Nutrition and Intestinal Rehabilitation of Children With Short Bowel Syndrome

Classification and Etiology of Short Bowel Syndrome

Exact measurement of the remnant intestine remains difficult even with the help of radiographic assessment (14.). The residual bowel length can be expressed as the percentage of the theoretical small bowel length (15.) instead of the actual length remaining.

Classification of SBS into 3 types is helpful for understanding the different outcomes: type 1 with end enterostomy or entero-rectal anastomosis with absence of the colon, type 2 with jejuno-colic anastomosis, and type 3 with jejuno-ileocolic anastomosis preserving the ileocecal valve (ICV) and the colon (16.) (Fig. 1).

In children, the conditions that most commonly lead to extensive small bowel resections are necrotizing enterocolitis (NEC), midgut volvulus, gastroschisis, intestinal atresia (including apple peel syndrome), and extensive aganglionosis. Total/Near Total intestinal aganglionosis (TIA/NTIA) is an uncommon and life-threatening form of Hirschsprung disease (HD). Most of the time, neonates with TIA/NTIA have SBS type 1 with less than 50 cm of normally innervated small intestine beyond the ligament of Treitz (2.) (Table 1 and Fig. 1).

TABLE 1. Causes of short bowel syndrome in childhood

Numerous factors determine SBS prognosis and duration of PN dependency: the underlying diagnosis, the type and condition of intestinal segments preserved, the presence/absence of the ICV and/or colon, a long-term enterostomy versus a primary anastomosis, any associated motility disorders, especially in intestinal atresia and gastroschisis (17.), the number of surgical procedures, as well as the age of the patient at the time of surgery (18.,19.). However, information about long-term growth and final adult height is lacking.

Epidemiology

Accurate estimates of neonatal SBS incidence are difficult to determine because of variation in the definition of SBS between institutions, difficulty of tertiary care referral centers to accurately determine their referral population, and problems ensuring complete follow-up of the cohort. Those difficulties are indicated by differing estimations of the incidence of SBS by different teams around the world.

The Canadian Collaborative Study Group estimated a neonatal SBS prevalence of 4.8/1 million population with an incidence of 24.5 per 100,000 live births [95% confidence interval (CI) 12.1–36.9] that was much higher in infants born before 37-weeks’ gestation compared with term newborns (353.7/100,000 live births vs 3.5/100,000 live births) (20.). In a report from the beginning of the century, overall mortality rate of neonates with SBS was 37.5%, 3 times higher than that of a matched cohort of infants without SBS. Intestinal failure associated liver disease (IFALD) was the cause of death in 60% of SBS (21.); this picture has changed due to advance in PN support (22.).

In the United States, the annual incidence was estimated to be 3–5/100,000 live births with a mortality rate of 26% (13.).

In France, the prevalence was estimated according to the number of patients on long-term home-PN. The French Home-PN registry demonstrated a prevalence of IF from all etiologies of around 10/million individuals <18 years of age and an annual incidence around 20–30 new cases per year (9.). Among this total population of children on home-PN 40%–50% were suffering from SBS. The reported mortality rate was 3.6% including all home parenteral nutrition (HPN) patients irrespective of the cause for long-term IF (9.).

Furthermore, the most recent report from the British Association of Parenteral and Enteral Nutrition (BAPEN) reported a steady increase of children with SBS-IF in the United Kingdom from 18 reported in 1992 passing through 99 reported in a single point prevalence survey in 2012 to the final number of 185 reported from the prospective registry in 2019 (23.)

Consequences of Intestinal Resection

SBS is characterized by a compromised intestinal absorptive capacity due to a severely reduced mucosal surface area resulting in diarrhea, fluid and electrolyte imbalance, and failure to thrive (5.). The functional consequences of SBS depend not only on the length, surface area, and site of the resected small intestine but also on the functional capacity of the remaining intestine (Table 2). The cause of resection and age of the patient at the time at which surgery was carried out also influence the capacity of the remaining gut to function and the potential for adaptation. Term neonates have a small bowel length of 250 ± 40 cm which doubles during the last trimester of gestation, suggesting that a short bowel remnant does not have the same prognosis in a preterm infant as in a full-term baby (15.). The same difference exists between neonates and older children. Thus, if the surgical resection occurs earlier, the opportunity for adaptive growth is greater (5.). The severity of intestinal resection can be classified according to the length of the small intestine measured along the anti-mesenteric border at surgery. Although, currently a clear definition of SBS according to length is lacking, the majority of recent literature defines ultrashort bowel syndrome as a remnant small bowel of <25–10 cm (or <20%–10% of expected small bowel) (24.-26.). Other determinants of IF in SBS are the portion of the small bowel resected (jejunum, ileum, ileocecal valve, colon) (Fig. 2), and the functional integrity of the residual intestine (27.-29.).

TABLE 2. Consequences of different intestinal resections

Pathophysiology of SBS, including the consequences of anatomical changes, has been extensively reviewed (30.).

Adaptation After Extensive Intestinal Resection

Soon after small bowel resection, the physiological process of adaptation of the remaining small bowel starts. Adaptation is a slow process accompanied by a gradual increase in absorption capacity for nutrients, electrolytes, and minerals (25.,30.). This comprises muscular hypertrophy (increased bowel diameter and wall thickness) and hyperplasia of the intestinal mucosa. This mucosal hyperplasia is characterized by an increased number of enterocytes per unit of SB length, an increased rate of enterocyte proliferation, and an increased villous height and crypt depth (64.). In animals, it was shown that epithelial hyperplasia following gut resection results in increased mucosal mass, including higher mucosal wet weight, higher protein content as well as higher DNA and RNA content per unit of bowel length which results in increased number of enteroendocrine cells (65.,66.).

The complex regulation of gut mucosal growth involves a multitude of factors including hormonal mediators such as enteroglucagon, glucagon-like peptides neurotensin, peptide YY (PYY), growth hormone, and insulin-like growth factor. Enteroglucagon, progenitor of glucagon peptide 1 (GLP-1) and 2 (GLP-2), and PYY seem to play a pivotal role by stimulating cell turn over, motility and absorptive capacity of the small intestine especially after small bowel resection (67.-69.).

Nutritional substances in the lumen are central in this process and they can act directly stimulating mucosal hyperplasia and expression of their own transporters (70.) or indirectly by modifying microbiota composition and stimulating trophic gastrointestinal hormone secretions (71.).

Major nutritional adaptative substances:

Besides nutrients, other factors seem to play an important role in the mucosal adaptation process:

Thus, adaptation of the remaining small bowel is dependent on the tight interaction of endogenous (enterohormomes, EGF, polyamines) and exogenous (nutritional substances) factors; this emphasize the usefulness of early EF after resection (82.). The goal of intestinal adaptation is to develop the ability to withdraw artificial nutrition thanks to a compensatory increase in the mucosal surface area and absorption capacity. The 2 indicators which have historically been used to predict weaning from PN in children: anatomy (residual bowel length measured at most recent major surgical resection and proportion of preserved colon) and plasma citrulline may be reliable prognostic factors in SBS patients (16.,83.-89.). Recently, monitoring citrulline plasma levels during the course of SBS adaptation has been shown to predict PN weaning (90.). In clinical practice the degree of intestinal insufficiency may be measured by both citrulline plasma levels and the percentage of PN required for normal body weight gain and longitudinal growth (16.,90.). PN requirements remain, therefore, the best measure of the degree of intestinal sufficiency in this setting (Table 3).

TABLE 3. Setting the scene for pediatric short bowel syndrome

Parenteral Nutrition Support

How to Deliver Enteral Feeds After Surgery?

Prospective randomized trials on when to start EF and the speed of increment of enteral feeds in high-risk patients are not currently available. Therefore, recommendations are primarily based on physiology and clinical experience. Management should be adapted and tailored on a case-by-case basis (117.). As stated earlier enteral nutrition plays a pivotal role in intestinal adaptation, prevention of cholestasis, and maintenance of a healthy microbiota in children with high risk (118.). This is true also for preterm newborns (119.). In particular 2 metanalyses have been performed on refeeding practices after NEC (120.,121.). These found that EF within 5 days of NEC diagnosis did not seem to be associated with adverse outcomes, including NEC recurrence.

Even after emergency GI surgery, early feeding may be feasible in patients without severe shock or bowel anastomosis instability. Early feeding was defined as initiation of EF within 48 hours after surgery (122.). For all those reasons, even if based on low quality evidence, starting EF as soon as possible within 48 hours after surgery has been recommended (92.). Most centers start feeding as soon as stool was passed (123.). Gastric output in this specific context may not reflect appropriately enteral nutrition tolerance since gastric hypersecretion is part of the intestinal adaptation process after resection (56.).

The resting and adaptation phase progression/alteration of feeds depends on the anatomy of the residual small intestine and the presence of the colon and should be guided by GI tolerance (stool volume, lack of nappy rash/perianal excoriation, abdominal comfort, etc) and PN weaned while allowing for growth and biochemical stability (12.).

There is very little evidence regarding the route of enteral nutrition administration (oral or artificial feeding device) and the mode of administration (continuous vs intermittent) (Table 4). Oral feeds are more physiological and should be offered whenever possible, while aggressive weaning and gut overload syndrome should be avoided (8.,12.,124.). Oral feeding promotes the release of epidermal growth factor (EGF) from salivary glands, increases GI secretion of trophic factors, and helps prevent feeding disorders (125.). For the aforementioned reasons, in the attempt of pursuing a more physiological mode of feeding, oral feeding should be the preferred route whenever possible after surgery. If the oral route is not accessible (eg, mechanical ventilation, clinical deterioration, immaturity) it is of utmost importance to promote complementary oral stimulation as soon as possible in order to prevent oral aversion and feeding disorders (11.). In some selected cases mucous fistula refeeding might be considered even if an international consensus is still lacking (126.).

TABLE 4. Advantages of different modes of enteral feeding.

A recent survey from the European Reference Network of Intestinal Congenital Abnormalities (ERNICA) consortium outlined the various clinical practices regarding early feeding in intestinal rehabilitation (123.) such as intermittent feeding use in half of the screened centers and the rest divided into continuous or a combination of both, when the oral route is not possible. On one hand, continuous provision of enteral nutrition may increase mucosal contact and improve receptor saturation, therefore, potentially improving absorption per unit of length of intestine (127.). However, continuous tube feeding eliminates fasting periods and may suppress neuromuscular function with a decrease of intestinal motility and thus, increasing the risk of SIBO and IFALD (12.).

There is no evidence on how to progress with EF but in clinical practice stool output is used as a proxy for intestinal adaptation and enteral nutrition tolerance. Some experts recommend that the upper acceptable limit of stool output is 40–50 mL/kg/day (10 bowel movements) (128.) while other experts recommend an upper limit of 20 mL/kg/day (6 bowel movements) (129.). Even without strong references, stool output should be interpreted according to the anatomy of the remnant intestine and SBS type. The ERNICA questionnaire (36 questions) about strategies used to introduce enteral nutrition postoperatively and start complementary food/solids in infants with SBS associated IF was developed and sent to 24 centers in 15 countries. There was wide variation in early EF practices in infants with SBS. Ten of 24 centers in the ERNICA survey (41.7%) increased EF by 10–20 mL/kg/day.

Recently Channabasappa et al (130.) published an algorithm for the increments of feeds considering stool output and stool frequency which was included in Figure 4.

In rare cases of children developing SBS-IF at an older age the diet regime applied before intestinal resection should be resumed as soon as possible after surgery taking into consideration the new intestinal anatomy and feeding tolerance.

Choice of Feeds

The decision of what type of enteral feed to initiate is controversial in pediatric SBS patients and depends on several factors. The age of the child, the cause and type of SBS (Fig. 1), as well as the intestinal motility are relevant issues. Although data are limited, HM (breast milk or donor milk) is recommended as a first-line enteral regimen for infants with SBS-IF, as it contains immunoglobulins, prebiotic substrates such as HM oligosaccharides, bioactive proteins, and growth factors (growth hormone and EGF) that together promote a healthy microbiome and intestinal adaptation (92.,130.). In addition, it has been hypothesized that the use of HM may result in fewer days of dependence on PN (131.) and may reduce the risk of IF-associated liver disease (132.). In the already mentioned ERNICA survey, maternal breast milk was the first choice in 23 of 24 centers, while donor breast milk, standard preterm/term formula, extensively hydrolyzed, or amino-acid based feed were the second choice. Furthermore 22 of 24 centers introduced weaning foods by 6 months of age. However, what type of complementary foods were used varied significantly among the units and seem to reflect cultural differences amongst European countries. Fifteen (62.5%) centers in the ERNICA survey used nutritional fortifiers.

When HM is not available various different types of formula have been suggested, ranging from polymeric formulae to extensive hydrolyzed formulas (EHFs) or AA-based types (133.) (Table 5).

TABLE 5. Summary of pros and cons of different types of feedings

Extensive hydrolyzed proteins, which provide dipeptides as well as AA, confer a significant absorptive advantage over AA formulas alone (134.). Most of EHFs do not contain lactose and usually have an increased concentration of MCTs. Components such as MCT can be better absorbed in the case of rapid transit, bacterial overgrowth, and bile acid depletion, which occurs after surgical NEC (135.). For this reason, hydrolyzed protein formulas have been popularized for use in patients in whom absorptive capacity is limited (136.). Nonetheless, the possible disadvantages of EHF should be considered, since they do not satisfy the elevated requirements in preterm infants (137.). Furthermore, studies in animals suggest that intact macronutrients (e.g. long-chain fatty acids) lead to better intestinal adaptation than do medium-chain fatty acids (138.). In this regard, the only randomized study comparing hydrolyzed versus non-hydrolyzed formula in children with SBS is an old trial involving only 10 children and designed to compare intestinal permeability (136.). This trial did not find any significant difference between the 2 groups.

Children with SBS, especially young infants, may have a significant risk of secondary non-IgE-mediated intestinal allergic disease (139.). It is thought that decreased intestinal barrier function, especially caused by SIBO may predispose infants with intestinal failure to allergic gastrointestinal diseases (140.). Children with SBS often have small intestinal dilatation and dysmotility, which results in SIBO. Secondary inflammatory changes in the mucosa may occur which result in permeability to macromolecules, sensitization, and secondary allergic enteropathy. In a retrospective review of 37 children with neonatal SBS over a 10-year period 3 patients (<10%) developed cow’s milk allergy (CMA) (139.).

In this specific setting, when based on clinical parameters an allergy is suspected; EHF and even AA formulas may have a significant therapeutic advantage in the management of SBS in patients younger than 2 years; these advantages are independent of the absorptive differences and are primarily related to the anti-allergic properties of the extensively hydrolyzed proteins or AAs (141.). Many studies involving HFs and AA formulas are small and retrospective, and there is no definitive evidence that one type is better than the other (142.). Capriati et al recently published a review of the enteral nutrition in neonatal SBS. They included 10 studies (822 patients) separated into 3 cohorts: EHF fed patients, EHF + HM-fed patients, and AA + HM-fed patients. There was no difference in the rate of enteral adaptation or PN duration among the groups. They concluded that HM should be used when possible in SBS patients and that HF is just as effective in adaptation as AA-based formulas (143.). However, AA formula does not contain high amounts of MCT as compared with the commercially available EHFs that usually contain an energy supply of 50% lipids of which 40%–60% is MCT.

Currently, there is no evidence to recommend HM fortifiers for SBS children; fortifiers may alter enteral feed osmolarity (144.) possibly inducing increased stool output.

Family Communication

Despite being a chronic condition, which frequently impacts not only on physical but also on psychosocial and emotional functioning, literature on the role of counseling/information and proper communication is scarce in this setting. A recent cross-sectional study on children and caregivers perspective of SBS demonstrated a lower degree of health care satisfaction, mainly due to lack of information and communication in children older than 5 years (145.).

As highlighted in a retrospective study surgical counseling may be necessary even prenatally because of the frequent association between prenatal diagnosis of congenital intestinal abnormalities and SBS (146.,147.). In other non-congenital conditions such as NEC or volvulus acute counseling may be necessary and should involve information on prognosis (29.,148.,149.) and possible complications (150.-152.).

After intestinal surgery an appropriate discussion with the family should take place to explain SBS diagnosis and include information on prognosis (including expected weaning possibilities) (25.,153.), impact on daily family care [impact daily assistance and quality of life (QoL)] (154.-156.), and possible complications (157.,158.).

For all those reasons, in order to build a sustainable long-term relationship with SBS children and their families, we strongly suggest not to neglect 2 important steps of communication:

Discharge and Family Education

Retrospective collection of data from children suffering from IF has demonstrated that HPN may be associated with less IF-related complications and improved QoL compared with hospital PN (159.). Furthermore, accessing HPN program seems to improve children’s QoL by allowing them to participate in the same activities as their healthy peers (160.,161.). For those reasons, HPN is now considered the standard of care for patients with prolonged/irreversible IF and should be considered for all children likely to require PN for more than 12 weeks (3.). However, HPN may have some limitations, for example, the ability to perform rapid changes in PN formulation depending on the specific HPN program (159.). Children considered for home discharge should thus have sufficiently stable fluid and electrolyte needs and tolerate cyclical PN while maintaining blood glucose levels.

Retrospective analysis of patients on HPN has demonstrated a 30% rate of hospital readmission within 30 days of discharge both for adults and children (162.,163.), even if only 20% of those readmissions were related to IF complications. Specific studies on HPN children are lacking even if we know from clinical practice the high burden of medical care of those patients.

The positive impact of a structured training program in preventing end stage nutritional failure was demonstrated in a retrospective cohort of children listed for intestinal transplant (164.,165.). This comes from retrospective data and we are not aware of any prospective data from small centers where a structured training program for intestinal rehabilitation is not available (166.). Despite limited evidence, we strongly recommend a structured training program to teach families how to manage HPN, in order to prevent hospital readmissions and complications. At least one (ideally both two) parent(s)/caregivers should have been trained by a nutrition nurse and the home has to be suitable for HPN (3.). Families should be carefully prepared for discharge on HPN, with extensive explanation about safe medical environment, hand hygiene, central venous catheter (CVC) care, HPN preparation, infusion pump usage, and infection control (163.).

Recently, various methods have been described in order to improve CVC and PN management such as online and video simulation which may be used in addition to face to face training for safe, on-demand rehearsal of techniques such as accessing the CVC for blood sampling, connection/disconnection of the PN bag, dressing change of the catheter exit site, and pump operation (167.-169.).

As stated in the ESPGHAN pediatric PN guidelines and confirmed by recent evidence (170.-173.), taurolidine locks should be in use in long-term PN (174.). When taurolidine is not available, ethanol locks may be considered (174.).