Intestinal failure (IF) is a condition in which severe malabsorption results in the need for lifesaving artificial nutrition provided through a parenteral route. IF may be reversible or irreversible, depending on a number of factors such as the underlying diagnosis and the treatments used to restore intestinal competence. Irreversible IF in children remains a challenge; nevertheless, relevant advances in the management of some of the diseases that have resulted in IF continue to advance the effectiveness of our treatment regimens. Because IF is relatively rare and because the etiologies of IF are varied, there is a serious paucity of evidence-based data. There are not enough data to provide the scientific foundation needed to form treatment guidelines or for the creation of universal standards for the care of such patients. Our aim is to report on the European experience in treating young patients with IF, with particular emphasis on our preference to provide conservative treatment before recommending intestinal transplantation.
Definition of IF and Nutritional Failure
IF is commonly defined as a critical reduction of the gut mass or its function below the minimum needed to absorb nutrients and fluids required for adequate growth in children and weight maintenance in adults (1). In clinical practice, intestinal sufficiency is indirectly measured by the percentage of parenteral nutrition (PN) required to grow. Indeed, the 2 indicators that have historically been used to predict weaning from PN in children, residual bowel length measured at final surgery, and serum citrulline, although helpful, have not proven to be greatly reliable prognostic factors in children when IF is caused by short bowel syndrome (SBS) (2,3). PN requirements remain, therefore, the best measure of the degree of intestinal sufficiency in this setting (4,5). Along with the definition of IF, a new concept of nutritional failure in children can be defined as the failure to provide long-term nourishment to a child either by natural or by artificial means. PN may be a limited option for a child with IF because of severe liver disease, overwhelming catheter sepsis, or loss of venous access (6). According to the present long-term graft's and patient's survival following intestinal transplant (ITx), IF itself may be a debatable indication for ITx, whereas nutritional failure remains a clear indication (6).
Causes of IF
In developed countries, IF in children is most commonly caused by congenital or neonatal intestinal diseases that can be divided into 3 groups: disorders with a reduced absorptive surface (eg, the SBS); disorders with an intact, although inefficient, mucosal surface (eg, the congenital enterocyte disorders); and disorders with an intact mucosal surface but with extensive motility dysfunction (eg, chronic intestinal pseudo-obstructions, CIPOs) (1). In developing countries, protracted (but treatable) postinfectious diarrhea is the leading cause of IF in children (7).
Definition and Etiology
SBS is a disorder characterized by a compromised bowel absorptive capacity because of a severely reduced mucosal surface resulting in diarrhea, fluid and electrolyte imbalance, and malnutrition (8). SBS usually follows extensive surgical resection, leaving the bowel length below a critical value for adequate nutritional supply (9). At term, neonates have a small bowel length of approximately 250 cm and their intestines lengthen substantially during the first year of life. Preterm infants have an even greater potential for bowel growth. Thus, if the surgical resection occurs earlier, the opportunity for adaptive growth is greater (10).
The cutoff length for SBS is related to a number of factors. In general, SBS occurs after massive resections leaving <40 cm of viable small bowel; nevertheless, a residual bowel length of only 15 to 40 cm has been associated with bowel adaptation when resection occurs around the full term. Important factors (other than the length of bowel remnant) that determine whether IF will develop include the underlying diagnosis, the type of segments preserved, the long-term maintenance of a stoma versus a primary anastomosis, the presence of the ileocecal valve, as well as the age of the patient at the time of surgery (11–15).
Despite the focus on the physical aspects of the small bowel, SBS is a functional rather than an anatomical entity; in fact, other factors are relevant to the development of SBS, such as the functionality of the residual bowel, especially the presence or absence of normal motility. Thus, the major determinants of IF in SBS are the age of the patient at the time of resection, the extent and the portion of the small bowel resected, and the functional integrity of the residual intestine (13).
In children, the conditions most commonly leading to extensive small bowel resections are necrotizing enterocolitis (NEC), intestinal atresia, gastroschisis, and extensive aganglionosis in Hirschsprung disease (Table 1). Of these, gastroschisis and Hirschsprung disease are almost always associated with dysfunctional motility in the residual bowel and hence often result in more serious IF (14).
Management of SBS
The management of SBS involves measures aimed at promoting small bowel adaptation and villous hyperplasia by enteral (oral or tube) feeding, by providing normal somatic growth with PN, and by optimizing the bowel absorptive surface through nontransplant surgical techniques (16). Issues related to long-term PN will be discussed toward the end of the article.
When possible, the gastrointestinal (GI) tract is preferred for feeding because it is the most physiological and safest way to provide nutrition. PN should ideally not be stopped until adequate intake and growth can be achieved with oral and/or enteral feeding alone. The optimal strategies for enteral feeding, including oral versus tube feeds and continuous versus bolus, remain a matter of debate. Nutrition, particularly enteral feeding, is the primary treatment of SBS. The advantages of oral feeding, when feasible, include the maintenance of sucking and swallowing functions along with the interest and enjoyment associated with eating, as well as the stimulation of hormones released by the GI tract promoting adaptation. For example, oral feeding promotes the release of epidermal growth factor (EGF) from salivary glands, increases GI secretion of trophic factors, and helps prevent feeding disorders (17). Sialoadenectomy in animals significantly attenuates ileal villus height, total protein, and DNA content after small bowel resection. Both systemic and oral EGF administration has been shown to reverse the effect of sialoadenectomy (18).
Enteral or oral feeding should be started as soon as possible after surgery. In an infant, breast-feeding should be encouraged (19,20) or when this is not possible, an extensively hydrolyzed formula can be used (21). Human milk may be considered the first choice in this setting, followed by whole protein feeds if tolerated. Human milk contains a number of factors believed to support the developing neonate's immune system including nucleotides, immunoglobulin A, and leukocytes (22,23). Human milk also contains glutamine and growth factors, such as EGF, which possibly promote bowel adaptation (24,25). Extensively hydrolyzed formulas are mostly used when whole protein feeds are not tolerated, whereas amino acid–based formulas (AABF) have been used as a last resort. AABF are generally used in the treatment of food allergies or in case of milk protein hydrolyzate intolerance (26). True food allergies have been rarely documented in children with SBS. Andorsky et al (27) reported less intestinal allergy by using AABF, without clearly defining the criteria for the diagnosis of allergy. Two retrospective studies report that the use of an AABF was associated with earlier weaning from PN and also a reduced rate of allergies (28,29); however, the small sample sizes and the lack of a control group in these studies limit the applicability of these findings to all of the children with SBS and IF.
Feeds may be administered either continuously or in frequent small boluses, initially aiming at supplying a minimal enteral feeding (20–25 kcal · kg−1 · day−1) and then increasing gradually as tolerated with larger boluses during the day and continuous feeding overnight (30). Continuous feeding appears to improve growth and adaptation through the optimization of absorption by challenging the enteral capacity (31). Complementary food can be started early. Tolerance is evaluated by measuring stool output and by the observation of vomiting, irritability, and intestinal distension. Stool output should be no more than two-thirds of feeding volume. Many factors can affect stool volume in SBS, including the length of the residual intestinal segment, the type of segment (the more proximal the resection the larger the fluid and sodium losses), the mucosal and endoluminal variables (residual enzymatic activity and absorptive capacity, bacterial overgrowth), the presence of the colon that can absorb large amounts of water, sodium, medium-chain triglycerides (MCTs) and peptides, as well as carbohydrates metabolized to short-chain fatty acid (SCFA) (32). Enteral feeding is the most important factor to promote adaptation; nevertheless, it is important to avoid overfeeding, which may worsen fluid, minerals, and nutrients malabsorption, and may result in severe perianal skin lesions. Carbohydrate intolerance presents with frequent and liquid stools, the presence of reducing substances, and stool pH <6. Bile salts malabsorption should be suspected in children with no ileocecal valve and/or colon, high stool volume, and perianal maceration, especially if there are no stool-reducing substances. In these subjects, a trial of cholestyramine should reduce the diarrhea. Fluid losses in these patients are often accompanied by sodium loss and depletion; enteral sodium supplements should therefore be provided with the aim of maintaining urinary sodium >20 mmol/L, with a sodium/potassium urinary ratio of at least 2:1 (33). Once enteral feeding has been maximized, several complications may still continue or arise; small bowel bacterial overgrowth (SBBO) is common especially in patients with no ileocecal valve and those having abnormal motility. A common finding in these patients includes dilated loops of bowel containing residual nonabsorbed nutrients (34–36) (Fig. 1). SBBO can further cause mucosal inflammation and increased permeability that in turn may lead to sensitization and allergy as well as bacterial translocation, sepsis, and cholestasis (37–40). It has been shown in an experimental model that intestinal-derived lipopolysaccharide activates Kupffer cells through Toll-like receptor 4 (TLR4) signaling (41,42); activated Kupffer cells are probably involved in the pathogenesis of IF-associated liver disease (IFALD) (43). In some patients with dysmotile intestinal loops (intestinal atresia, gastroschisis, NEC), and liver disease, aggressive continuous tube feeding is often attempted with the aim of weaning a patient off PN, which may be the main cause of liver injury. Overaggressive feeds may lead to the loss of self-regulation of intake and may also result in abdominal discomfort, intestinal distension, and SBBO with subsequent further mucosal and liver injury (41–43). Surgical procedures such as tapering-lengthening or serial transverse enteroplasty aim not only to enhance the bowel length but also to reduce the diameter of dilated intestinal loops, thus improving its motility and reducing the risk for developing SBBO (44–47). A recent 5-year follow-up cohort study after serial transverse enteroplasty technique (STEP) confirms the efficiency of this procedure. Interestingly, intestinal absorptive capacity as assessed by D-xylose, a marker of carbohydrate absorption and mucosal integrity, and plasma citrulline, a marker of small bowel enterocyte mass, both increased significantly postoperatively (47). This suggests that STEP procedure, by reducing SBBO, restores small intestinal mucosa integrity and improves villous size within the first weeks following the procedure.
Clinical manifestations of feeding intolerance such as abdominal distension, bloating, and nausea, caused by colonic microbiological hypermetabolism, should be monitored. D-lactic acidosis may occur in some children during the process of bowel adaptation and even long after weaning off of PN. Indeed, Lactobacilli and other bacteria, including Clostridium perfringens and Streptococcus bovis, when present, may ferment nonabsorbed carbohydrate to D-lactic acid, which cannot be metabolized by L-lactate dehydrogenase (32,48). These microorganisms may proliferate in the acidic environment of the colon that is the result of the metabolism of unabsorbed carbohydrate to SCFAs. D-lactic acidosis presents with encephalopathy (ataxia, blurred speech, decreased consciousness) and should be considered when there is a high anion gap metabolic acidosis in the setting of normal serum lactate levels. Preventive measures for D-lactic acidosis include the reduction of carbohydrate intake, followed by antibiotics (such as metronidazole or cotrimoxazole) when dietary changes fail (49).
Special care should be taken to monitor children with SBS who are off of PN. They should be screened from time to time for nutritional deficiencies ((50) [references 51–136 are available as supplemental digital content at http://links.lww.com/MPG/A156]). Hyperphagia should be considered a sign of intestinal insufficiency and counteracted by feeding optimization or restoration of partial PN to preserve long-term growth (15,51).
Hormonal Therapy and Other Adaptive Treatments
Hormonal therapy is promising in the management of infants with SBS; nevertheless, the results of recent trials have largely reduced the enthusiasm around this therapeutic option (52–56). Recombinant human growth hormone has had inconsistent results and high rates of reported adverse effects in several published adult trials (54). A few studies of recombinant human growth hormone alone or in combination with glutamine have been carried out in PN-dependent children with SBS; despite some decrease in PN requirements during treatment, these trials showed little benefit on body composition and mucosal absorption in the long term (53,55,56).
Glucagon-like peptide 2 (GLP-2) is produced by the L-cells of the terminal ileum in response to luminal nutrients and has a trophic effect on the intestine, promoting absorption and adaptation (57). GLP-2 has been shown to increase the surface area of the gut mucosa, up-regulate nutrient absorption, improve gut-barrier function, increase intestinal blood flow, and decrease bone resorption. Patients with low levels of GLP-2 following the resection of the terminal ileum and/or the ileocecal valve improved intestinal absorption and nutritional status after treatment with GLP-2 (58). Other relevant treatments associated with a trophic effect on the bowel mucosa such as the SCFAs may be beneficial in children with SBS (59). Finally, there is also interest in the use of other trophic factors such as EGF and insulin-like growth factor-1 in children with IF and SBS (60).
Nontransplant Surgery for SBS
Several surgical strategies have been used in the last 30 years to improve the intestinal function in children with SBS having rapid transit time, dilated bowel loops, and insufficient absorptive capacity. Early experience has been reported with the technique called longitudinal intestinal lengthening and tailoring (LILT), and more recently with the STEP (61,62). Indications for bowel-lengthening surgery include the presence of a large intestinal diameter (>3 cm) for at least 20 cm of small bowel and a minimum total bowel length of 40 cm.
The LILT operation involves the longitudinal resection of a dilated loop of the small bowel between the peritoneal leaves of the mesentery; this gives origin to 2 hemiloops, each having its own blood supply. These are sutured longitudinally and anastomosed sequentially, doubling the length and halving the diameter of the operated loop (Fig. 2). The advantages of the LILT procedure include the maintenance of the normal orientation of the muscular fibers allowing more physiological peristaltic contractions, and the possibility to further perform a STEP procedure on the operated segments. The disadvantages are represented by the risks of vascular complications during the operation, making LILT more technically demanding as compared with the STEP procedure (63).
The STEP operation involves the use of a surgical stapler applied sequentially, from alternating and opposite directions to the dilated loop, in a transverse, partially overlapping, fashion creating a zigzag-like channel of approximately 2 to 2.5 cm in diameter (Fig. 3). This operation has the great advantage of being simple and reproducible. Recent evidence suggests that substantial benefits may result from a lengthening procedure in children with IF and SBS. These techniques also alleviate the consequences of severe bowel dilatation in children with SBS (64,65).
A review by Thompson and Sudan (45) recommended that surgical bowel lengthening should be considered in any chronically PN-dependent patient when there is substantial bowel dilation, regardless of the remaining bowel length.
Congenital enteropathies are a group of rare disorders causing IF in infancy and early childhood. Children with these disorders have early or even neonatal onset of chronic diarrhea that requires PN support (66).
The etiology includes defects in nutrient and electrolyte digestion, absorption, and transport, and disorders of enterocyte differentiation and polarization. Clinically, it is important to differentiate protracted from intractable diarrhea of infancy, the latter being irreversible. Figure 4 proposes a simple algorithm to approach newborns and infants with severe diarrhea.
The most common causes of intractable diarrhea of infancy are microvillus inclusion disease (MVID, also known as microvillus atrophy), tufting enteropathy (also known as intestinal epithelial dysplasia), syndromic or phenotypic diarrhea, and autoimmune enteropathy; the latter is not considered to be intractable unless all of the available treatments fail. Several genes responsible for these disorders have recently been identified by linkage studies based on genomewide analysis of polymorphisms, adding new tools for the diagnosis of intractable diarrhea of infancy (67).
In newborns with MVID, a congenital enteropathy caused by MYO5B mutation, severe watery diarrhea develops in the first few days after birth and can rapidly reach a total fecal output of 200 to 300 mL/kg of body weight per day. Diarrhea does not stop during fasting and causes life-threatening electrolyte and acid-base imbalances, rapid and severe dehydration, and hypovolemic shock. Children with MVID are invariably dependent on continuous PN; growth failure is extremely common and nearly half have some developmental delay. Many patients develop liver disease leading to intestinal and liver failure within the first few years of life. Presently, the survival rate for these children is approximately 70%, including those patients (up to half) requiring intestinal or liver-intestinal transplantation (68,69).
Similarly to children with MVID, infants with congenital tufting enteropathy (CTE) develop a severe neonatal diarrhea that persists through bowel rest. Often there is a family history of consanguinity and neonatal deaths related to severe diarrhea and dehydration. Indeed, CTE has been found to be associated with mutations in the gene encoding for epithelial cell adhesion molecule (EpCAM) (70). There are reported clusters of cases in the Arabic Gulf area (71). Infants with CTE typically experience a worsening of diarrhea during continuous enteral feeding, even when given extensively hydrolyzed or AABF, resulting in failure to thrive and protein-energy malnutrition. Diarrhea is usually less severe in infants with CTE than in children with MVID; some patients may be weaned from PN. Nevertheless, most remain PN dependent and eventually require ITx (72). Expert histological review of duodenal biopsies is the key to making the diagnosis of these 2 severe causes of IF in children (73). The so-called congenital sodium diarrhea has been reported as related to SPINT2 mutations (74). Clinical presentation, associated extraintestinal disorders, and histological features suggest a link between CTE and congenital sodium diarrhea (75).
Syndromic diarrhea (SD), also known as phenotypic diarrhea or trichohepatoenteric syndrome, is a rare congenital bowel disorder characterized by intractable diarrhea and woolly hair. SD has recently been associated with mutations in TTC37 encoding the uncharacterized tetratricopeptide repeat protein thespin (76,77); however, SKIV2L mutations have been recently reported in 6 individuals presenting with typical SD but having no variation in TTC37 (78).
This disorder is characterized by life-threatening diarrhea in early infancy, immunodeficiency, liver disease, trichorrhexis nodosa, facial dysmorphism including prominent forehead and cheeks, broad nasal root and hypertelorism, hypopigmentation, and cardiac defects. The patient's hairs are woolly, easily removed, and poorly pigmented. Severe and persistent diarrhea starts within the first 6 months of life in most cases and is accompanied by severe malabsorption leading to early and relentless protein energy malnutrition with failure to thrive. Liver disease, including extensive hepatic fibrosis and cirrhosis, affects about half of such patients. There are presently no specific biochemical profiles in these patients, although a functional T-cell immune deficiency with defective antibody production has been reported (79). Microscopic analysis of the hair shows twisted hair (pilitorti), aniso- and poikilotrichosis, and trichorrhexis nodosa. Histopathological analysis of small intestine biopsies shows nonspecific villous atrophy with low or no mononuclear cell infiltration of the lamina propria, and no specific histological abnormalities involving the epithelium. The pathophysiology remains unknown. Early management consists of total PN. Some infants have a milder phenotype requiring partial PN or only enteral feeding. Prognosis of this syndrome is poor, but most patients now survive, and about half of the patients may be weaned from PN at adolescence. Even treated patients have a final short stature (76).
INTESTINAL MOTILITY DISORDERS
Motility is essential to drive nutrients from the stomach to the colon. Normal digestion, absorption, and secretion of the bowel are not possible if motility is not adequate. This is why patients with similar residual lengths of the bowel or even a full-length gut may have vastly different outcomes in terms of intestinal sufficiency. In SBS, the motility of the residual bowel appears to be more relevant than its length to achieve adaptation (80). Children born with gastroschisis experience the consequences of injured intestinal motility from prolonged exposure to the amniotic fluid during prenatal life and, despite a normal bowel length, often do not achieve PN independence (81).
Intestinal motility is under the control of the enteric nervous system that is functionally independent from the central nervous system and is therefore efficient even in completely disconnected bowel loops, such as ITx. Normal motility is achieved through the transmission of the signals from the enteric nervous system to the enteric smooth muscle generating healthy peristaltic waves. Therefore, motility disorders may derive from either enteric nerve or muscle dysfunction. Although several GI conditions are classified among the motility disorders, only a few can lead to IF: gastroschisis, extensive Hirschsprung disease, and CIPOs. This review focuses on CIPOs, which represent the cause of IF in approximately 15% of all pediatric cases (1).
CIPO is a descriptive term pooling together several disorders of the enteric muscles or nerves. Thus, it may have heterogeneous features but has a similar phenotype characterized by recurrent bouts of intestinal obstruction without demonstrable mechanical occlusion.
CIPOs may be due to several diseases that can be either congenital or acquired. The most severe forms are usually congenital and present shortly after birth with episodes of intestinal obstruction with the surprising finding of normal intestinal appearance at laparotomy. Repeated such surgeries can negatively affect the course of the disease (82).
CIPOs have been conventionally divided into 2 groups, according to the pathogenesis of dysmotility: neuropathies and myopathies, the former caused by the involvement of the enteric nervous system and the latter by the dysfunction of intestinal muscles. CIPOs caused by muscle dysfunction are rare but seem to be more severe (83). Urinary tract disorders, such as megacystis and megaureter, can be associated with both neuro- and myopathies causing CIPOs. These should be managed by experienced urologists, although, surprisingly, they may be better tolerated than other more common obstructive urinary tract disorders (84).
The diagnosis of CIPOs is based on clinical factors. Careful assessment for mechanical causes of bowel obstruction must be pursued first. Tools helpful to suspect a severe motility disorder include radiological and histological evaluations and, if feasible, GI manometry (85); however, in our experience, intestinal manometry has never been conclusive for either the diagnosis or the optimization of the care of patients with CIPOS, whose management is mainly based on clinical and radiological features. The presence of a systemic autoimmune disease as well as severe infections and endocrinopathies suggests an acquired form of CIPOs that sometimes can be managed by treating the underlying illness.
In CIPOs, a plain abdominal x-ray typically shows air-fluid levels and dilatation of the bowel loops. Contrast studies, such as the barium small bowel follow-through study, are helpful to rule out mechanical obstruction, but may not reveal motility abnormalities. Transit may actually and surprisingly appear normal.
Congenital forms of CIPOs can be misdiagnosed as Hirschsprung disease, even resulting in surgery; however, surgical biopsies reveal normal enteric ganglia. In these cases, bowel resections should be avoided (82). When CIPO is strongly suspected, laparoscopic full-thickness biopsies may support the diagnosis with a minimally invasive procedure. Nevertheless, histological hallmarks are scant and the sample should be evaluated in referral centers by expert pathologists who have experience in similar cases and who have access to special facilities such as specific immunohistochemistry and electron microscopy, allowing the recognition of immune-mediated conditions, congenital neuromuscular disorders, and mitochondrial cytopathies (86).
Nuclear medicine may become a helpful tool to evaluate gastric emptying and intestinal motility in children. This test has the great advantage of being minimally invasive and reasonably standardized (87). Further studies, however, are required to validate such tests in children with CIPOs.
The management of IF caused by CIPOs should be based on a multidisciplinary intervention by medical, surgical, and allied professional personnel. Children with CIPOs almost invariably require some surgical intervention. The major barriers to food progression in a patient with fruitless propulsive strength are the natural GI tract bottlenecks: the pylorus and the ileocecal valve. These can cause a functional occlusion of the gastric outlet or small bowel clogging, which can be easily resolved by the formation of a gastrostomy (or jejunostomy) and an ileostomy, respectively. The formation of a stoma can improve the quality of life and reduce symptoms in up to 50% of children with CIPOs (81). Rarely, it is possible to localize the segments of the bowel more responsible for the dysmotility symptoms; in such cases, a loop resection can improve the intestinal transit and allow enteral nutrition and a return to a more normal life (82). Near-total small bowel resection has been proposed as treatment of CIPOs in some cases (88).
Optimizing nutrition in a child with CIPOs is challenging; it requires a skilled clinical evaluation and a thorough knowledge of all of the available tools. Different strategies should be tailored to each patient. Because of the heterogeneity of the syndrome, a key issue is to adapt the treatment/management to each individual patient according to age at onset, severity, and the outcome of surgical procedures such as a primary ileostomy (82). These children need to maintain the ability and the pleasure to eat normal food and this can be permitted by taking small and frequent meals with liquids or, in more severe cases, by using the gastrostomy as a venting device; the known benefits of delivering enteral feeding in children with IF make it mandatory to attempt intermittent gastrostomy closure and gastric or gastroduodenal low-fiber feeding (89).
Only a few medications have been shown to improve GI motility in patients with an intact enteric nervous system. Erythromycin at low or full antibiotic dosages may improve gastric emptying in children with CIPOs and gastroparesis (90). Several other drugs with a demonstrated effect on gastric motility, such as the serotonergic agents cisapride and tegaserod, have been withdrawn from the market because of the occurrence of rare but severe cardiac adverse events including arrhythmias, heart attacks, and strokes (90). Colonic acute pseudo-obstruction can be managed successfully by the infusion of the anticholinergic drug neostigmine, but this drug has not been tested in a long-term regime (91).
Children with CIPOs may experience SBBO and can thus occasionally benefit from a course of antibiotics such as metronidazole, aminoglycosides, or cotrimoxazole. These drugs should be prescribed only upon clinical suspicion rather than regularly, to avoid the emergence of bacterial resistance (34).
A French multicenter study including 105 children, 18 with prenatal diagnosis and 80 younger than 12 months at onset, showed that early age at presentation, PN dependency, and the number of surgical procedures were associated with a poor prognosis (92). In the most severe forms of CIPOs, children end up with an ileostomy, a gastrostomy with almost permanent aspiration because of gastroparesis, frequent bowel obstructions, and total PN dependency. Patients with such a poor quality of life may benefit from transplantation that should include the stomach (ie, modified multivisceral transplantation) (93).
LONG-TERM HOME PN
Home PN was first used in the early 1980s; since then, major advances have been made that allow for full nutritional support of children and adults with temporary or permanent IF at home. The improvement in the management of preterm infants and various treatments of children with severe GI diseases have led to new expansive use of home PN. Survival of children receiving prolonged PN depends mainly on the underlying diagnosis and has increased dramatically during the last 3 decades; nevertheless, complications such as catheter-related sepsis, liver disease, and loss of venous access can seriously challenge the clinical stability of patients with IF (94–96).
The expertise required to prescribe PN both at home and in the hospital usually comes from a dedicated hospital-based nutritional team who has a thorough knowledge of energy expenditure, nutrients and elements requirements by age, appropriate central catheter handling, and awareness of the risks and complications of long-term PN. Home PN must be tailored to the single patient and their families, always maintaining the goal of counteracting the deleterious aspects of IF (97,98).
Discussing the guidelines of PN in children is beyond the purposes of this review; official guidelines and position statements on central catheter handling and PN prescription have been published by the pertinent societies (99,100). We limit the discussion here to the issues on which recent research has given new insights that may lead to changes in the standard practices and recommendations.
IFALD is probably the most relevant and persistent complication affecting children with intestinal insufficiency and long-term PN. The prevalence of the disorder is unknown because there is no established definition of liver disease in this setting and it is unclear whether IFALD should be diagnosed on the basis of clinical, biological, or histological criteria. Indeed, there are insufficient data on the degree and type of liver involvement in patients with long-term PN (101,102).
The main factors contributing to liver injury in these patients are recurrent catheter-related sepsis, SBBO with bacterial translocation, and release of enterotoxins, which also negatively affect liver function, and a paucity of oral and enteral nutrition.
Factors affecting the onset and face of liver disease that are specifically related to PN are the inadequate supply of amino acids, the administration of excessive amounts of glucose, the duration of the infusion period, the inappropriate use of lipid emulsions, and finally micronutrient imbalances (Fig. 5). It should be stressed that the most important factors leading to IFALD are those related to individual patient characteristics and, importantly, the episodes of sepsis (103–106).
IFALD develops frequently at early ages, especially in premature infants in whom liver immaturity, frequent sepsis, and NEC facilitate liver inflammation and severe damage (105,106). At this young age, PN must be administered continuously for 24 hours each day and catheter-related sepsis is common.
An important role in this process is played by liver inflammation caused by extrahepatic infections in which microbial products brought to the liver through the bloodstream, either directly or through production of cytokines, lead to alterations of bile flow. The inflammation associated with these changes may cause rapid fibrosis and eventually biliary cirrhosis with end-stage liver disease (107).
A frequently cited observational study performed by Cavicchi et al suggested a link between fat intake and liver disease. In this retrospective trial of 90 patients with permanent IF, the likelihood of developing liver complications was tested against several clinical features; the authors found that the duration of PN, the dose of lipids (triglycerides rich in long-chain ω-6) >1 g · kg−1 · day−1, and a length of functional bowel <50 cm were factors strongly associated with the onset of chronic cholestasis (108).
Colomb et al found that a reduction or the discontinuation of the administration of soybean-based lipid emulsion along with the addition of α-tocopherol was associated with reversal of cholestasis. Interestingly, this strategy seemed also to increase the platelet count (109).
Extremely similar results were obtained in the study by Ganousse, who showed that the improvement of cholestasis depended also on maintaining an appropriate protein/energy ratio in PN, achieving cyclic rather than continuous PN infusion and with the use of MCTs plus α-tocopherol (110).
Recent data suggest that IFALD is a multifactorial disease in which the use of soybean oil–based emulsions in PN may represent the major culprit (111). Several factors should be taken into consideration when choosing a fat emulsion for parenteral use: the content in essential fatty acids (EFAs), the ratio of ω-6/ω-3, the polyunsaturated fatty acid (PUFAs) content, the amount of MCTs, and the quantity of α-tocopherol and phytosterols.
The probable detrimental effect of ω-6 FAs on liver function is provided by studies that showed that fat emulsions based on pure fish oil (containing ω-3 FAs) have been successful as rescue therapy in pediatric patients with SBS affected by severe liver disease (112). The infusion of exclusively ω-3 FAs ultimately changed the management of these patients because it allowed the reduction of the intake of proinflammatory ω-6 and phytosterols while increasing the amounts of α-tocopherol, a powerful antioxidant (101,107,113,114).
The evidence gathered on the beneficial effects of fish oil in these patients has led to its use in clinical practice; however, 2 different approaches have been developed in the United States and Canada as compared with Europe. In North America, only a pure fish oil solution (Omegaven, Fresenius Kabi, Bad Homburg, Germany) is available on the market, whereas in Europe, it is possible to use also an emulsion containing a mixture of soybean oil (30%), coconut oil (30%), olive oil (25%), and fish oil (15%) (SMOF-lipid, Fresenius Kabi, Bad Homburg, Germany).
Conversely, some concern has been raised about providing fish oil as the sole source of lipids for a long period of time. Fish oil provides less essential ω-6 fatty acids (ω-6 FAs) than that presently recommended in infants and young children (101,111). Furthermore, Omegaven (pure fish oil) can only be given at lower infusion rates compared with SMOF-lipid. Omegaven may not be able to provide enough energy to sustain growth. Thus, the combination of several types of oil by mixing soybean oil (rich in ω-6 FAs), coconut oil (rich in MCTs), olive oil (rich in MUFAs), and fish oil (rich in ω-3 FAs) appears to promote better growth while limiting hepatic toxicity (115). Phytosterols contained in soybean oil have been found to be associated with liver disease progression, and their exclusion from intravenous lipid emulsions may also be beneficial in children receiving PN. Clayton compared the level of phytosterols in plasma of healthy subjects, patients with mild hepatic dysfunction, and those with severe dysfunction who received soybean oil emulsion (rich in sterols), and found a link between liver damage and phytosterols plasma levels (116–119).
Regarding the presence of tocopherol in lipid emulsions, we should emphasize that there are different preparations of tocopherol; α-tocopherol is the form with far greater antioxidant activity. Although soybean oil emulsions contain a high amount of γ-tocopherol (which has 25% of the antioxidant power as compared with α-tocopherol), lipids based on fish oil are rich in the most powerful antioxidant vitamin E, α-tocopherol (114,120,121). To ensure a proper antioxidant power in lipid preparations, it is advisable to add 0.5 mg of α-tocopherol per gram of PUFAs (101,102).
SMOF-lipid emulsion contains soybean oil (30%), coconut oil (30%), olive oil (25%), and fish oil (15%), with an important addition of α-tocopherol (200 mg/L), equal to that found in the emulsion based only on fish oil (Omegaven) (115). A randomized, double-blind, controlled trial on 60 preterm babies stratified by body weight has analyzed a set of parameters (clinical data, laboratory data, FAs in plasma and red blood cells, plasma levels of α-tocopherol and phospholipids) after infusion of PN with SMOF-lipid or soybean oil–based emulsion. The SMOF-lipid emulsion increased the content of eicosapentaenoic EPA and docosahexaenoic acids (DHA) and reduced the ω-6/ω-3 ratio, improving also liver function tests (122).
Another study evaluated the long-term effects of the lipid mixture SMOF-lipid versus a soybean oil–based preparation in pediatric patients receiving home PN. This randomized, double-blind study involved 28 children who received >4 infusions of PN per week for 4 consecutive weeks. The infusion was administered in 12 to 14 hours overnight. At the end of the study, no differences between biochemical and nutritional outcomes were recorded, but there was a clear association between the use of SMOF-lipid and a significant decrease in bilirubin levels, which conversely increased in the soybean oil–based group (123).
A confirmation of these findings comes from the study of Muhammed et al who examined the effect of the switch from a soybean-based lipid emulsion to SMOF-lipid in 17 children with cholestasis. The subjects were assigned to a treatment group receiving SMOF-lipid and a group receiving soy-based lipids. During a period of 6 months, the use of SMOF-lipid was associated with a marked statistically significant reduction in the levels of bilirubin when compared with the soy-based lipid group (124).
We can therefore conclude that recent studies have emphasized the superiority of fish oil–derived lipid emulsions as a major advance for the management of patients receiving long-term PN. Preparations with pure fish oil are effective in improving cholestasis, but their use as the sole source of lipids may not provide enough evidence to meet EFAs requirements, especially for long-term use (125). Nevertheless, although some randomized controlled trials have demonstrated the beneficial effect of SMOF-lipid versus soy-based lipid emulsion (122,123), no studies have compared SMOF-lipid with Omegaven in these patients.
THE IMPORTANCE OF A MULTIDISCIPLINARY TEAM
Pediatric IF is a multifaceted condition requiring the competent contributions of several medical and allied health professionals for inpatient and outpatient care. Therefore, the formation of a multidisciplinary team is vital to achieve optimal results (126).
The IF team should ideally include a staff specializing in surgery, gastroenterology, and nutrition, a pediatric dietician, and a nurse experienced in handling central venous catheters and PN infusion. Special consideration should be given to the link between the hospital team and the home care team. Fostering coordination of surgical, medical, and nutritional management is vital to provide high-quality, integrated care of patients with IF, thus improving remarkably the survival of these patients (127). A study by Beath et al has demonstrated that the 3 most important issues in the management of children with IF include a good and early link between primary care givers and IF programs, the presence in the program of both intestinal rehabilitation and intestinal transplantation expertise, and participation in the network of the organizations providing home PN solutions. Collaborative strategies must be developed to reduce mortality and morbidity in patients with IF, especially for those who are referred for permanent IF or intestinal transplantation (128).
NUTRITIONAL FAILURE AND REFERRAL FOR INTESTINAL TRANSPLANTATION
Although a large percentage of children with IF can survive with long-term PN, a proportion of patients eventually develop life-threatening complications such as severe septic episodes, fluid and electrolyte imbalance, loss of venous access for PN, and end-stage liver disease (1,129). In these patients, nutrition has failed both in the enteral and in the parenteral routes. We refer to such patients as patients with “nutritional failure.” These patients should be referred for transplantation.
Nevertheless, relatively few advances have been made in the field of intestinal and multivisceral transplantation in the last 10 years, with no significant improvement in the long-term patient and graft survival. According to the ITx registry, approximately 2500 ITxs have been carried out so far in 79 worldwide transplant centers, of whom half are alive. Among 1351 transplanted children, the 5- and 10-year graft survival rate is reported as approximately 50% and 30%, respectively; the 5- and 10-year patient survival rate is similar, approximately 50% and 30%, respectively. In patients with a functioning graft, approximately 60% have normal function, whereas 40% require partial PN or intravenous fluids (130). These sobering figures mandate the adoption of all of the relevant strategies to avoid intestinal transplantation until new protocols to achieve a better outcome are available.
There is probably a different threshold for intestinal transplantation on both sides of the Atlantic Ocean. In accordance with the European approach, we are more reluctant to refer a child for ITx and are inclined to support long-term home PN, which is cost-effective and provides a better quality of life. Support for this view comes from Pironi et al, who recently performed a 3-year prospective study (including both adults and children) on long-term PN for IF, and compared “noncandidates” for ITx (no indications nor contraindications), with “candidates” (who had an indication according to the Center for Medicare and Medicaid Services definitions, and a high risk of death or morbidity according to the American Society of Transplantation position paper) (131,132). The results showed that only patients with nutritional failure caused by IFALD or major catheter complications had an increased risk of death while receiving home PN, thus supporting its use as the primary treatment for IF. Therefore, it was suggested that ITx should be used only as a lifesaving procedure. Although experienced transplantation centers have suggested that the role of ITx should be expanded to a preemptive/rehabilitative procedure applicable to all patients with irreversible IF (133), further recent findings have shown that home PN is the treatment of choice for IF in adults as well as in children (134–135). We therefore believe that only nutritional failure should be regarded as a clear indication to intestinal transplantation.
In children, an early referral is essential to optimize the long-term outcome of severe IFALD. Central venous catheter–related major complications may be indications for a preemptive intestinal transplantation in selected patients (136).
In conclusion, treatment of permanent IF has made remarkable strides in the last few decades. The establishment of multidisciplinary intestinal rehabilitation programs at leading centers has improved the survival of children with IF, whereas the morbidity associated with both IF and PN has significantly decreased. Recent advances in the knowledge of factors implicated with PN complications and improvements in the medical and surgical management of SBS will result in better outcomes for these patients in the near future. Isolated liver Tx for patients with SBS who have the potential of bowel adaptation is no longer required. It is interesting to note that the most recent International Intestinal Transplantation Registry report at the XII Small Bowel Transplant Symposium (Washington DC, September 2011) showed early evidence of a worldwide trend of 20% reduction in the number of pediatric ITx (130). Major efforts are needed to improve the outcome of intestinal transplantation that will likely remain part of the armamentarium required to prolong the survival of children with life-threatening complications of IF. Nevertheless, the European experience has led to support a more conservative approach more inclined to home PN, limiting referrals for transplantation only to children with nutritional failure.
We thank Steve Hardy, MD, from the Children's Hospital at Harvard University, for help in reviewing the manuscript.