small bowel lengths exceeding this are sometimes unable to transition from PN due to poor motility and/or residual malabsorption. Conversely, neonates with as little as 10 cm of small bowel can at times be weaned from parenteral nutrition. For research purposes, the duration of PN dependence has been used to objectively identify patients with intestinal failure. According to the American Society of Parenteral and Enteral Nutrition (ASPEN) consensus definitions, intestinal failure may be defined by PN dependence for 60 out of 74 consecutive days[5].

The more premature the child, the greater the capacity for intestinal growth and this may be why recent studies have indicated that, when standardized for bowel length, neonates with necrotizing enterocolitis have a more favorable outcome than other etiologies of SBS [6][7]. Although investigations are somewhat conflicting, the presence of an ileocecal valve may be a secondary factor for a more favorable prognosis in SBS. Conceptually, it is important to realize that an ileocecal valve is a marker for residual ileum and that the presence of this highly specialized area for absorption may in fact be an underlying determinant of weaning from PN rather than the presence of the valve per se. Colonic length plays a relatively modest role in the genesis of SBS. The colon is the site of fluid and electrolyte absorption and its nutrient transport role is limited to the transfer of short chain fatty acids. Nonetheless, the prompt closure of enterostomies is associated with improved SBS outcome. Indeed, the reestablishment of bowel continuity sometimes significantly reduces fluid requirements for these patients [4].

see also Intestinal Lengthening Procedures

Content in this topic is referenced in SCORE Short Bowel Syndrome/Intestinal Failure overview

Epidemiology

What diagnoses lead to intestinal failure?

The most common forms of intestinal failure encountered by pediatric surgeons are those with true short bowel syndrome (SBS) rather than mucosal enteropathies or pseudo-obstruction. The major causes of pediatric SBS vary somewhat according to the clinical setting surveyed, however, in a large intestinal rehabilitation program they are the following.

• necrotizing enterocolitis (NEC) 35%
• intestinal atresia 25%
• gastroschisis 18%
• malrotation with volvulus 14% [8]

The remainder of SBS etiologies comprise a compendium of rarer diagnoses such as Hirschsprung disease extending into the small bowel, trauma and vascular pathologies. NEC incidence is highly dependent upon birth weight and those centers with large numbers of extremely low birth weight infants will have a more pronounced preponderance of NEC-associated SBS [9].

What is the incidence of intestinal failure?

A large group of hospitalized neonates, encompassing 16 tertiary care centers in the United States, was found to have an SBS incidence ranging between 0.7 and 1.1%, depending upon birth weight [10]. A population based study from the province of Ontario, Canada, calculated the incidence of SBS to be 24.5 per 100,000 live births [11].

The occurrence of SBS appears to be much higher in infants born less than 37 weeks estimated gestational age as compared to term newborns (353.7/100,000 live births versus 3.5/100,000 live births, respectively) [11].

Pathophysiology

Following massive small bowel resection there are significant fluid and electrolyte losses and alterations in the microbiome.

What molecular or genetic mechanisms play a role in the diagnoses associated with intestinal failure?

The most common conditions accounting for intestinal failure in the pediatric population include necrotizing enterocolitis, gastroschisis, atresias, and midgut volvulus[12]. Other causes in which intestinal length is normal, but absorption and digestion are impaired include intestinal pseudo-obstruction, and various mucosal enteropathies such as tufting or microvillus inclusion disease.

While genetic factors undoubtedly play a role in the pathogenesis of many of these diagnoses, there does not appear to be any common genetic or molecular mechanism to account for intestinal failure. In a small number of cases, a true congenital short bowel syndrome (CSBS) has been described [13]. Although the normal length of the small intestine at birth is approximately 275  cm, patients with CSBS have an average intestinal length of roughly 50 cm. The identification of homozygous and compound heterozygous mutations in Coxsackie and adenovirus receptor-like membrane proteins in CSBS patients has confirmed an autosomal recessive pattern of inheritance in most affected families [14]. More recently, an X-linked mutation in filamin A has been described resulting in a similar gut phenotype in addition to intestinal pseudo-obstruction[15]. Additionally, mutations in the gene coding for tetratricopeptide repeat domain 7A (TTC7A) have been identified as monogenic causes of severe intestinal diseases such as very early onset inflammatory bowel disease, congenital enteropathies, or multiple intestinal atresias leading to intestinal failure[16][17]. Notably, TTC7A mutations have also associated with combined immunodeficiency, which complicates their management and leads to a high mortality rate[17][18].

What is the natural history of patients with intestinal failure?

Following massive small bowel resection (SBR), there is a compensatory response in the remnant bowel termed adaptation [19]. Adaptation after SBR provokes alterations which affect intestinal morphology, the kinetics of cell turnover and overall intestinal function [20]. The bowel increases in length and caliber and there is hyperplasia and hypertrophy of all of the intestinal layers. The mucosal surface area is also increased as villi become taller and crypts deepen. In addition to these morphologic changes, the rates of enterocyte turnover are enhanced as demonstrated by increased proliferation and programmed cell death (apoptosis) [21]. Functional adaptation, as gauged by digestive and absorptive enzyme activity per unit area, is augmented as well.

The significance of adaptation is illustrated clinically in patients who were previously unable to tolerate full enteral feeding but are ultimately able to wean completely from parenteral nutrition (PN) over time. If this adaptation is incomplete, the patient is committed to a lifetime of PN and its associated morbidity.

Multiple mechanisms and mediators have been proposed for the initiation and maintenance of the postresection adaptation response. Indeed, serum levels of multiple peptides, hormones and growth factors have been found to be elevated following massive intestinal resection. The three main factors involve hormones, enteral nutrients and pancreaticobiliary secretions.

After massive SBR, there is an initial period of significant fluid and electrolyte loss. The subsequent initiation of enteral feeding leads to a period of dynamic intestinal adaptation. Finally, a chronic or plateau stage is reached after one to two years. It is during this final phase that the problems of malabsorption, diarrhea, and chronic nutritional deficiency occur if adaptation is incomplete.

Initial postoperative management requires attention to fluid and electrolyte replacement. Large volumes of fluid may be lost from an ostomy and replacement with a solution containing comparable electrolyte concentrations is often required to avoid imbalance. Post-resection hypergastrinemia and gastric hypersecretion occur almost immediately after intestinal resection and can result in massive fluid and electrolyte losses, the potential for the development of peptic ulcers and compromised intestinal absorption. Gastric hypersecretion initially responds well to pharmacologic acid blockade and usually resolves spontaneously.

Diarrhea is inevitable after extensive SBR and is due to multiple factors including reduced absorptive surface area, increased intestinal transit, gastric, small intestinal and colonic hypersecretion, increased osmolality of colonic contents and intestinal bacterial overgrowth. In addition, colonic hypersecretion results from increased spillage of bile acids and fat into the colon where they are converted into secretory stimulants by luminal bacteria. Pancreatic lipase activity is reduced in the setting of a more acidic duodenal lumen which leads to impaired micelle formation and reduced fat absorption.

Unabsorbed carbohydrates entering the colon are metabolized by bacteria into short chain fatty acids and contribute further to diarrhea by a direct osmotic effect as well as by stimulating the colon to secrete fluid and electrolytes. Rarely, children with extensive SBR develop D-lactic acidosis. This occurs secondary to alterations in colonic pH and growth inhibition of Bacteroides species thereby fostering the growth of acid-resistant anaerobes capable of producing D-lactate.

Due to a greater content of luminal free fatty acids which bind and prevent calcium absorption, the bioavailability of dietary oxalate is increased leading to hyperoxaluria and nephrolithiasis.

Mineral deficiencies are common and aggressive supplementation is necessary. Sodium and potassium losses may be substantial and require prompt replacement. Malabsorbed fatty acids form luminal soap complexes with calcium and magnesium. While vitamin D and calcium supplements are usually helpful in preventing calcium deficiency, magnesium deficiency is more difficult to treat as enteral administration of most magnesium salts is often associated with diarrhea.

The clinical response to massive intestinal loss is dependent upon several key features including the length and site of intestinal resection and possibly the presence of an ileocecal valve. Jejunal resection is usually well tolerated since the ileum has the greatest capacity to adapt. Several hormones responsible for inhibiting gastric secretion are mainly produced in the jejunum and therefore jejunectomy is more likely to result in gastric hypersecretion. In contrast, the ileum has a pronounced effect in slowing intestinal transit. Thus, ileal resection generally results in more rapid intestinal transit. The ileum is also an essential site for the absorption and recycling of bile salts. As such, extensive ileal resection is associated with depletion of the bile salt pool leading to a higher incidence of cholelithiasis and fat malabsorption. The colon should be preserved whenever possible in cases of major SBR as the colon not only provides an absorptive surface area but also slows intestinal transit and may passively absorb macronutrients.

What role does the microbiome play in the pathophysiology of intestinal failure?

Small bowel bacterial overgrowth (SBBO) and catheter-related bloodstream infections are two of the most common complications experienced by patients with intestinal failure and directly impact morbidity and mortality [22]. SBBO generally results from the development of dilated loops of intestine with impaired peristalsis. This anatomic alteration sets the stage for stasis, disruption of the enteric flora, secretory diarrhea, malabsorption, gut mucosal inflammation, D-lactic acid production and bacterial translocation into either the portal circulation or mesenteric lymph nodes. Despite these well-known events, data supporting the occurrence of bacterial translocation and microbiological features of SBBO in humans is both limited and indirect. Prior studies have utilized cultures of duodenal aspirates and the absorption of various sugar markers as a surrogate for intestinal permeability. Hydrogen and/or ¹⁴C-D-xylose breath testing has also been used. Reliance on culturable organisms alone is restricted by the fact that 50% of bacterial species in the gut cannot be cultured.

Recent advances in high throughput sequencing of the 16S ribosomal rRNA gene of luminal gut bacteria have established a significant association between the intestinal microbiome and various intestinal epithelial and metabolic responses from a wide spectrum of diseases and conditions. With 16S sequencing, massive SBR has been shown in several animal models to be associated with significant alterations in the gut microbiome[23]. In general, the dysbiosis in intestinal disease tends to be characterized by an increased abundance of pro-inflammatory bacteria such as gram negative Enterobacteriae and a deficiency in anti-inflammatory bacteria such as Clostridia bacteria[23]. The few pediatric clinical studies in the literature have involved small patient numbers. In one report, Lilja, et al., analyzed the gut microbiota in eleven children with SBS and found a reduced bacterial diversity associated with an increased relative abundance of Proteobacteria [24]. A confounding variable of this study was that the majority of patients on PN were receiving antibiotics at the time of stool sampling and six patients had already weaned from PN. In another study, there was reduced bacterial diversity in 23 children with intestinal failure, with a relative abundance of Proteobacteria in patients that required PN contrasted by an overabundance of Lactobacilli in patients that had already weaned from PN [25]. In the PN patients, Proteobacteria were associated with a greater degree of liver injury. These data offer the possibility that the gut microbiome may be a major contributor in the pathogenesis of cholestasis and hepatic injury in patients with intestinal failure. Dysbiosis as a result of small bowel resection has also been associated with poor growth outcomes[26].

Mayeur et al noted a marked dysbiosis in fecal microbiota in sixteen patients with SBS. They demonstrated a predominance of the Lactobacillus/Leuconostoc group while Clostridium and Bacteroides were underrepresented [27]. The presence of fecal lactate (56% of patients) was used to define a lactate accumulator group (LA) while the absence of fecal lactate (44% of patients) defined a non-lactate accumulator group (NLA). The LA group had lower serum HC0₃ levels and were at risk of D-lactate encephalopathic reactions. Furthermore, all patients in the NLA group, and those preferentially accumulating the L-isoform in the LA group, never developed D-acidosis. The D/L fecal lactate ratio may therefore be a relevant index to predict the risk for D-lactate encephalopathy.

The paucity of published data regarding direct interrogation of the microbiota in the setting of intestinal failure represents a significant gap in our understanding of this important morbidity. Closing this gap will direct a more informed scientific rationale for current therapeutic interventions such as antibiotic administration, prebiotics, probiotics, operative reduction in small bowel caliber or even future interventions such as microbiota manipulation via fecal transplantation.

Prevention

The goal of initial management of the patient with possible short bowel syndrome is to minimize intestinal loss. Different operative strategies and techniques can be helpful in this regard including primary bowel lengthening in appropriate patients.

What intraoperative strategies at the initial operation can help to limit the extent of short bowel syndrome?

While it may not be possible to prevent the primary disease process that leads to bowel loss in the infant or child, one of the first goals of treatment is to minimize the extent of bowel injury. These strategies apply primarily to disease processes that cause progressive bowel ischemia rather than anatomic bowel loss (e.g. intestinal atresia) or dysfunctional bowel (e.g. Hirschsprung disease). The operative principle in all cases in which extensive bowel loss is anticipated should be the preservation of as much small and large bowel as possible even if the bowel demonstrates some evidence of ischemia. In general, the surgeon should resect only bowel that is completely necrotic and not amenable to the approaches described below. There are no prospective data directly comparing these techniques to preserve bowel length.

An initial strategy to avoid bowel loss is to avoid laparotomy all together. Although a significant subset of patients require laparotomy, primary peritoneal drainage for necrotizing enterocolitis in the extremely low birth weight premature neonate is a well established treatment option. In primary peritoneal drainage, bowel is not resected but the perforation is drained while aggressive medical therapy continues. Although initially viewed as a bridge to laparotomy, primary peritoneal drainage has increasingly been evaluated as a definitive therapy for necrotizing enterocolitis [28]. The advantage of this approach is that 27 to 64% of these infants have been reported to regain intestinal continuity without requiring salvage laparotomy and resection [29][30][31][32][33].

Another damage control strategy that may be useful in a small subset of patients with extensive ischemia and necrosis of the bowel is the patch, drain and wait approach advocated by Moore and colleagues [34]. With this technique in the setting of severe necrotizing enterocolitis, areas of perforation or significant ischemic necrosis can be patched by imbrication or serosal approximation with an adjacent bowel loop or suture fixation of omentum. Minimal bowel is resected, penrose drains are placed bilaterally for wide drainage of the peritoneum and a gastrostomy may be created. An interval period of several weeks to months follows to allow the inflammatory process to subside and the injured intestine to form a fistula or spontaneous ostomy at the drain site.

Additional strategies to preserve bowel length include the clip and drop approach and reconstruction with proximal diversion. The clip and drop technique involves resecting clearly necrotic segments of bowel but preserving any marginal bowel. Bowel segments are closed with a nonabsorbable suture or a vascular clip to avoid extensive intestinal reconstruction in a critically ill neonate. In most cases, a second look laparotomy is performed within 24 to 48 hours to reassess these segments and plan a more definitive form of intestinal reconstruction. A disadvantage of this approach is that clipping the bowel can result in distention of the bowel lumen which can lead to venous congestion and a paradoxical increase in ischemia.

Proximal diversion can be performed using a loop of relatively healthy bowel located proximal to the diseased segments. Resection of the necrotic distal bowel is performed with loose anastomoses performed to reconstruct segments of perfused bowel. In this setting, the surgeon should err on keeping marginal bowel in place since the fecal stream is diverted proximally. Some surgeons advocate reconstruction of multiple bowel segments using an intraluminal stent such as a feeding tube or small catheter. The disadvantage to this approach in patients with ultra short bowel anatomy is that a short segment of bowel may be lost during ostomy closure. In general, there is little advantage to the creation of multiple enterostomies as this may compromise bowel length; however, a mucus fistula can allow access to the distal bowel for future refeeding of ostomy effluent or contrast study evaluation prior to establishment of intestinal continuity.

Imbrication of the bowel involves suture fixation of the proximal and distal seromuscular layers over an area of advanced ischemia or even partial thickness necrosis. The imbrication is performed along the long axis of the bowel to minimize the risk of stricture formation [35]. The technique of imbrication can be used to preserve bowel that has partial ischemia, but not circumferential necrosis.

Some patients may benefit from temporary abdominal closure in the setting of extensive peritoneal inflammation. If there is any suspicion that the intra-abdominal pressure is high and will affect postoperative bowel perfusion, the abdominal wall can be closed with a temporary dressing; definitive abdominal wall closure can take place days later when the inflammation has resolved and the patient is more stable. Additional advantages of a temporary closure include ease of performing a second look laparotomy and visualization of bowel perfusion directly through the closure material.

What is the role of second look laparotomy?

A second look laparotomy involves re-exploring the abdomen at a defined time interval after the first operation. Usually, the second look laparotomy is performed 24 to 48 hours after the initial laparotomy. This time period is chosen to allow any ongoing ischemia to evolve so that final decisions regarding the viability of marginally perfused bowel can be made. In many cases, some of the techniques to preserve bowel length described above are used in the initial operation with a plan to return to the operating room at an interval date. It is important to make the formal decision for a second look laparotomy at the conclusion of the initial abdominal exploration so that the surgical plan does not change between providers or multidisciplinary care team members.

The second look laparotomy is particularly useful in disease entities that lead to progressive bowel ischemia such as necrotizing enterocolitis and volvulus. Specific new findings at the time of second look laparotomy that may alter clinical management include culture positive abdominal fluid, new perforations in marginally perfused bowel and evolution of ischemic segments into frank necrosis. In many cases, a second look laparotomy can help preserve bowel length in that shorter lengths of bowel are ultimately resected than originally considered at the initial exploration. This finding suggests that the assessment of bowel perfusion and viability at the time of initial laparotomy may not be sensitive and that bowel integrity may need time to declare itself [36].

During second look laparotomy the surgeon must assess for the viability of bowel segments, drain any complex fluid collections and make a formal plan on intestinal reconstruction such as ostomy creation, serosal imbrication or bowel anastomoses. The abdominal wall is usually closed definitively at the conclusion of the second look laparotomy. In some cases, a third look laparotomy may be indicated when bowel ischemia continues to progress or additional data are needed to determine recommendations for continuation of care.

Second look laparotomy can often be challenging in the neonate with necrotizing enterocolitis as these patients may continue to have respiratory and hemodynamic instability in the postoperative setting. In this cohort of critically ill premature infants, second look laparotomy is usually reserved for patients with significant bowel loss that will result in intestinal failure. In some cases of extreme bowel loss, data gathered at the second look laparotomy such as bowel length and perfusion of the liver or stomach, can be used to counsel families about the prognosis of intestinal failure and alternative approaches such as end of life care.

What is the role for primary lengthening and/or tapering procedures at the initial operative intervention?

Many causes of neonatal bowel obstruction, including intestinal atresia, gastroschisis with atresia, and in utero midgut volvulus, can result in a dilated proximal small bowel remnant. The dilated remnant is at risk for poor motility, bacterial overgrowth and a difficult anastomosis given the size discrepancy with the small caliber distal bowel. The options for surgical revision of the dilated small bowel remnant at the initial operation include resection of the most dilated portion, proximal diverting ostomy, tapering enteroplasty and bowel lengthening procedures.

The most straightforward operative technique is resection of the most dilated portion of the proximal small bowel. The surgeon resects the distal aspect of the dilated bowel until a reasonable diameter of the small bowel is reached that more closely approximates the diameter of the distal bowel. While this technique works well it is contraindicated to resect a significant length of bowel in a patient with short bowel anatomy. Therefore, resection should only be considered when it appears that the remaining bowel length will be sufficient for the patient to wean off parenteral nutrition in the early postoperative period.

A proximal diverting enterostomy may allow the dilated small bowel to decompress over several weeks with a plan for interval laparotomy and establishment of intestinal continuity. The disadvantages of a proximal enterostomy include significant fluid and electrolyte losses, the need for parenteral nutrition until the distal bowel can be placed back into continuity and potential additional bowel resection at the time of enterostomy closure. In some cases, even diversion will not allow the small bowel diameter to decompress significantly over time.

Tapering enteroplasty involves resecting the antimesenteric portion of the dilated bowel. The enteroplasty is performed over a large red rubber catheter or chest tube (at least 24 French in diameter) placed in the intestinal lumen to help size the residual bowel [37]. If cautery is utilized, then the bowel lumen is reapproximated using a single or double layer of absorbable suture. After tapering enteroplasty is performed, the proximal bowel is anastomosed to the distal bowel remnant to reestablish intestinal continuity. Tapering enteroplasty has been shown to be safe in the newborn, effectively reduces the diameter of the proximal bowel to better match the diameter of the distal bowel and leads to more effective peristalsis, reducing stasis and bacterial overgrowth. It is commonly utilized in the newborn with a bowel obstruction and dilated proximal bowel. The disadvantages are the long suture or staple line and the fact that a certain amount of mucosal surface area is sacrificed in order to taper the bowel lumen. Some authors have advocated for intestinal plication in which the antimesenteric surface of the bowel is folded into the lumen and a continuous seromuscular suture is used to maintain the plicated bowel [38]. While this latter technique avoids resection of the bowel wall, the long-term effectiveness of the technique is debated.

The serial transverse enteroplasty (STEP) has recently been utilized as a primary bowel lengthening and tapering procedure in the newborn with dilated proximal bowel and associated short bowel anatomy. Case reports have demonstrated its safety in this patient population [39]. Recently, data on primary STEP procedures in the newborn with congenital short bowel anatomy were reviewed from the International STEP Data Registry. This analysis identified fifteen infants who had undergone primary STEP with a relative increase of small bowel length of 50%. The mean bowel diameter after STEP was 1.7 cm. There were no short-term complications from this procedure. In this series, however, only 27% of patients attained enteral autonomy and 69% of patients subsequently redilated their small bowel requiring a repeat STEP procedure. The authors concluded that a primary STEP can be performed safely, but should be limited to select cases of severely short and dilated bowel in an effort to minimize stoma creation [40]. Another analysis of 111 patients enrolled in the international STEP Data Registry demonstrated that 47% of patient attained enteral autonomy following their first STEP, with a median time to reach enteral autonomy of 21 months. Overall post-STEP mortality was 11% in this study [41].

A recent publication describes an alternative to the primary STEP procedure [42]. Wales et al have attempted to leave the dilated proximal bowel intact with resection only of non-viable bowel. The proximal bowel is essentially left obstructed and a gastrostomy tube is placed for venting and sham feedings. In this small series, the authors found a greater than expected increase in the remnant bowel length at the time of re-exploration and performed a delayed primary STEP at approximately four months of age.

The long term outcomes of primary STEP in neonates with congenital short bowel remain to be determined. There remains considerable debate as to the relative merits of performing a primary STEP in the initial newborn period versus a tapering enteroplasty with an interval bowel lengthening procedure at a later date if needed due to inadequate intestinal adaptation and continued dependence on parenteral nutrition.

Classification

An ordered classification of intestinal failure is facilitated if one first considers the question, "Is actual small bowel loss or resection the primary cause?"

Yes

Short Bowel Syndrome

The prime determinant of outcome for short bowel syndrome is residual small bowel length [4]. However, it should be understood that all forms of SBS, and particularly gastroschisis, sometimes have an associated motility disorder. If motility issues persist, antroduodenal and colonic manometry may be useful in further defining the diagnosis and guiding therapy. For the pediatric surgeon, it is also of significance to note that Type IV intestinal atresia may be associated with severe combined immunodeficiency and a TTC7A mutation [43].

No

Mucosal enteropathy

Mucosal enteropathies (e.g. villous inclusion disease, tufting enteropathy) are diagnosed via endoscopic biopsy.

Pseudo-obstruction

Pseudo-obstruction may be further subdivided into neuropathic, myopathic, or combined etiologies. The diagnosis may be confirmed by manometry and genetic testing.

Myopathic pseudo-obstruction is associated with megacystis and the risk of renal compromise. Mitochondrial disease has also been reported in the context of pseudo-obstruction [44].

Assessment

Patients with intestinal failure should undergo a thorough clinical and anatomic assessment. Many of these variables are associated with the ultimate prognosis in children.

What anatomic factors determine prognosis in patients with short bowel syndrome?

The prognosis of children with intestinal failure has improved dramatically over the past few decades. Most young children with intestinal failure will now survive into school age and beyond, even with the shortest bowel lengths.

In many studies, residual small bowel length has been found to correlate with the prospect of achieving enteral autonomy and parenteral nutrition (PN) independence. Bowel length should always be measured and recorded at each laparotomy in children with short bowel anatomy by placing a long suture along the antimesenteric border of the bowel. Prior to the current era of multidisciplinary intestinal rehabilitation it was thought that at least 35 to 40 cm of remaining small bowel was needed to achieve a 50% probability of enteral autonomy [4]. More recent data from the Pediatric Intestinal Failure Consortium (PIFCon) demonstrate that residual small bowel length greater than 41 cm has the greatest combined sensitivity and specificity to predict enteral autonomy [45]. The PIFCon data found that every 1 cm of residual small bowel was associated with a four percent increase in the odds of attaining enteral autonomy. However, there are many reports of patients successfully weaning off parenteral nutrition with shorter bowel lengths.

Therefore, it is likely that bowel length alone is not the only anatomical determinant of long term prognosis in children with intestinal failure. For example, small bowel that is dysfunctional due to poor motility or excessive dilation may not function effectively regardless of its overall length. The specific location of the remaining bowel may also be important to a patient’s prognosis. It is thought that ileum is better suited for intestinal adaptation than the jejunum due to the increased absorptive capacity of the ileum. Several studies have found that the presence of the ileocecal valve increases the likelihood of achieving enteral autonomy as it may help to slow the passage of enteric contents and lead to increased luminal absorption. Alternatively, it may be that the benefit associated with the ileocecal valve actually corresponds to an increased length of remaining ileum in a patient with short bowel syndrome.

In a similar fashion, glucagon like peptide-2 (GLP-2) is produced by the L-cells in the ileum. Exogenous administration of GLP-2 has been shown to improve intestinal adaptation after bowel resection [46][47]. Therefore, it is possible that the benefit of longer remnant ileal length is secondary to increased or preserved GLP-2 production.

The length of the remaining colon likely plays a role in the overall prognosis for enteral autonomy in children with short bowel syndrome. Colon is difficult to measure after surgical resection and standard lengths of colon at different ages are not well established. Since the colon plays an important role in water and electrolyte absorption, a longer colon remnant may lead to enhanced luminal absorption of nutrients and a better overall prospect of weaning off parenteral nutrition.

Finally, it is known that intestine continues to exhibit significant linear growth until term gestation. Therefore, the bowel length measured in a premature neonate (for example, in a patient with necrotizing enterocolitis) may underestimate the actual length of the small bowel remnant. In this way, it is possible that a premature infant with a specific bowel length may have a better prognosis than an older infant or term neonate with the same remaining bowel length. To address this issue, residual bowel length can be used to calculate a percent expected bowel length based on published norms for gestational age-appropriate intestinal length, which takes into consideration the gestational age of the infant at birth [48]. The use of percent expected bowel length rather than absolute residual bowel length is generally preferred when assessing likelihood of attaining enteral autonomy, particularly given the large proportion of intestinal failure patients with a history of prematurity.

What chemical and radiologic parameters determine the prognosis in patients with short bowel syndrome?

There are several nonanatomic factors that may influence treatment options and prognosis in children with intestinal failure. These factors should be considered in the initial and ongoing assessment of a child with short bowel syndrome.

Citrulline is a free amino acid byproduct of glutamine and proline metabolism. It is made almost entirely by the small bowel enterocyte. In both adults and children with intestinal failure, plasma citrulline levels have correlated with small bowel mass and can predict the likelihood of enteral autonomy. In a review of 27 patients with short bowel syndrome, Fitzgibbons et al found that a plasma citrulline level of 15 µmol/L had an 89% sensitivity and 78% specificity of predicting future parenteral nutrition independence [49]. No patient in this study with a plasma citrulline level of 12 µmol/ or less attained enteral autonomy. Baseline measurement of citrulline may therefore be an important prognostic indicator in patients with intestinal failure.

Liver function tests should be checked at baseline and specified intervals during intestinal rehabilitation. At most intestinal rehabilitation programs, liver function is assessed at each clinic visit and during inpatient admissions. Significant elevation of liver transaminases, gamma-glutamyl transferase and conjugated bilirubin are seen in intestinal failure-associated liver disease (IFALD). IFALD can progress to cirrhosis and portal hypertension with subsequent coagulopathy and thrombocytopenia. If IFALD is suspected, a liver ultrasound with Doppler flow should be performed to evaluate for portal hypertension. Data have shown that children who present to intestinal rehabilitation programs with advanced IFALD have a significantly increased risk of mortality.The surgeon should carefully consider the timing of any laparotomy in the setting of progressive liver disease with portal hypertension as coagulopathy and bleeding can occur.

Fluoroscopic contrast studies, including small bowel follow through and contrast enema, can reveal significant bowel dilation, dysmotile bowel segments, abnormal intestinal transit times and anastomotic strictures. These contrast studies are routinely performed in the work up of a new patient with short bowel syndrome or an established patient who develops a significant change in enteral feeding tolerance. Contrast radiographic studies can help guide management in several ways. In a child with a significantly dilated small bowel remnant who has plateaued in their enteral advancement, a bowel lengthening and tapering procedure may be beneficial. Delayed transit through a dilated segment may indicate dysmotility while rapid transit to the colon or rectum indicates a poor functional absorptive capacity of the remnant bowel. Both findings can make advancement of enteral feeds more challenging. Fluoroscopic contrast studies can also help determine the relative length of small and large bowel in a patient who does not have reliable operative data in terms of intestinal length. These measurements however are inferior to operative measurement which should always be performed when possible [50].

Medical Treatment

What immediate medical interventions should be considered following the development of short bowel syndrome?

The management of acutely acquired short bowel syndrome, such as severe necrotizing enterocolitis or malrotation with midgut volvulus, can be considered in terms of immediate interventions, stabilization and long term care. The stages of stabilization (provision of adequate and appropriate nutrition, achievement of euvolemia) and long term care (re-establishment of intestinal continuity, optimizing intestinal adaptation, achievement of enteral autonomy) are covered elsewhere in this topic and others. Treatment in cases of more elective surgical intervention (atresia, long segment Hirschsprung disease) follows the same trajectory, but with more emphasis on the second and third phases of care than the initial critical care management.

The immediate phase of medical management can be considered in terms of three primary principles:

• life-saving intervention
• intestinal preservation
• physiologic balance

Children with necrotizing enterocolitis or intestinal ischemia from midgut volvulus often present in extremis, and can deteriorate while being prepared for the operating room. Verifying a secure and definitive airway, appropriate ventilator management and adequate intravenous access are necessary in this window. In addition, a determination of the best location and method for surgical therapy must be a conscious decision requiring a discussion amongst the surgical, neonatal/pediatric, nursing and respiratory therapy teams. Intensive care units are increasingly utilizing on-site procedure or operative suites to limit the risk of transport to the operating room. In addition, less complex or bridging procedures such as primary peritoneal drainage in perforated necrotizing enterocolitis can be performed to control the abdominal insult until definitive surgical therapy can be safely performed.

In the immediate postoperative period, the paramount consideration must obviously be preservation of life such that standard critical care management strategies including airway stabilization, optimization of ventilation and oxygenation and provision of adequate circulatory support take precedence over other factors. These then carry over into the principle of intestinal preservation, particularly in cases of intestinal ischemia or injury such as necrotizing enterocolitis and volvulus. Thus, circulatory support to maximize splanchnic perfusion is preferred. This often involves significant fluid and blood product administration. Close monitoring for undiagnosed etiologies of poor perfusion (e.g. patent ductus arteriosus, abdominal compartment syndrome) is also important. Dovetailing with this medical support are surgical principles discussed previously such as the use of multiple-look laparotomies and limiting intestinal resection to only clearly necrotic tissue, thereby allow ongoing resuscitation and potential salvage of marginal intestinal segments. Temporary abdominal closure with silos or other similar constructs can allow for decompression of the abdominal cavity as well as constant monitoring of tissue appearance and perfusion. This must be done, however, with meticulous fluid management as copious fluid losses can occur [51][52].

Some specific medical interventions are beneficial in the early postoperative period following significant intestinal resection. Sudden intestinal resection results in a hypersecretory state secondary to hypergastrinemia. This can result in both increased fluid and electrolyte losses via proximal enterostomies and a risk for peptic ulcer formation and massive gastrointestinal hemorrhage. Acid blocker therapy, usually with intravenous histamine receptor blockers or proton pump inhibitors, is therefore an important acute adjunct [51][53].

Finally, the initial period is marked by high fluid losses from profound ileus. Frequent monitoring of fluid and electrolyte status via clinical and laboratory parameters with appropriate adjustment in the content of administered fluid is key to managing this period until fluid loss and repletion are reliably balanced. In addition, adequate nutrient provision, which can only be performed via parenteral nutrition (PN) during this early period, must likewise be achieved to optimize nutrition and replenish energy stores. The combination of these therapies often results in complicated schemes of standing parenteral nutrition which is adjusted daily, maintenance intravenous fluid based on changing electrolyte needs and replacement fluids based on measured outputs (gastric, wound, etc.) [51][54].

What methods exist for parenteral support to optimize nutrition and minimize complications?

Prior to the advent of parenteral nutrition, mortality in patients with intestinal failure was uniform. With the seminal work of Dudrick and Rhoads in the 1960s, survival improved drastically to approximately 70%. While PN initially saved lives, the complications stemming from the need for parenteral nutrition led to the majority of morbidity and mortality. Over the following decades, with improvements in management and standardization of care, the initiation of multidisciplinary centers of excellence and particular attention to limiting the complications of parenteral nutrition, this survival has now improved to well over 90% [3][52].

Parenteral nutrition requires intense monitoring to verify the provision of adequate macronutrients for homeostasis and growth and close evaluation of micronutrients and vitamins. This was especially apparent at times when there were shortages in the international supply of various parenteral micronutrient products. The nutritional aspects of parenteral nutrition provision are covered in Nutrition Medical Treatment.

There are three major complications that can arise from long term parenteral nutrition: vascular complications secondary to the need for central venous access, liver injury (secondary to the administration of parenteral nutrition, lack of enteral nutrition and the high occurrence of septic events) and central venous catheter related blood stream infections.

The need for long term vascular access in small children results in the potential for significant vascular complications. The complete loss of central venous access over time is the dreaded extreme of this issue. Indeed, it can be life threatening in children dependent on parenteral nutrition and is itself an indication for consideration of intestinal transplantation. Management strategies aimed at reducing the risk of vessel thrombosis or the loss of vascular access include catheter salvage during episodes of central line infection and meticulous technique to limit infection (see learning objective on central line associated blood stream infection minimization below). In addition, the provision of parenteral nutrition above a dextrose concentration of 15% should be limited to central lines with their tips located within the superior vena cava or right atrium. Inferior access is generally not ideal due to higher risk of thrombosis and infection, but when required is best used with a catheter tip at the inferior vena cava-right atrium junction [55].

What medical therapies exist to minimize the risk of catheter related infections?

One of the most injurious complications observed in patients with intestinal failure is the development of central line associated blood stream infections (CLABSI). CLABSI in children with intestinal failure can be life threatening as these patients are often among the most fragile patients, have pre-existing metabolic and fluid homeostasis issues, and have a higher risk of infection with more aggressive bacteria (i.e. gram negative rods). The mechanisms of infection include skin exit site infection, introduction of infection secondary to catheter access and bacterial translocation through an abnormal intestinal mucosal barrier. Baseline infection rates among intestinal failure patients are greater than ten per 1000 catheter-days, which is higher than immunocompromised hosts seen in the oncologic setting [55][56][57]. CLABSI has been shown to impart significant morbidity and a high cost of care. Thus, effective strategies for the treatment and prevention of CLABSI are paramount, with the goal of preserving access sites, limiting life-threatening sepsis and decreasing hospitalizations and lengths of stay.

The treatment of CLABSI in the setting of pediatric intestinal failure has become increasingly uniform - in part secondary to consensus guidelines which have recently been updated [58]. The diagnosis is contingent on vigilant evaluation of any child with an indwelling central catheter for fever or other signs of sepsis. Empiric blood culture and treatment with broad spectrum antibiotics is warranted. A general strategy of catheter salvage, in light of the need for preservation of long term central venous access, is utilized. Patients with temporary catheters, fungal sepsis or signs of hemodynamic or physiologic instability warrant immediate catheter removal. In the more stable patient, catheter salvage can be utilized with serial (daily) blood cultures, tailoring of antibiotics as speciation and sensitivities become available and demonstration of the catheter clearance with a prolonged antibiotic course (usually fourteen days) [55][57][58]. Additional adjunct modalities which may optimize catheter salvage include the use of antibiotic locks to treat luminal and extraluminal bacterial biofilm development. Using this strategy, catheter salvage rates reach up to 80%.

With the advent of a checklist strategy, care bundles and pressure from payors to reduce or eliminate iatrogenic infections, prevention strategies have garnered increasing attention in recent years [57][59]. Perhaps the most important of these is the standardization of catheter handling in the hospital. While efforts have concentrated on in-hospital management, similar strategies are being employed by parents administering home parenteral nutrition. Examples include meticulous attention to handwashing, the use of chlorhexidine and/or alcohol based cleansing solutions for catheter dressing changes, appropriate scrubbing of access hubs and intermittent routine replacement of intravenous tubing. For patients with long-term catheters, effective reductions in CLABSI have been noted with the use of polyurethane or silicone catheters rather than polyethylene catheters (all current commercial catheters are now made of polyurethane or silicone), single rather than multiple lumen catheters (one CLABSI avoided for every 20 catheters placed) and the use of tunneled catheters or permanently implantable devices [57].

One of the more exciting interventions which has been recently utilized in the prevention of CLABSI is the ethanol lock. While specific protocols vary by center, a 70% ethanol solution is used as a catheter lock while the patient is being cycled off parenteral nutrition (e.g. four hours a day, three days a week). With one protocol, the use of ethanol locks was shown to drastically reduce catheter infection rates in intestinal failure patients from 9.9 to 2.1 per 1000 catheter-days [56]. There are some reports that ethanol locks can result in a slightly higher risk of catheter breakage. Indeed, an ethanol lock is better suited for silicone-based catheters than polyurethane catheters since they have less risk of molecular dissolution with ethanol exposure. In addition, in vitro reports and some case reports have suggested the precipitation of heparin in the presence of 70% ethanol solution may lead to increased line occlusion rates leading most intestinal rehabilitation programs to avoid the infusion of heparin containing solutions through catheters in which ethanol locks are being employed [60]. Nonetheless, these risks are marginal compared to the significant reduction in catheter infection rates seen with the use of ethanol locks and most intestinal rehabilitation programs have adopted this prevention strategy with welcome success.

Antibiotic locks have been discussed in CLABSI prevention but their utility is limited due to concerns for antibiotic resistance and the narrower therapeutic range [61][62]. Although it is not available in the USA, Taurolidine is used in other countries and has been shown to decrease CLABSI rates without reported adverse events [63]. Tetrasodiumethylene diamine tetra acetic acid (EDTA) 4%, also known as Kitelock, is a non-antibiotic, antimicrobial lock solution that has been shown to be safe in preventing CLABSI in adults and children [64]. It is approved for use in Canada and Europe with studies ongoing in the United States.

What is small bowel bacterial overgrowth?

Small bowel bacterial overgrowth can be a cumbersome complication in patients with intestinal failure and can occur in more than 50% of this patient population [65]. During adaptation, bowel dilation leads to disordered motility. Stasis of intestinal contents results in an ideal environment for overgrowth of resident luminal bacteria. Bacterial overgrowth leads to inflammation, which in turn can lead to mucosal injury, gastrointestinal bleeding and bacterial translocation. In addition, the bacteria themselves can increase bile acid deconjugation leading to further malabsorption and can produce toxic byproducts such as D-lactic acid that lead to a specific syndrome of ataxia, mental status changes, and acidosis [65][66]. Intestinal failure patients with new feeding difficulties, a plateau in feeding advancement, worsening dysmotility or the development of abdominal distention, pain and/or gastrointestinal bleeding potentially suffer from bacterial overgrowth. A diagnosis of bacterial overgrowth is objectively made in one of two ways: demonstration of an excess of bacteria >105 CFU/mL in duodenal aspirate fluid or a glucose or lactose hydrogen breath test with either a double peak or a peak >20 parts per million hydrogen above basal within 90 minutes of laculose ingestion [5]. In the absence of objective data, symptoms such as those noted previously can be indicative of bacterial overgrowth, particularly if the symptoms respond to a course of empiric antibiotic therapy. Typical treatment is empiric with the administration of an enteral antibiotic course for one week. Broad spectrum antibiotics effective against anaerobes and Gram negative bacteria are utilized. Commonly used antibiotics in our center include metronidazole, ciprofloxacin, and the intravenous form of gentamicin administered enterally, since this typically has minimal systemic absorption. The use of probiotic formulations to prevent or treat small bowel bacterial overgrowth has been reported; however intestinal failure patients with indwelling central venous catheters have had devastating complications of CLABSI traced back to the probiotic microbe being administered [67]. As a result, we advocate against the use of probiotics in this patient population.

Endoscopy with quantitative duodenal aspirate cultures can be used to help guide therapy against bacteria that may be responsible for the patient’s symptoms. These cultures and sensitivities allow for tailored therapy. Even in the absence of quantitative cultures, the appearance of blunted villi and subacute inflammation on biopsies can be indicative of small bowel bacterial overgrowth in patients with appropriate clinical symptoms. Endoscopic intestinal biopsies can also help in demonstrating other pathology such as allergic enteritis or colitis, marginal ulceration or stricturing at surgical anastomoses or more generalized intestinal inflammation which may benefit from anti-inflammatory therapy. In one single center study of intestinal failure patients undergoing endoscopy, 24 of 27 patients were found to have abnormalities on gross examination, duodenal culture or microscopic examination of intestinal biopsies. Of these, twenty patients had a change in management as a result of the endoscopic findings [68].

Surgical interventions in the form of resection or tapering operations (STEP for D-lactic acidosis) have been used for the treatment of small bowel bacterial overgrowth and its complications with some success. They serve as additional therapeutic options in patients who fail more conventional maneuvers [69].

What hormonal agents have been utilized in intestinal failure?

There are a plethora of hormones, growth factors and peptides that have been investigated as a means to enhance intestinal adaptation responses in various animal models. These include growth hormone (GH), insulin-like growth factors 1 and 2, insulin, epidermal growth factor, hepatocyte growth factor and glucagon-like peptide 2 (GLP-2) [70]. In general, the intestinotrophic effects of these growth factors result in stimulation of enterocyte proliferation, attenuation of enterocyte apoptosis, villus elongation, crypt deepening and enhanced absorption and digestion of luminal nutrient. The most widely evaluated hormones in adult humans have been GH and GLP-2.

Growth hormone was one of the first hormones that demonstrated an enhanced ability to wean from PN in humans [71]. Multiple subsequent reports have supported this finding [70]. Patients who benefited the most (i.e. complete discontinuation of PN) appear to be those with 70 to 100 cm of remnant small bowel and without an intact colon. Patients with extreme SBS (less than 70 cm) had less success in achieving autonomy from PN.

GLP-2 is another well studied intestinotrophic factor. This growth factor is synthesized in enteroendocrine L-cells of the distal ileum and proximal colon. GLP-2 exerts its effects through the GLP-2 receptor which has been identified on intestinal enteroendocrine cells, enteric neurons and subepithelial myofibroblasts. Substitution of glycine for alanine at position 2 within the 33 aminoacid peptide creates a synthetic analog of GLP-2 (Teduglutide, Takeda Pharmaceuticals U.S.A., Inc.) that is resistant to enzymatic degradation and significantly extends its half-life. Teduglutide has been demonstrated in two randomized, controlled trials to significantly reduce the need for parenteral nutrition in pediatric patients [46][47]. The trials showed that the majority of patients (69%) on 0.05 mg/kg subcutaneous Teduglutide daily were able to wean their parenteral nutrition requirements by at least 20% over the 24-week study period. 11.5% of patients on this dose achieved enteral autonomy during the study period. The safety profile was similar to that previously reported in adults. After the results of these trials showed such a favorable response to Teduglutide therapy, it gained FDA approval in 2019 for use in pediatric patients over 1 year of age. Teduglutide currently requires daily injections, but there are ongoing trials of other GLP-2 analogues (Glepaglutide from Zealand Pharma and Apraglutide from VectivBio) that may be longer lasting necessitating fewer injections [72][73]. A combination GLP-1 and 2 co-agonist is also currently being tested in animals with the theory that native enteroendocrine hormones work in concert with each other leading to an additive effect [74]. As with any growth factor that stimulates cell proliferation, there is a concern for possible development of neoplasia as a consequence of administration of Teduglutide; however, this risk has not yet been demonstrated in the adult follow up studies to date. Given this theoretical risk, patients that are on Teduglutide require close follow up with serial screening endoscopies and labs to diligently monitor for any developing neoplasias.

Medical Decision Making

How should enteral nutrition be titrated?

After significant bowel resection the shortened intestine must undergo structural and functional adaptation to improve the nutrient absorptive capacity of the remnant bowel. This process, however, is optimized through stimulation with enteral feeding. It appears that a combination of mechanical, luminal and humoral factors contribute to this process marked histologically by an increase in villus height and crypt depth, crypt epithelial cell hyperplasia, bowel lengthening (marginally), and bowel dilation [75][76].

The provision of enteral nutrients seems to be key to the initiation and optimization of adaptation. It also serves a significant protective role in preventing or limiting intestinal failure associated liver disease. For this reason, the attention of the health care team must be focused on the appropriate timing of initiation and advancement of enteral feeding in patients with intestinal failure.

Following the initial period of postoperative ileus, the return of bowel function should mark the beginning of feeding attempts in the absence of other contraindications to feeding. These include concern for anatomic obstruction (e.g. stricture), pronounced hemodynamic instability or ongoing concerns for intestinal ischemia. Given the significant variation in the etiology of intestinal failure, underlying congenital abnormalities and prematurity, the ability to initiate feeding will vary from patient to patient.

Several decisions occur concurrently when feeding is to be initiated. First, the formulation of the diet needs to be clarified. In most cases, breast milk is chosen preferentially because its growth factors and immunoglobulins may aid intestinal adaptation. Breast milk may also be protective against intestinal failure associated liver disease. In cases where breast milk is not available, donor breast milk may be chosen if the patient meets institution-specific criteria; otherwise a formula must be used. In severe intestinal failure, amino acid based elemental formulas have been associated with improved outcomes [4][77].

The second decision involves the route of administration. In severe cases, most practitioners would choose continuous feeding over bolus feeding. Despite the developmental benefits to bolus feeding, which more closely mimics normal gastrointestinal physiology, the primary goal should be weaning parenteral nutrition and the achievement of enteral independence. In most cases of short bowel syndrome, this is best accomplished with continuous tube feeding which allows constant saturation of nutrient receptors and maximizes intestinal absorptive capabilities. Continuous feeding has been shown to improve intestinal nutrient absorption, weight gain and overall enteral tolerance in patients with intestinal failure [77].

Once feeding is initiated, monitoring for tolerance and determining rates of advancement are based on clinical factors. In the absence of symptoms indicating ongoing ileus or obstruction, the primary determinant of feeding advancement is stool or stoma output. A feeding advancement guideline for neonates and infants with new short bowel syndrome is provided (guidelines for enteral feeding) as an example [77]. Note that while this practice guideline contains assessment of reducing substances to monitor for malabsorption and gastric residuals to monitor gastric emptying and feeding tolerance, these are no longer standards in our daily care but are adjuncts which may be used on a case-by-case basis. The fundamental tenet of the guideline is the advancement of feeds by 10 mL/kg/day only once daily with close monitoring of stool or stoma output and an assessment of the hydration status. Parenteral nutrition (PN) is simultaneously weaned as feeding is advanced to maintain appropriate weight gain, including catch up goals. If feeding tolerance plateaus, caloric fortification may allow for additional enteral calories. In addition, adjuncts such as loperamide and cholestyramine may be used with caution to reduce stool output.

In concert with feeding advancement, age appropriate initiation of oral feedings may be attempted as per the practice guideline. These attempts should be methodical and should not occur at the expense of ongoing feeding advancement. Furthermore, the assistance of a speech and feeding therapist is often required due to the development of oral aversion. Weaning from parenteral nutrition and achieving enteral independence should remain the primary goal.

What lipid formulations are available?

see Nutrition Complications

The first commercially available intravenous fat emulsion (IVFE) was composed primarily of cottonseed oil (1956). Its use was associated with severe hepatotoxicity, jaundice and fevers and it was soon withdrawn. An IVFE made from soybean oil was subsequently developed and became the most commonly used product in the United States (Intralipid®, Fresenius Kabi). Soybean oil based IVFE’s are composed primarily of linoleic acid (approximately half of calories) with significant caloric contributions from α-linolenic acid (nine percent), thus making essential fatty acid deficiency (EFAD) unlikely at the standard pediatric dosing of 3 g/kg/day.

Intralipid composition

Essential fatty acid deficiency is a concern in this vulnerable population, occurring as early as a few weeks after lipids have been aggressively restricted or removed entirely, due to the association of docosahexanenoic (DHA) and eicosapentaenoic (EPA) acids with normal early brain development. The predominance of proinflammatory N-6 polyunsaturated fatty acids (PUFA) in soybean oil based formulations, however, has contributed to concern that they play a direct and deleterious role in the development of intestinal failure-associated liver disease (IFALD). Other characteristics of soybean oil-based IVFE which are thought to contribute to IFALD include the presence of phytosterols and a relative lack of α-tocopherols [78].

In an effort to create a more balanced and optimal lipid emulsion, subsequent formulations have decreased the component of N-6 fatty acids in favor of combinations of N-3 polyunsaturated fatty acids, medium chain triglycerides (MCT), olive oil, fish oil and α-tocopherol. MCTs contain saturated fatty acids and, therefore, are resistant to the oxidative stress often observed in critically ill patients. IVFE containing olive oil in combination with soybean oil were developed to decrease the amount of PUFAs while maintaining linoleic and α-linolenic acid levels. Small trials from Europe have shown that a soybean and olive oil product (Clinolipid®, Baxter) is safe and effective in children, but it remains FDA-approved for adult use only [79].

Clinolipid composition

A lipid emulsion composed solely of fish oil (Omegaven®, Fresenius Kabi) is composed almost entirely of anti-inflammatory N-3 PUFAs and has shown promise in its ability to reverse the biochemical signs of cholestasis. It is currently recommended to be used as a rescue therapy for IFALD and has been approved by the FDA for use in children. There have been case reports of reversal of cholestasis in patients with IFALD who were switched from SMOFlipid to fish oil monotherapy [80].

Omegaven composition

The last alternative lipid formulation is composed of soybean, olive and fish oils as well as MCTs (SMOFlipid®, Fresenius Kabi) and is perhaps the most physiologic of all the products discussed thus far. It contains N-3, N-6 and N-9 fatty acids as well as substantial amounts of MCTs and α-tocopherol [81]. SMOFlipid was approved by the FDA for use in adults in 2016, although it is widely used off-label for pediatric intestinal failure patients. It has been demonstrated to be safe and effective in children and infants[82], and was associated with a decreased serum bilirubin in a randomized controlled trial [83].

SMOF composition

A variety of studies have been published in an attempt to answer the question of which lipid product to use and at which dose. It is clear that soybean based IVFE dosed at three g/kg/day is associated with a significant incidence of IFALD which, at its worst, leads to cirrhosis and liver failure requiring transplantation. In an effort to decrease this risk, many institutions use a lipid restriction strategy to limit the lipid dose to no more than one g/kg/day. Data from these centers has demonstrated that lipid restriction can reverse cholestasis once it has developed. While early restriction has been associated with the development of mild EFAD, a slight increase in the dose reversed this without adverse sequelae. Lipid restriction has now been shown to prevent the development of cholestasis without affecting somatic growth or essential fatty acid levels [84]; many intestinal failure programs empirically use this dosing scheme in neonates anticipated to need long term PN [85]. Other centers have pioneered the use of fish oil based IVFE at a dose of one g/kg/day; these patients maintain their somatic growth while avoiding EFAD. A recent study comparing equivalent doses (one g/kg/day) of soybean based with fish oil based IVFE’s demonstrated no differences in the maximum or median levels of conjugated bilirubin. Patients also maintained somatic growth and experienced no EFAD [86]. Importantly, many of the studies were relatively short in duration and not all assessed for adequacy of somatic growth or the development of EFAD.

Because of the heterogeneity of these studies, questions remain regarding the optimal lipid dose and formulation to prevent IFALD and EFAD while providing adequate growth. Due to this uncertainty, it is recommended that the levels of linoleic, alpha-linolenic, and mead acid in addition to the triene to tetraene ratio all be considered when evaluating these patients for essential fatty acid deficiency [87]. A wide variety of lipid products exist, each with advantages and disadvantages, although the only IVFE FDA approved for use in children is a soybean oil based product. With this limitation in mind, lipid restriction may be the safest and most feasible option to prevent and manage IFALD in PN dependent patients. While a number of centers in the United States have adopted the use of fish oil based IVFE when cholestasis develops, these protocols also use a lipid restriction strategy thus raising the question of whether the observed reduction in PNALD is due to the lipid formulation itself or the dose administered. Further investigations are warranted to determine the optimal lipid product and dose that will minimize the development of IFALD while avoiding EFAD and its potential downstream effects on neurodevelopmental outcome.

Indications for Surgery

How often do children with intestinal failure require operative intervention? What technical considerations should be given in repeated operative intervention?

Intestinal failure has evolved from a diagnosis with certain mortality to a diagnosis with long-term survival over 90% since the advent of parenteral nutrition, the institution of dedicated nutrition services and the advent of interdisciplinary care. As a result, mortality in this patient population has been replaced by morbidities in the form of long-term complications, high cost, and a remarkable burden of care.[88][89]

In this light, one of these morbidities for intestinal failure patients is the surgical burden they endure. While it is clear-cut that children with short bowel syndrome as a result of surgical resection will require initial surgical operation(s) to correct their emergent pathology, what is now becoming clear is that these children continue to require additional abdominal surgical intervention.

One study, evaluating patients from 14 intestinal failure centers around North America, described the marked surgical burden of these patients. In 272 patients with intestinal failure over nearly 3 years of follow-up during a cohort observation period, the median number of abdominal procedures was 4, with an interquartile range of 3 to 6 and an absolute range of 1 to all the way up to 15 abdominal procedures for one patient. Notably, these procedures included exploratory laparotomies and bowel resections as would be expected, but also a large number of procedures for enteral access management, ostomy creation/revision/closure and autologous intestinal reconstruction.[89]

Multivariable analysis of the variables associated with this high surgical burden demonstrated that higher surgical burden was not correlated with longer parenteral nutrition (PN) duration, transplantation or death. However, children with higher surgical burden sustained more sepsis episodes and were less likely to be cared for at a center with transplant capability. In addition, there were no specific patient factors associated with higher surgical burden overall, though patients with necrotizing enterocolitis (NEC) were statistically more likely to undergo bowel resection than non-NEC patients.

Overall, the understanding that intestinal failure patients sustain a high surgical burden is paramount in planning their care, properly managing parental and care team expectations and in operative planning. The surgeon caring for these children must be well-versed in repeat laparotomy, with great care taken to minimize unnecessary bowel resection, limit scar formation and manage complications.

When and what type of intestinal lengthening and tapering procedures should be used for children with intestinal failure?

The two most commonly used Autologous Intestinal Reconstruction Surgery (AIRS) procedures that have been used include the longitudinal intestinal lengthening procedure (LILT) and the serial transverse enteroplasty (STEP) procedures. Both of these procedures are described in the "Intestinal Failure Procedures" topic.

While LILT remains a part of the pediatric surgeon’s arsenal, the STEP procedure has become the procedure of choice for the surgical treatment of intestinal failure. Since its initial description in 2003, the procedure has been noted to have three clear cut potential applications which have been borne out in subsequent analysis. These include the lengthening of the patient with short bowel syndrome and failure to progress in terms of enteral independence, the child with isolated dilation and significant bacterial overgrowth and the newborn with congenital SBS and bowel dilation secondary to atresia [90][91].

The prerequisites for performance of the STEP procedure are intestinal failure and bowel dilation amenable to tapering and lengthening. The dilation, unlike for LILT, does not need to be uniform as each individual application of the stapling device can be tailored to the desired new intestinal channel diameter [90]. In addition, unlike with LILT, the STEP can be performed repeatedly as needed, though repeated STEP procedures likely have diminishing returns [41].

In general practice for most surgeons treating patients at major intestinal failure centers, patients with short bowel syndrome, ongoing parenteral nutrition dependence and a plateau in the advancement of enteral nutrition should be considered for STEP or other autologous intestinal reconstruction if they do not meet standard criteria for transplantation. Preoperatively, medical management (treatment of bacterial overgrowth, non-specific inflammation, dysmotility, malabsorption) should be optimized and contrast studies should be performed to verify adequate intestinal dilation and the absence of stricture or other anatomic anomalies. Post-operatively, patients should continue to be followed closely as AIRS can have delayed complications such as staple line bleeding and bowel re-dilation [92][93].

When and how should a transplant evaluation be performed in an intestinal failure patient?

The indications for referral to an intestinal transplant program were originally adopted in 2001 by CMS to include

• impending or overt liver failure due to PN-induced liver injury
• thrombosis of two or more central veins
• a single episode of line-related fungemia, septic shock or acute respiratory distress syndrome
• frequent episodes of severe dehydration despite intravenous fluid supplementation in addition to PN [94]

These criteria have been recently challenged given the significant improvements that have occurred in the management of intestinal failure over the past two decades. More recent data suggests that patients may have better long-term survival with prolonged PN management than with early intestinal transplantation. In one single center study, Burghardt et al found that three new criteria had a high predictive value for the need for intestinal transplantation.

• two or more intensive care unit admissions
• persistent direct bilirubin of greater than 75 mmol/L (4.4 mg/dL) despite optimized lipid strategies
• loss of more than three central venous access sites [95]

When two of the three criteria were present there was a 98% probability of needing an intestinal transplant. Despite the recent improvements in care, early referral to a transplant center is still indicated when a patient is believed to be approaching these milestones as time is necessary for evaluation and listing to be completed. In addition, wait list times vary widely depending on patient size, blood type and organs needed so one cannot assume that a transplant can be performed within a short time after referral.

Once a patient is referred for intestinal transplant evaluation, a review of records and studies must be completed to determine the indication for transplant and to rule out any contraindications to transplant. This involves an interdisciplinary group including hepatology, gastroenterology, surgery, nutrition, social work, pharmacy, infectious disease, financial services coordinator and any other services or medical specialties that are deemed necessary to optimize the care of the patient. From a surgical standpoint, one must understand the history of operative interventions that have led to the current anatomy to develop an appropriate surgical plan for replacement of the appropriate organs. Once the evaluation is completed and the patient is deemed to be an appropriate candidate for intestinal transplantation, they may be listed but given the uncertainty of wait list time, a plan must be developed with the referring intestinal failure team for seamless ongoing management of the patient to avoid unexpected development of complications that may be contraindications to transplantation such as uncontrolled sepsis or a complete loss of intravenous access.

When do both intestines and liver, versus one or the other, get transplanted in the intestinal failure patient?

During the evaluation of a patient with intestinal failure for possible transplantation, the surgeon must decide which organs to include in the transplant. The typical intestinal failure patient will be stable on PN with permanent intestinal failure and essentially no chance of attaining full enteral autonomy. This may be due to either a physical lack of sufficient bowel absorptive surface area (i.e. extreme short bowel syndrome) or a functional lack of bowel absorptive surface area (i.e. motility disorder) or a combination of the two (i.e. gastroschisis). In this case, the patient will definitely require an intestinal transplant to have a chance of gaining enteral autonomy and the decision to include additional organs like liver or stomach will depend on the function of those other organs.

Generally, if there are any signs of decompensated liver disease including coagulopathy, ascites, persistent jaundice or signs of portal hypertension including thrombocytopenia or gastroinestinal bleeding, a liver should be included in the graft (liver-intestine or multivisceral transplant). The degree of liver disease requiring a liver inclusive graft varies from center to center and a liver biopsy may be useful to help guide this decision. In cases of abdominal catastrophe or stomach dysmotility the stomach may also be included. There are some cases where the liver is in good condition but the stomach is required and these cases should undergo a modified multivisceral transplant (intestine, pancreas and stomach).

The final situation is those cases where the patient has made good progress towards enteral autonomy (greater than 50%) but the liver starts to show signs of decompensation that limit any additional progress. If the anatomic situation is such that the intestinal failure team believes there is enough intestine to achieve full enteral autonomy, the transplant team may elect to proceed with an isolated liver transplant with the hope that the patient can continue to wean from PN following elimination of liver disease and portal hypertension. Given the significantly better survival of isolated liver transplant versus any type of intestine transplant, this option may be worth the risk in select cases. If the patient eventually proves to be recalcitrant to PN weaning, an isolated intestine transplant may be performed at a later date.

Surgical Decision Making

Intestinal continuity should be established early. Patients with intestinal failure have unique vascular and enteral access needs. Decision making regarding operations to increase bowel length are complex.

When should intestinal continuity be re-established?

Stomas are common in infants and children with intestinal failure and predominate among those who have short bowel syndrome due to necrotizing enterocolitis. While creation of an enterostomy can often result in an earlier return to enteral feeding, stomas are associated with significant complications such as prolapse, retraction, obstruction, parastomal hernia and skin breakdown. Several reports in the literature cite overall complication rates of 24 to 68%, with strictures, wound infections and stoma retraction or prolapse being the most common morbidities [96][97]. More proximal stomas can be associated with significant fluid loss and an inability to initiate or advance enteral feeds. While stoma formation remains a standard of care, particularly in neonates who are low birth weight or extremely premature, primary resection and anastomosis can be considered in larger neonates without significant comorbidities or whose respiratory status is not substantially compromised.

The optimal timing of stoma reversal has been debated for decades with some practitioners advocating early take down and others preferring to wait until the patient’s medical conditions have stabilized. Numerous reports in the literature have shown that early take down is safe and effective [98][99] although the definition of what constitutes “early” differs widely among the studies [100]; early reversal is generally accepted to occur within 8 to 10 weeks of enterostomy creation [101]. Most surgeons tend to wait at least four weeks and often six weeks from the last abdominal procedure and until the infant’s weight is over 2 kg. Provided that the patient is otherwise stable, operative intervention is generally straightforward at this point. Data from Lee et al. demonstrated that birth weight was the most significant factor in predicting postoperative complications, supporting the decision wait until the infant weighs over 2 kg [102]. However, a more recent study reported no difference in perioperative outcomes when stratifying by weight at the time of enterostomy reversal, suggesting that a weight less than 2 kg at the time of surgery was not associated with worse outcomes [103]. Regardless of their weight, neonates with significant lung disease and/or congenital heart disease are best treated with delayed stoma closure. In all situations, an inability to advance enteral nutrition because of high ostomy output may lead to a dependence on PN which can cause worsening cholestasis. In this situation, the benefits of enterostomy reversal and the ability to advance enteral feeds will often push the care team toward earlier operative intervention.

The benefits of early reversal include the ability to advance enteral nutrition, decrease the dependence on parenteral nutrition (PN) and decrease the risk of the infectious and metabolic complications of long term PN. Enterostomy reversal may also improve the management of fluid balance as enteral losses are often decreased with restored bowel continuity. Finally, nutritional status is often compromised by proximally located stomas. Therefore, reversal may be associated with improved growth. Although the trend is to reverse stomas as early as is feasible and safe, infants with a distal stoma who are on full enteral nutrition and who have a stable home environment may benefit from being discharged from the hospital to undergo reversal some months in the future.

What vascular access options exist for patients with intestinal failure?

Virtually all patients with intestinal failure will require central venous access at some point for parenteral nutrition. In many cases, patients will require central venous access early in life and for chronic duration. Hence, it is important to have a strategy for central access in these patients that will provide durable functioning central access that will also preserve vascular access options for the long term.

The initial vascular access for parenteral nutrition in many intestinal failure patients, especially infants, will be a peripherally inserted central venous catheter (PICC). The advantage of a PICC is that its entry point is in a peripheral vessel and, therefore, in the short term it is not likely to cause injury to a central vein that will limit further access options in the future. In other patient populations there is some evidence that the long-term use of a PICC for parenteral nutrition can increase the risk of venous thrombosis that can extend into the great vessels [104]. A PICC is also more likely to be accidentally pulled out of the correct anatomical position. Once a child is identified as a patient who will require PN beyond several weeks it is preferable to insert a tunneled catheter. Tunneled catheters also have the advantage of a lower risk of central line associated blood stream infection.

Initial placement of a tunneled central line in a patient with intestinal failure is usually in the subclavian or internal jugular vein by a percutaneous approach. The percutaneous approach allows for preservation of the vein without ligation. Some centers will start with a branch vessel, like the external jugular, facial or even saphenous vein via a cut down approach for placement of a tunneled catheter (see Steps of the Procedure). This technique reduces the risk of direct injury to a major vessel and, in theory, preserves the larger vessels for future access. There are no published data beyond institutional consensus and preference to guide the selection of a vessel for vascular access in children with intestinal failure. There are data to show that lower extremity central lines are more prone to infection and this pattern is likely accentuated in children with short bowel syndrome who may have high volume loose stool output. Therefore, long term lower extremity venous access in intestinal failure patients is not recommended [105]. If a femoral or saphenous vein must be used for access the line should be tunneled to an exit site down the leg and preferably along the lower thigh to avoid contact with stool and the diaper. Lower extremity access also results in a chronic indwelling catheter in the inferior vena cava which may ultimately be needed for the vascular supply to an intestinal transplant graft. Finally, a lower extremity line may be easier to tangle and can interfere with crawling and walking in the young infant or toddler. For these reasons, long term central venous access in the upper body (e.g. the internal jugular or subclavian veins) is preferred in the short bowel patient.

Beyond a PICC, the preferred catheter type for central access in a child with intestinal failure is usually a single lumen tunneled catheter. An additional lumen is usually not necessary in a child on PN and data demonstrate that double lumen catheters have an increased risk of infection compared to single lumen devices [106]. A subcutaneous port is not indicated because the line needs to be accessed daily and for long durations at a time, thereby making needle access complicated and cumbersome in these patients [105].

There are now well established data from multiple centers that ethanol lock therapy for central lines reduces the incidence of central line associated blood stream infection in children with intestinal failure[107][108][109]. Ethanol locks are contraindicated in polyurethane catheters since the ethanol can degrade the plastic material. Therefore, silicone based central venous lines should be used when available in patients with short bowel syndrome. (see Prevention)

What enteral access options exist for patients with intestinal failure?

One of the underlying principles of intestinal failure management is the provision of enteral calories as early as possible even if the remnant intestinal length is so short that only trophic feeds are possible. Trophic feeds consist of the administration of a small volume (10 mL/kg/day in a neonate) of a balanced enteral formula to allow for positive gastrointestinal benefit rather than for nutritive value. These gastrointestinal benefits are thought to improve intestinal immune function, intestinal blood flow and may decrease bacterial overgrowth, translocation and episodes of sepsis. Intestinal adaptation is a complex, multistep process which is aided significantly by enteral nutrition - particularly fat intake. Enteral feeds also help maintain the gut integrity thereby decreasing infectious complications caused by bacterial translocation. The process of intestinal adaptation and the transition to full enteral nutrition may take years to occur and can be characterized by periods of time when enteral advancement is more difficult than others. Children with short bowel syndrome have increased caloric requirements due to chronic malabsorption and may not be able to meet these needs orally - particularly if oral aversion is present. It is for all these reasons that essentially every child with intestinal failure requires durable enteral access.

Enteral access is divided into broad categories based on

• the mechanism of placement (open, laparoscopic or endoscopic)
• how long the device is needed (short- or long-term)
• the portion of the intestine that is fed (gastric or small bowel)

In general, nasoenteric feeding tubes should be minimized in this population as most children are expected to need enteral support for longer than the four to six weeks. Nasogastric and nasoenteric tubes may become dislodged. While caregivers can be trained to place nasogastric tubes and use of a bridle, particularly for nasoenteric tubes, may decrease the incidence of inadvertent tube dislodgement, nasoenteric tubes usually need to be placed via fluoroscopy or endoscopy with a requirement for radiation exposure and general anesthesia, respectively. Feeding tubes placed through the nose may also be associated with epistaxis and sinusitis. Finally, because these tubes are of a small caliber, they may clog with unacceptable frequency. It is generally preferred that children with short bowel syndrome not be discharged from the hospital with nasogastric or nasoenteric feeding tubes in place.

Because enteral tubes can be placed via open, laparoscopic, endoscopic, ultrasonographic, or fluoroscopic approaches, there are virtually no absolute contraindications to the placement of a long term device although abdominal adhesions, changes in anatomy due to prior procedures and chronic abdominal distention may make the procedure more technically challenging. A good practice in presumed intestinal failure patients is to place enteral access at the time of the patient’s definitive operation (i.e. final abdominal closure and/or enterostomy take down).

Enteral access algorithm

Open Stamm gastrostomy was the standard approach for enteral access until percutaneous endoscopic gastrostomy was introduced in the 1980’s. With the advent of laparoscopy and ultrasonography/fluoroscopy there are now four reliable techniques for the placement of gastric feeding tubes. A recent systematic review of these approaches in pediatric patients found that percutaneous endoscopic gastrostomy (PEG) placement was associated with more complications compared to the laparoscopic approach. All other comparisons analyzed found no difference in complication rates although there was a paucity of data regarding the image-guided technique [110]. While PEG placement is highly successful, its complication rate has been reported to be as high as 10% [111]. Gastrocolocutaneous fistula is probably the most feared complication of this technique and occurs when the transverse colon is interposed between the stomach and abdominal wall. The likelihood of this complication may be increased in children who have had numerous laparotomies and whose colon and stomach may be relatively fixed due to intra-abdominal adhesions. Gastrocolocutaneous fistulae may present immediately after tube placement or may become apparent after the first tube change. Almost all require operative correction [112]. More common complications related to gastric feeding tube placement include leakage, local infection, overgrowth of granulation tissue and a persistent fistula after tube removal.

Gastric feeding is preferred over small bowel feeding as it is more physiologic and allows for both bolus and continuous formula administration. Blenderized diets, which are being used increasingly in a variety of disorders, are more easily provided by gastrostomy as tube replacement due to clogging is straightforward and can be done at home, thus obviating the cost and inconvenience of a medical encounter.

Children with significant gastroesophageal reflux or dysmotility may only tolerate postpyloric feeds. Placement approaches for postpyloric tubes are the same as for gastric tubes. The newest technique – direct percutaneous endoscopic jejunostomy tube (D-PEJ) – has been shown to be safe in the pediatric population although the largest series involved only five patients [113]. Downsides to postpyloric feeding include the need for continuous feeds and the fact that the duodenum, an area of significant nutrient absorption, is largely bypassed. Finally, for children with extremely short intestinal length, feeding beyond the stomach leaves very little small bowel exposed to the nutrient admixture. Complications of percutaneous jejunostomy tube placement are similar to those of percutaneous gastrostomy tube placement.

Many children with intestinal failure benefit from access to both the stomach and the small bowel via combination tubes which can be placed via open, laparoscopic or percutaneous routes. An advantage of these tubes includes the ability to feed the jejunum while decompressing the stomach. Sham oral feedings may be facilitated with gastric drainage. Medications, which often clog the smaller bore jejunostomy tubes, can be administered into the stomach without affecting enteral nutrition. Combination tubes have the significant disadvantage that they need to be replaced by fluoroscopy or endoscopy with their associated risks including repeated exposure to ionizing radiation. The most common complications unique to jejunal feeding tubes are proximal migration of the tube tip and clogging of the tube. Both can often only be corrected by tube replacement. Standard gastroenteric tubes (versus those with a Dobhoff-type extension) are at risk for perforating the small bowel in children less than six kg.

Enteral feeding tubes are a critical component in the care of children with intestinal failure. A variety of tube options exist, and their use must be individualized for each patient.

Steps of the Procedure

What technical considerations regarding vascular access are required in intestinal failure patients in order to ensure long term success?

Durable long term vascular access is a priority in children with intestinal failure who are dependent on parenteral nutrition (PN). There are several technical aspects of line placement that should be considered to achieve long term central access.

It is paramount to avoid ligating the large veins that act as conduits for central access in this population. Children with intestinal failure may require years of central access and, therefore, it is important to maintain all options for vascular access. Hence, percutaneous techniques for line placement performed under ultrasound guidance are preferred in this cohort to avoid excessive manipulation and potential injury of the vessel. Long term central venous access in the upper body is preferred.

If an open cut down procedure is required the venotomy should be performed using a small incision with an angled blade that can be reapproximated around the catheter using a pursestring suture or equivalent. An alternative option for performing the venotomy is to use a large bore intravenous catheter (i.e. 16 gauge needle) to create the venotomy and then thread either the wire or the end of the catheter directly into the lumen of the vein. This allows for access to the vein without circumferential dissection or clamping of the vessel.

Another option to avoid manipulation of a large vein is to start with a branch vessel. In this way, some centers will routinely place the initial tunneled catheter in the external jugular, facial or greater saphenous vein in a young child with intestinal failure. This technique can be performed using an open cut down technique without significant risk of injuring a major vessel. Future central access can then be performed percutaneously via the subclavian or jugular veins.

Details of the catheter itself can be important when placing a central line in a child with intestinal failure. Single lumen lines most commonly suffice since these children need the line only for parenteral nutrition and, occasionally, intravenous antibiotics. Lines placed in a young infant or toddler who will need the line for a prolonged duration of time should be placed at the junction of the superior vena cava and right atrium or slightly deeper in the mid right atrium. This allows the line to remain central as the child grows and may minimize the need to replace a central line for malposition.

Central lines in children with short bowel syndrome require replacement over time due to infection, thrombus, malposition and line breakage. Initial attempts at repairing the external portion of a central line if possible are indicated to avoid excessive procedures to replace the line. At some point, a line that has been repaired multiple times likely needs replacement to avoid infection - although there are no data to indicate the optimal timing for a line replacement in this scenario. When a central line needs to be replaced consideration should be given to replacement of the line over a wire. In this way, the line is dissected and accessed in the subcutaneous tissue just before it enters the vein. The line is carefully divided and a wire is placed through the line and into the right atrium under live fluoroscopy. A new line, with a separate tunnel, can be placed over the wire to preserve the same venous access site. Alternatively, the cuff can be dissected free and a Transend^® wire can be passed into the right atrium through the intact line. The line can then be replaced over the wire.

What unique approaches for vascular access may be used in the patient with diminishing venous access options?

Children with intestinal failure require long term PN and, therefore, are at risk for common venous access site thrombosis and stenosis from chronic central venous access. Preservation of central line access sites is a priority in the care of children with short bowel syndrome. Therefore, significant venous thrombosis should be treated with anticoagulation and serial ultrasound imaging to evaluate for patency. Large veins with significant clot burden have been noted to recanalize after appropriate treatment with anticoagulants. In addition, balloon dilation and stenting can be performed with the help of interventional radiologists to actively improve the patency of large veins with clot burden or achieve recanalization of chronically occluded central veins [105][114]. Less invasive methods of venoplasty such as the utilization of soft guide wires and dilators or noncompliant angioplasty balloons should be attempted prior to using more aggressive techniques such as sharp recanalization, cutting balloons, or stents [105]. There are reports of central lines placed successfully through stents in the superior vena cava, inferior vena cava, and iliofemoral veins, although recurrent thrombosis remains a potential complication [115][116].

When the common upper and lower extremity sites for central venous access are no longer available alternative options are required. Percutaneous transhepatic venous access is an established technique to obtain central access [117][118]. The procedure is less invasive than other alternative forms of central venous access and can be readily performed under ultrasound guidance. Transhepatic catheters are prone to dislodgement and changes in position due to anatomical factors and diaphragmatic movement during respiration. Authors have recommended affixing the catheter to the peritoneal lining by a small incision to reduce the risk of dislodgement.

In a similar fashion, procedures to insert percutaneous translumbar catheters have been described [114][119]. These lines are performed with the patient in a prone position and the catheter is most often tunneled toward the inferior flank. Direct percutaneous cannulation of the inferior vena cava has been described even in the setting of a completely thrombosed vena caval filter. Retrograde cannulation of large upper extremity veins through a thrombosed superior vena cava and brachiocephalic veins has also been described [120]. In this technique, retrograde access of the axillary or internal jugular is obtained from the groin by maneuvering a guidewire past the clot burden in the superior vena cava and brachiocephalic vein. A cutdown approach is then performed to access the tip of the guidewire in the vein and a new line is threaded along the wire and position in the proximal vena cava.

The most invasive technique for alternative central venous access involves direct cannulation of the right atrium [121]. This procedure can be performed using a right anterior thoracotomy or a median sternotomy. In addition to its invasive nature this technique also carries the risks of catheter dislodgement and injury to mediastinal structures. Another innovative approach to central venous access involves thoracoscopic guidance of subpleural needle access to central venous structures.[122]

These alternative techniques for central venous access are usually performed as a last resort when more standard options no longer exist. Many require advanced preoperative imaging (including computed tomography angiography) to evaluate the relevant vascular anatomy. All of these techniques mandate a dedicated multidisciplinary team of skilled providers specializing in pediatric surgery, interventional radiology, cardiac surgery and pediatric anesthesiology.

How are the LILT and STEP procedures performed?

see Intestinal Lengthening Procedures

What other options are available for management of short bowel syndrome and how are they performed?

Its comparative technical simplicity and ability to create a uniform caliber of small bowel, despite variable degrees of underlying intestinal dilation, have resulted in the serial transverse enteroplasty (STEP) operation being currently the most popular bowel lengthening and tapering surgical procedure. Historically there have been other innovative attempts to treat short bowel syndrome. In the early 1980s the longitudinal intestinal lengthening and tailoring (LILT) operation, was described by Bianchi. Using the anatomic principle that the leaves of the mesentery supplying the small bowel can be divided into two this procedure partitions the dilated bowel in half and then by sliding and anastomosing the resultant segments of small bowel in an isoperistalic manner it tapers and lengthens the intestine. Due to the geometry of this operation variable intestinal dilation cannot be accounted for and angulated anastomoses are a risk. Another venerable but ingenious operation for short bowel syndrome was described by Kimura. It involves suturing the small intestine to either the liver or abdominal wall and then allowing the bowel to parasitize blood supply. At a second operation the bowel loop is then divided longitudinally and anastomosed with the resultant lengthened segment deriving half its blood supply from the liver or abdominal wall. Needless to say this is also a technically challenging approach.

Other operations have been suggested to specifically augment absorption or slow motility in short bowel syndrome. Theoretically recirculating loops of intestine may improve nutrient absorption, though given the presence of baseline disordered motility and tendency to bacterial overgrowth in short bowel syndrome, they cannot be recommended as a practical option. Attempts to specifically slow bowel motility have also been suggested. Placing a short segment of intestine (10 cm or less) in an antiperistaltic orientation can create a partial bowel obstruction and hence slow transit. Unfortunately, this may also result in proximal bowel dilation, bacterial overgrowth, abdominal pain and vomiting. Interestingly some authors have suggested temporarily placing an antiperistaltic loop of bowel, or partially obstructive circumferential mesh around the intestine, to purposely dilate proximal bowel and in this fashion allow for a subsequent bowel lengthening operation. Although superficially appealing this technique adversely affects motility and hence hinders bowel rehabilitation during the time of partial obstruction. Hence, natural adaptive processes are masked and the necessity for a bowel lengthening operation becomes almost a certainty.

Complications

What is the long term morbidity of intestinal failure?

Hepatobiliary, renal, bone, nutritional and metabolic comorbidities are common in children with intestinal failure. The length and type of intestine resected directly affects the comorbidities that can be expected in each individual. Carbohydrate and fat and protein absorption, for example, occur primarily in the duodenum and jejunum whereas vitamin B₁₂ and bile salts are only absorbed in the ileum. In general, loss of the proximal small bowel is better tolerated than more distal resections as the ileum has a greater ability to adapt than the jejunum.

In addition to the medical comorbidities noted above, psychosocial morbidities and neurodevelopmental disability are also important morbidities in these patients, as children with intestinal failure have reported a decreased quality of life and decreased performance in school [123][124][125].

Hepatobiliary

Intestinal failure associated liver disease (IFALD) is a significant potential complication of parenteral nutrition. While recent data suggest that the lipid component of parenteral nutrition bears a direct injurious role in this process, IFALD is nonetheless multifactorial, initially manifesting as cholestasis and ultimately leading to cirrhosis and liver failure if not prevented or reversed early in its course. (see Medical Decision Making) The long-term implications of IFALD are unclear, but ongoing hepatic inflammation could potentially increase the risk for development of hepatocellular carcinoma later in life [126]. However, intestinal failure patients with biopsy-proven cirrhosis can remain clinically stable for prolonged periods of time and have been shown to have similar long-term transplant-free survival as those without cirrhosis [127]. Thus, ongoing efforts to minimize liver injury and to achieve enteral autonomy remain paramount to reducing hepatobiliary morbidity in these patients. Several nonlipid parenteral nutrition related factors which can lead to liver injury include the provision of excess calories and repeated episodes of sepsis. Therefore, meticulous line care and thoughtful provision of parenteral calories are paramount in the prevention of IFALD.

Protective measures are also helpful in preventing IFALD. These include the provision of any amount of enteral calories if the intestine is at all able to tolerate feeding and the use of cyclic parenteral nutrition (PN). Cyclic parenteral nutrition is felt to improve liver dysfunction by reducing hyperinsulinemia, promoting free fatty acid mobilization and the more efficient use of metabolic substrates thus reducing liver stress. The use of cyclic parenteral nutrition, if able to be done safely with respect to glucose infusion and serum glucose stability, has been shown to prevent or reverse development of cholestasis [128][129].

In regards to the lipid component itself, many authors have advocated for a lipid minimization strategy to limit the hepatic toxicity seen with infusion of soybean based lipid sources. While the specifics vary by center, lipid minimization has been shown, with some mixed results, to limit the development of IFALD when used selectively in patients expected to require long term PN support (i.e. greater than three weeks) [85][130]. The lipid dose most often chosen is 1 g/kg/day, which has been proven to not result in biochemical or clinical evidence of essential fatty acid deficiency. Nonetheless, such a strategy mandates close attention to this potential risk and additional study is required to evaluate the possibility that lipid restriction in at-risk neonates could result in neurological impairment. Over the past decade, the use of different sources of parenteral lipid, focusing on the use of omega-3 fatty acids, has gained traction. Even in the era of hepatoprotective PN, which is defined as lipid restriction to 1 g/kg/day or use of fish oil-containing lipid emulsions, recent studies have shown ongoing elevation of hepatic transaminases, even after bilirubin levels normalize, suggesting persistent low grade liver inflammation despite resolution of cholestasis [126].

After cholestasis, biliary sludge and cholelithiasis are the most common hepatobiliarymorbidities seen in children with intestinal failure. Risk factors for cholelithiasis include a relative lack of bile salts due to disruption of the enterohepatic circulation and biliary stasis due to a lack of enteral feeds [131]. The duration of parenteral nutrition use has also been shown to be an independent risk factor for the development of gallstones in children with a prior ileal resection [132]. It may, therefore, be reasonable to consider early prophylactic cholecystectomy in this population. Gallstones and biliary sludge are best prevented with enteral nutrition. Ursodiol is commonly used in patients with intestinal failure to promote bile flow, aid in the dissolution of cholesterol gallstones, and help prevent gallstone formation in those with biliary sludge [133][134].

Bacterial overgrowth

Children with evidence of intestinal bacterial overgrowth may have dysfunctional and dilated small bowel remnants. In addition, bacterial overgrowth can lead to an increased incidence of central line infection and sepsis. Long term enteral antibiotics may be needed to treat bacterial overgrowth in these patients, typically utilized in a cyclic fashion (i.e. one week out of the month). Some intestinal rehabilitation centers will use hydrogen breath tests or endoscopic-guided duodenal aspirates to assess for bacterial overgrowth; other centers have treated bacterial overgrowth empirically.

Central line infections and venous thromboses

Frequent central line infections and venous thrombosis increase the risk of severe sepsis and may lead to loss of central venous access options.Sepsis and loss of venous access leads to a poor long term prognosis in this population. Loss of viable central venous access options is an indication for intestinal transplantation in children with short bowel syndrome who cannot wean from parenteral nutrition. Central line infections and venous thromboses should be aggressively treated with appropriate intravenous antibiotic courses and anticoagulation therapy when identified. Thrombolytics such as tPA are often used empirically when thrombosis is suspected to be the cause of catheter occlusion [135]. Prophylactic anticoagulation in patients with long-term central venous catheters has been suggested as a way to decrease rates of catheter-related thrombosis and occlusion [136][137]. However, in patients with intestinal failure, the risks of bleeding complications must be weighed against the benefit of thrombosis prevention, and a clear consensus has not yet been reached. Ethanol lock therapy should be strongly considered in patients with recurrent central venous catheter infections. Retrospective data has showed a significant reduction in catheter-associated blood stream infections using ethanol lock therapy in children with intestinal failure [56]. If ethanol lock therapy is not an option, Kitelock (EDTA 4%) therapy could be considered instead [64].

Nutrition Deficiencies

Vitamin, mineral and trace element deficiencies are common in children with intestinal failure. Bile salts derived from the liver are a crucial component of the enterohepatic circulation and are almost completely absorbed in the ileum. Extensive ileal resections, then, are most associated with deficiencies in the fat soluble vitamins A, D, E and K. Vitamin B₁₂ supplementation is also commonly needed in those who have undergone extensive ileal resection. Other water soluble vitamins such as B₂, B₆ and C are absorbed along the length of the small bowel; deficiencies are related more to the length of bowel remaining than the specific type. Sodium supplementation may be required in children with excessive diarrhea or enterostomy output. Patients with total body sodium depletion will present with poor growth, and a urine sodium can be checked to help identify if they need supplementation. Trace element deficiencies (chromium, copper, iron, magnesium, selenium, zinc) are relatively common and their levels should be monitored at least every three to six months for infants and children on chronic PN[138][139].

Renal Disease and Nephrolithiasis

Children with intestinal failure are at increased risk for kidney disease due to recurrent episodes of dehydration and sepsis, as well as potential exposure to nephrotoxic agents such as antibiotics or anti-inflammatory medications. Children with an ileal resection, but with an intact colon, are at risk for developing renal oxalate stones. Under normal conditions dietary oxalate combines with enteral calcium to create an insoluble salt which is not absorbed. In the setting of fat malabsorption, however, enteral calcium binds with free fatty acids leaving the oxalate free to be reabsorbed. High levels of serum oxalate lead to oxaluria and nephrolithiasis. These stones tend to form within the first few years after surgery and may be associated with an obstructive uropathy. This risk can be minimized by avoiding high oxalate foods which include beets, rhubarb, spinach and chocolate [138].

Osteopathy

Metabolic bone disease develops when the intricate interplay between calcium, phosphorus, magnesium and vitamin D is disrupted. Children with intestinal failure are at risk of developing osteopathy due to segmental intestinal resection, chronic malabsorption of vitamin D and calcium as well as the need for low concentrations of calcium in parenteral nutrition. Children whose caloric intake is predominantly from parenteral nutrition are at increased risk of calcium and phosphorus deficiency as limited amounts of both can be added to the solution before a precipitate occurs. In a recent retrospective review of children with intestinal failure, 50% were noted to have evidence of metabolic bone disease as defined by a bone mineral density Z-score less than one as measured by dual-energy x-ray absorptiometry (DXA). The primary risk factor for metabolic bone disease in this cohort was the duration of parenteral nutrition [140]. Bone mineral status should be assessed by DXA at least every few years if children continue to require a significant portion of their calories from parenteral nutrition. Treatment is geared toward appropriate supplementation of calcium and/or vitamin D, along with weight-bearing exercise to stimulate bone mineralization.

Gastrointestinal Bleeding and Chronic Intestinal Inflammation

Children with intestinal failure frequently develop GI bleeding, which has several potential etiologies including allergic enteritis (with eosinophilic infiltration of the bowel mucosa), bacterial overgrowth, or anastomotic ulceration. Additionally, they can develop chronic inflammation not attributable to one of these above etiologies, causing a presentation similar to that of inflammatory bowel disease (IBD). In this subset of intestinal failure patients, it may be beneficial to treat with anti-inflammatory medications such as the 5-ASA class of drugs (ie. sulfasalazine, mesalamine, or balsalazide), or even glucocorticoids. In a recent study, patients with chronic inflammation treated with glucocorticoids were shown to have significant improvement in enteral tolerance and cessation of GI bleeding [141]. However, the benefits of improved enteral tolerance must be weighed against the known risk of stunted growth associated with prolonged steroid use. Anti-inflammatory medications can be administered enterally or rectally, depending on the portion of bowel most affected. Treatment of focal ulceration with either topical anti-inflammatory therapy or endoscopically guided surgical resection should be determined based on individual patient considerations.

Quality of Life

Health-related quality of life (QOL) is increasingly recognized as a valuable metric in the management of pediatric patients with chronic diseases. Children with intestinal failure have significantly lower QOL scores compared to healthy children and have similar to worse QOL scores compared to children with chronic illness [124][125][142]. Family members of patients with intestinal failure also have reduced QOL scores [143]. This impaired quality of life becomes particularly evident once the children reach school-age, with poor school functioning and neurodevelopmental disabilities being major contributors [123][124][125]. One study evaluating QOL in a cohort of adult intestinal failure patients identified the presence of an ostomy and the absence of oral intake as two of the main determinants of reduced quality of life [144]. Another study in children with intestinal failure found an association between degree of parenteral support and poor QOL [145]. In amalgam, these reports highlight that efforts to improve patient and family quality of life are an important part of the overall care of patients with intestinal failure.

What are the molecular mechanisms and histopathologic evolution of intestinal failure associated liver disease?

Intestinal failure associated liver disease (IFALD) has distinct histopathologic features. Early microscopic findings include lipid accumulation in the hepatocyte (i.e. hepatic steatosis). It is thought that hepatic steatosis is caused by excess dextrose in the bloodstream leading to increased insulin activity. Insulin stimulates the hepatocellular uptake of glucose which can is converted to fat in the process of lipogenesis. With continued exposure to parenteral nutrition, histopathologic changes can progress to hepatocellular ballooning, periportal inflammation, bile plugging and bile duct proliferation. The tissue injury evolves to hepatic fibrosis which progresses to bridging fibrosis between portal triads. The latter stages of this process are demonstrated by scarring of the liver parenchyma including cirrhosis [146][147].

The exact mechanism for IFALD is thought to be multifactorial and involve processes both within the liver and secondary to changes in the intestine associated with short bowel syndrome. This is why the term parenteral nutrition associated liver disease has been recently changed to IFALD. Many mechanisms for IFALD have been hypothesized and tested in animal models although the exact molecular pathway for the pathogenesis of IFALD remains unknown .

One proposed mechanism involves phytosterols, plant-based sterols found in soybean oil based lipid emulsions [148]. Phytosterols have been shown to interrupt both farsenoid X receptor (FXR) signaling and the expression of bile acid transporters in the liver. FXR is an established regulator of bile production and flow in the liver; this receptor controls the genes involved in bile acid synthesis, uptake and export. Phytosterols are found in much higher concentrations in soybean oil based lipid emulsions than in fish oil-based lipid formulations. Human studies have shown a linear correlation between increasing duration of parenteral nutrition exposure and serum levels of phytosterols although the etiology for this relationship is not clear.

In animal models, it has been demonstrated that PN with soybean oil-based lipid emulsion in the presence of intestinal injury can lead to cholestatic liver disease. Using fish oil based emulsions in the same model does not lead to cholestasis. Interestingly, the administration of a phytosterol to the animal receiving fish oil led to cholestatic liver disease. In addition, animal models of IFALD have demonstrated that treatment with an FXR agonist can protect against liver injury.

Bacteria may play a role in this process [149]. In animal studies, the administration of lipopolysaccharide (LPS) from gram negative bacteria reduces the expression of hepatic FXR and bile salt export pumps. This in turn leads to the accumulation of bile within the hepatocyte. The use of antibiotics in IFALD animal models can prevent cholestasis.

Another role for bacteria in the pathogenesis of IFALD may involve toll like receptor 4 (TLR-4) on the Kupffer cell in the liver. Activation of the TLR-4 is known to initiate a cascade of proinflammatory cytokines in the liver which may, in turn, decrease bile flow. Mutant animals missing TLR-4 do not respond to LPS and do not exhibit liver injury from parenteral nutrition in IFALD models. One proposed mechanism for the early stages of IFALD is that LPS derived from intestinal bacteria translocates through the injured gut barrier, travels through the portal venous circulation and binds to TLR-4 in the liver. TLR-4 leads to activation of the Kupffer cell and cytokine release affecting bile synthesis and transport.

Finally, it is established that soybean oil-based lipid emulsions contain low concentrations of alpha-tocopherol, a form of Vitamin E. Alpha-tocopherol is a known antioxidant and is found in higher concentrations in fish oil-based lipid emulsions. It has been hypothesized that the low levels of alpha-tocopherol in soybean oil-based lipids may lead to an increased risk of oxidative stress and lipid peroxidation in the liver thereby causing the liver injury seen in IFALD[150].

Outcomes

What is the long-term outcome of patients with intestinal failure?

The long-term outcomes in children with intestinal failure have changed dramatically in the past decade. With contemporary management at a multidisciplinary intestinal rehabilitation program, a child with short bowel syndrome can be expected to survive into school age and beyond.

Previously, this cohort was at high risk for complications of intestinal failure including parenteral nutrition associated liver disease, coagulopathy and recurrent sepsis from central line-associated blood stream infections. Many children who were dependent on parenteral nutrition secondary to intestinal failure died early due to progressive liver failure, variceal bleeding or refractory sepsis. Within the past ten to fifteen years, significant advances have been made in the medical and surgical care of pediatric intestinal failure. These new treatment options have made long term intestinal rehabilitation safe and, in many cases, effective.

In most recent series, the current mortality for pediatric short bowel syndrome remains approximately 10% or less. Even children with the shortest bowel lengths have a reasonable chance of long-term survival; retrospective series of children with less than 20 cm of small bowel show similar mortality rates [151][152]. These data may be misleading as it includes only patients who are ultimately referred to intestinal rehabilitation centers. Hence, these data may not include young neonates with severe short bowel syndrome who do not survive the early stages of care in the neonatal intensive care unit.

The improvement in survival is likely due to a multitude of factors. Widespread implementation of multidisciplinary intestinal rehabilitation programs began in the early part of this century. These programs bring together pediatric surgeons, gastroenterologists, hepatologists, transplant surgeons, infectious disease specialists, dieticians, pharmacists and social workers to care for the child with short bowel syndrome. The goal of intestinal rehabilitation is to gradually advance enteral nutrition and prevent complications as the remnant bowel adapts to improve nutrient absorption. Retrospective data demonstrate that multidisciplinary intestinal rehabilitation programs are associated with improved survival in children with short bowel syndrome [2].

In many ways, intestinal rehabilitation has been made possible by other medical and surgical advances. For example, restriction of intravenous soy bean or fish oil based lipids in parenteral nutrition regimens may prevent liver disease in this population. Ethanol lock therapy has reduced the incidence of central line associated blood stream infection. Intestinal bacterial overgrowth can be treated with long term enteral antibiotic formulations. Safe techniques for bowel lengthening have been developed to provide increased intestinal absorptive capacity in appropriate patients. Together, these treatment options have reduced the risk of progressive liver disease in children with intestinal failure and have afforded the time needed to rehabilitate the bowel as it continues to adapt.

Another relevant outcome measure in this population is the achievement of enteral autonomy. The likelihood of enteral autonomy in children with short bowel syndrome varies from 30% to greater than 90% in the literature and is likely dependent on multiple factors including bowel length and anatomy, the etiology of intestinal failure, age at diagnosis and the development of complications from intestinal failure (see Complications). Recent data from the Pediatric Intestinal Failure Consortium (PIFCon) demonstrate that residual small bowel length greater than 41 cm best predicts enteral autonomy with 4% increase in enteral autonomy with each 1 cm of bowel length[7]. A recent single center review found the prospect of enteral autonomy in children with less than 50 cm of small bowel to be 23% at one year, 38% at two years, and 71% at five years into intestinal rehabilitation [153]. Data also demonstrate that children with ultrashort bowel syndrome (less than 20 cm of remaining bowel) may have a chance of weaning from parenteral nutrition with their native bowel. While variable, the available data demonstrate that enteral autonomy is possible in many patients with short bowel syndrome and the process of weaning from parenteral nutrition can take years. These data help to validate the recent emphasis on intestinal rehabilitation in patients with intestinal failure.

As outcomes in pediatric intestinal failure have improved, the condition has transitioned from a form of critical illness to a chronic disease. In this way, outcomes data must also address the quality of patient survival. Recently studies have started to address quality of life and neurocognitive outcomes in this cohort, although much of these data are preliminary in nature.

How does the morbidity of intestinal failure influence the outcome?

The comorbidities and complications that primarily affect mortality in intestinal failure are intestinal failure associated liver disease (IFALD) and central line associated blood stream infections. Nutrient deficiencies are also important potential causes of long-term morbidity.

Intestinal failure associated liver disease (IFALD) has historically been reported to be present in 25% to 33% of neonates with short bowel syndrome and it is associated with increased mortality. However, recent therapeutic advances have substantially reduced both its incidence and mortality. Risk factors for IFALD include low birth weight, prematurity, duration of parenteral nutrition and number of septic episodes. The resolution of the clinical manifestations of IFALD, including hyperbilirubinemia, has been accomplished with the institution of full enteral nutrition and newer hepatoprotective parenteral nutrition strategies (see Medical Decision Making). Of note, cirrhosis may be present even if serum indices of liver function normalize. Interestingly, biopsy proven cirrhotic IFALD pediatric patients have an excellent transplant free survival provided they ultimately transition to full enteral nutrition [127]. Contrary to previous conjecture, IFALD-induced cirrhosis does not appear to decrease the likelihood of transition to full enteral nutrition [127].

It is unknown what the long term consequences of neonatal IFALD truly are as the first major vanguard of surviving patients is only now nearing adulthood. Routine Doppler ultrasound examinations and alpha fetoprotein assessments are recommended particularly in patients with known cirrhosis. Promising new noninvasive techniques to further assess liver status in children with IFALD include the use of sequential intravenous stable isotopic breath tests (¹³C-methionine) [154] and liver elastography.

Catheter associated bloodstream infection (CLABSI) is another major potentially fatal complication of pediatric intestinal failure. The baseline infection rates in pediatric intestinal failure approach 10/1000 catheter days and the major causative organisms are gram negative bacteria. Standardized line care protocols reduce the risk of serious line infections. Additionally, the use of 70% ethanol locks in patients with a prior history of CLABSI has been associated with a decrease in line infections rates to 2.2/1000 catheter days [56]. The avoidance of tying off vessels when introducing central lines and the advent of novel interventional radiology techniques has greatly reduced loss of intravenous access as an indication for intestinal transplantation in pediatric intestinal failure [52].

Multiple nutrient deficiencies are frequent findings in pediatric patients with intestinal failure [155]. Recent studies also demonstrate that they have significant evidence of low bone mineral density when assessed by DXA scan at school age or older [156]. This puts the group at significant risk for the complications of osteoporosis as adults. Careful attention to appropriate growth, adequate calcium intake and normal vitamin D levels as well the encouragement of weight bearing exercise is necessary to minimize this complication.

In general, adequate long term studies of the morbidity consequent to pediatric intestinal failure are lacking.

Follow-Up

How are patients with intestinal failure best followed long term?

Children with intestinal failure do better when their care is coordinated by a multidisciplinary team whose clinical and scholarly focus is on the medical, surgical and nutritional management of these complex patients. Patients followed in these programs are more likely to wean from parenteral nutrition and have increased overall survival [2][157]. The initial management of children with intestinal failure is done on an inpatient basis but once clinical, metabolic and nutritional parameters have stabilized, most children can be managed as outpatients. Routine follow up in an intestinal failure clinic may be required on a frequent basis, especially during periods of increased somatic growth (e.g. infancy, puberty) when metabolic and nutrient demands change frequently. During periods of stability, however, most care can be provided at home once caregivers are trained to do so.

Improvements in the medical and surgical care of infants and children with intestinal failure has allowed for the development of a population of teenagers and young adults who have had intestinal failure since birth and whose complex care will need to transition from pediatric to adult providers. Barriers to this transition of care may occur at the level of the pediatric provider, the adult provider and from the patient and his/her family. Pediatric practitioners have often taken care of these children for years and may be reticent to transfer care to a provider they may not know well. Adult practitioners may have little experience in the management of intestinal failure and, therefore, may be unwilling to take these young adults into their practice. Patients and families may have a very hard time accepting that the transition needs to occur. Many children with intestinal failure have received all of their care by a single team over a long period of time. A transition to an adult provider may be internalized as abandonment and may create anxiety for parents who may no longer be accepted as the primary communicator regarding their child’s care [158]. Smooth transitions from pediatric to adult care occur when the process is open, individualized, gradual and involves participation by all stakeholders [159].

What is the value of intestinal failure programs?

The management of children with intestinal failure can be challenging from medical, surgical, nutritional and social perspectives. The potential benefits of multidisciplinary teams include more consistent care via standardized protocols, the development of clinical expertise, improved patient satisfaction by coordinated outpatient care and increased opportunities for scholarly activity. Investigators from a number of large volume intestinal failure programs have assessed patient outcomes following the implementation of a multidisciplinary intestinal failure program. Overall growth, individual nutritional parameters, the ability to wean from parenteral nutrition, survival and the avoidance of transplantation have all been shown to improve when children are managed by these teams [2][157][160][161]. Lower peak bilirubin levels and lower rates of sepsis have also been observed in these children [2][161]. These data support the idea that children with intestinal failure and parenteral nutrition dependence should be referred to a regional intestinal rehabilitation center as soon as possible and before cholestasis or liver failure begin to develop.

Most interdisciplinary intestinal failure teams are composed of at least one of each of the following

• gastroenterologist
• surgeon
• clinical dietitian
• pharmacist
• nurse
• social worker
• psychologist

Research and Future Directions

How is intestinal failure defined from a research perspective?

Animal models of short bowel syndrome are defined by a small intestinal resection of 80% or greater, although no such consensus exists for pediatric patients. In clinical investigation a purely functional definition of intestinal failure may be used. The simplest and most frequently applied is parenteral nutrition dependence for greater than 90 days [4]; however, some investigators have used a definition based on parenteral nutrition dependence for greater than 60 days [45][162]. More complex definitions based upon amalgams of functional and anatomic considerations have also been suggested. The current consensus definition to allow for comparisons of outcomes across studies and institutions is dependence on parenteral nutrition for 60 days out of 74 consecutive days [5][45].

What experimental studies are being done regarding altering the microbiome?

Through metagenomic and biochemical analysis, the intestinal microbiota of genetically obese mice have been shown to have an increased capacity for energy harvest from the diet [163]. In this study, transfer of stool into the gastrointestinal tract of germ-free mice resulted in a significant increase in body fat. Stools from obese mice displayed a proportional increase in Fermicutes phyla (which includes the genus Lactobacilli) in their intestinal lumen. It is therefore plausible to hypothesize that the altered intestinal microbiota in patients with intestinal failure adapts to provide greater energy harvest for the host.

What treatment modalities are currently being evaluated for augmenting the intestinal adaptation process?

In general, growth factors have been evaluated extensively in multiple animal models primarily as intestinotrophic agents. The major mechanism that has been evaluated to enhance postresection mucosal growth is via the stimulation of enterocyte proliferation. A more comprehensive understanding of the pathogenesis of the adaptation process may allow for improved clinical interventions designed to enhance adaptation beyond the simple initiation of cell proliferation. For example, rates of enterocyte cell death [21] are also elevated in the adapting bowel - probably to counterbalance the elevated rates of proliferation. As such, therapy targeted to reduce rates of cell death may also result in amplified adaptation responses.

In consideration of other cells within the intestine that may affect the magnitude of adaptation, the underlying mesenchyme has been demonstrated to signal the overlying epithelium for growth. In addition, growth factor delivery to the intestinal smooth muscle (as opposed to the epithelium) may ideally enhance intestinal lengthening as well as villus elongation. This is suggested by a study in which transgenic over expression of insulin-like growth factor specifically within the intestine smooth muscle cells resulted in significant gut lengthening after intestinal resection [164].

More recently, angiogenesis within the intestinal wall has been described as an essential component of adaptation [165]. It is presently unknown whether the angiogenic response to intestinal resection is the primary driver of mucosal growth or if mucosal growth is the primary signal for angiogenesis. Proangiogenic therapy may therefore be a useful therapy in the immediate postresection period.

The optimal diet to promote adaptation is still unknown. Diets that are high in fat [166] as well as protein [167] content have been shown in animal models to enhance adaptation. The optimal concentration and composition (omega 3 versus 6 for fat, simple amino acids versus dipeptides versus more complex proteins) have yet to be clearly elucidated.

Finally, alterations in the enteric microbiome associated with intestinal failure warrant further investigation. As previously described, sepsis and liver damage are well known morbidities associated with intestinal failure. Manipulation of the postresection gut microbiome may, therefore, be clinically valuable as a means to reduce septic complications as well as to enhance availability of luminal nutrient for the adapting host.

What treatment modalities are being considered for the expansion of existing intestine?

Although autologous intestinal reconstructive surgery (AIRS) such as the longitudinal intestinal lengthening procedure (LILT) and the serial transverse enteroplasty (STEP) can improve bowel function and nutrient absorption, their use is limited to patients with intestinal dilation. Distraction enterogenesis, or creating intestinal tissue by mechanical stretch, is a promising therapy for short bowel syndrome. The applied tension induces growth via several known molecular pathways, stimulates mesenteric neovascularization and stretched jejunal segments appear to function well when placed in continuity [168]. A number of creative approaches have been attempted in animal models [169][169]. Constructs minimizing operative interventions and allowing the bowel to remain unobstructed and in continuity appear preferable. A recent rodent investigation has demonstrated the feasibility of an extraluminal, self expanding, scalable, shape memory polymer system for distraction enterogenesis [170].

Another area of active investigation is the use of tissue engineering principles to augment intestinal surface area. Neonatal rat intestinal cells have been implanted into a scaffold and anastomosed to the remaining bowel with evidence of increased brush border enzymes [171]. Further the implantation of VEGF microspheres into the construct appears to increase epithelial proliferation and microcapillary density.

Patient Care Guidelines

What is the algorithm that should be followed in the management of feeding for the patient with intestinal failure?

guidelines for enteral feeding

Suggested practice guidelines for initiation and advancement of enteral feeding in the neonate and infant with intestinal failure. *Feeds should generally be held for 8 hours, then restarted at 75% of the previous rate. Supplemental intravenous fluids may be needed to account for volume loss from this adjustment. (From Brenn M, Gura KM, Duggan C. “Intestinal Failure,” in Manual of Pediatric Nutrition, 5^th edition. Kendrin Sonneville and Christopher Duggan, eds. Used with permission from PMPH-USA, LTD, Shelton, CT.)

What algorithm should be followed when considering vascular and enteral access in the patient with intestinal failure?

enteral access algorithm

Vascular Access

Initial attempts at securing vascular access should be made with an overall strategy of preserving all vascular access options for the long term. These strategies include

• avoiding long term lower extremity access to reduce the risk of cental line infection and thrombosis of the inferior vena cava
• start with a cut down approach for line access in a branch vessel (external jugular, facial, or greater saphenous veins)
• prioritize percutaneous techniques for central line placement to avoid the potential for ligation of a large vessel
• avoid ligation of a large vein even with cutdown placement is performed
• for malfunctioning catheters without evidence of infection, consider replacement over a wire rather than accessing a new vessel
• avoid long term use of a PICC line for chronic central venous access

Perspectives and Commentary

To submit comments about this topic please contact the editors at think@apsapedsurg.org.

Additional Resources

APSA parent and patient education materials on Short Bowel Syndrome (also known as short gut syndrome, short gut, small intestinal insufficiency)

Stay Current in Pediatric Surgery podcast Intestinal Failure

Discussion Questions and Cases

To submit interesting cases which display thoughtful patient management please contact the editors at think@apsapedsurg.org.

Additonal questions are in SCORE Short Bowel Syndrome/Intestinal Failure conference prep

[17]

[29]

References

Media

enteral access algorithm

STEP for D-lactic acidosis

Artist’s rendition of a STEP procedure being performed on the duodenum for the treatment of D-lactic acidosis in the setting of short bowel syndrome and small bowel bacterial overgrowth. (Reproduced with permission from Modi et al., J Pediatr Gastroenterol Nutr 2006.)

guidelines for enteral feeding

Suggested practice guidelines for initiation and advancement of enteral feeding in the neonate and infant with intestinal failure. *Feeds should generally be held for 8 hours, then restarted at 75% of the previous rate. Supplemental intravenous fluids may be needed to account for volume loss from this adjustment. (From Brenn M, Gura KM, Duggan C. “Intestinal Failure,” in Manual of Pediatric Nutrition, 5^th edition. Kendrin Sonneville and Christopher Duggan, eds. Used with permission from PMPH-USA, LTD, Shelton, CT.)