Prematurity

One of the earliest factors identified as a risk for IFALD is prematurity. Using low birth weight as a correlate of gestational age, Beale et al. [18] reported a 50% incidence of parenteral nutrition associated liver disease (PNALD) in infants born at <1,000 g, a five-fold increase in risk compared to infants born at >1,500 g. This finding has been supported in multiple studies [4, 19]. In a retrospective, multicenter analysis of 1,366 neonates who received ≥14 days PN, Christensen et al. [4] found that the odds ratio of developing PNALD was 30.7 in neonates with birth weight <500 g and 13.1 in the <750 g subgroup.

There are several potential factors to explain why the premature or low birth weight neonate may be at higher risk for IFALD development [1]. Immature intestinal barrier function and enterohepatic circulation of bile acids in the neonatal liver may predispose to cholestasis [20]. At a molecular level, this is likely related to developmentally regulated expression of nuclear receptors (e.g., farnesoid X receptor (FXR)) and their effect on expression of bile acid transporters [21]. Accumulation of toxic bile acids may precipitate secondary oxidant injury, to which the neonatal liver may be more predisposed because of antioxidant and other deficiencies (such as glutathione deficiency) [22].

Sepsis

Sepsis, typically related to venous catheter infections, is common in the pediatric intestinal failure population. A multicenter, retrospective analysis of 272 infants with IF found that 3.3 new catheter-related bloodstream infections occurred per person year in infants on PN [23]. The occurrence of infection has been well demonstrated as a key risk factor in the development of IFALD [19]. In a retrospective analysis of 42 post-surgical infants on PN ≥3 months, Sondheimer et al. [7] reported that the majority of patients who developed cholestasis did so within 14 days of an episode of infection and that patients who developed early infections (<6 weeks of life) were most likely to develop cholestasis. This clinical correlation between age of first infection and severity of IFALD was further supported histologically by Hermans et al. [24] who found, in comparing infants with IFALD that had severe liver fibrosis (n = 16) vs. normal/mild fibrosis (n = 14), incidence of infection was significantly greater and age at first infection was significantly lower in the severe fibrosis subgroup.

The relationship between sepsis and cholestasis is well established, and there are several pathophysiologic mechanisms that support this clinical association. Within the liver, circulating endotoxin (lipopolysaccharide) from bacteremia, as well as absorbed bacterial cell wall products through an abnormal mucosal barrier, activates Kupffer cells, stimulating the release of proinflammatory cytokines which generate an inflammatory cascade. Kupffer cell activation likely occurs through endotoxin binding to membrane receptors causing induction of TLR-4 dependent pathways [25]. At the hepatocellular level, endotoxinemia and subsequent release of cytokines by macrophages lead to down-regulation of bile acid transporters. On the basolateral hepatocyte aspect, decreased expression of NTCP and OATP lead to decreased uptake of circulating bile acids. On the apical canalicular level, decreased expression of BSEP and MRP-2 lead to accumulation of bile acids and other organic ions within the hepatocyte, stimulating further downstream hepatotoxic effects.

Intestinal and Biliary Stasis

Several clinical factors precipitate alterations in normal gastrointestinal motility in IF patients. There is often a requirement for prolonged lack of enteral feedings as a result of congenital intestinal malformations or post-surgical factors. Feeding intolerance due to malabsorptive diarrhea or dysmotility is common in the patient with SBS, and this may require interval periods of decreased or absent enteral feeds. As the foreshortened bowel undergoes adaptation, there is compensatory luminal dilation to improve the absorptive surface area. When the bowel is overly distended, motility is altered, promoting stasis. Alterations in enteral feeding lead to decreased circulation of enteral hormones, including cholecystokinin (CCK), glucagon, gastrin, and enteroglucagon [1]. Decreased secretion of CCK and other regulatory gastrointestinal hormones predisposes to stasis within the biliary tree as well as the intestinal lumen. Precipitation and formation of biliary sludge, exaggerated by anatomic alterations in enterohepatic circulation, namely in the patient with ileal resection, also contribute to the propensity for cholelithiasis and IFALD development.

In addition to affecting gut motility, intestinal stasis and post-surgical anatomic alterations (absence of ileocecal valve, bowel dilation, anastomotic narrowing or dysfunction) predispose the patient with IF to small bowel bacterial overgrowth (SBBO) and intestinal dysbiosis. Bacterial overgrowth is an important factor in the clinical management of patients with IF. Because objective evaluations for SBBO are difficult to obtain and interpret, there are no conclusive studies that directly link SBBO as a causative factor in IFALD. Theoretically, however, the accumulation and overpopulation of enteric organisms, often gram-negative bacteria, lead to enteritis, altered intestinal permeability with absorption of bacterial cell wall products (pathogen-associated molecular patterns), and bacterial translocation. Translocation, in turn, is likely to drive clinical episodes of gram-negative sepsis (catheter-associated bloodstream infections) in the IF patient. There may also be subclinical translocation of bacteria, without frank bacteremia, that further drives the activation of inflammatory cascades and the down-regulation of bile acid transport pumps within the hepatocyte and liver.

Underlying Surgical Disease

Necrotizing enterocolitis (NEC) is the most common etiology of SBS in the pediatric population. NEC occurs in premature infants, often presenting with acute, severe bowel necrosis and secondary peritonitis. The premature infant with compromised bowel and NEC may require emergent surgical management, bowel resection, and often prolonged periods of limited enteral feeding. Secondarily, the patient with SBS following NEC has prolonged PN requirement starting at an early age, typically initiated in a setting of peritoneal contamination or systemic sepsis, and continued in the setting of anatomically disrupted bowel integrity and impaired motility. It is no surprise, then, that the occurrence of severe NEC directly correlates with incidence of IFALD [26].

In addition to NEC, infants with gastroschisis and prolonged PN requirement have been implicated as having a higher risk for IFALD development and progression [4]. The patient with gastroschisis, especially with concomitant intestinal atresia and SBS, may have significantly dilated and often dysfunctional bowel, further predisposing to SBBO, altered intestinal permeability, PN requirement, and sepsis.

At present, there are no studies that conclusively link a defined post-surgical anatomy (e.g., jejunal resection, ileal resection) with occurrence of IFALD. However, the absence of the ileocecal valve (ICV) has been well established as a risk factor for both SBBO and prolonged PN requirement in comparison to intact ICV and colon after surgery or stoma takedown [27]. One can extrapolate that this anatomic variation may also predispose to IFALD based at least on the risk of prolonged PN exposure.

Excessive Energy and/or Carbohydrate Load

Excessive delivery of parenteral calories (>110–120 kcal/kg/day) or carbohydrate (>15–20 mg/kg/min) may lead to hyperglycemia, hyperinsulinemia, and hepatic steatosis. Steatosis is an important feature of the liver injury in adults with IFALD and less so in infants and children. In the present era, when attention to dose reduction of intravenous lipids has gained predominance in IF management, a compensatory increase in parenteral carbohydrate delivery has been used to maintain adequate provision of energy. Future studies should be entertained to re-examine the relationship between carbohydrate administration and liver injury.

Amino Acid Component

As a key macronutrient and energy source in parenteral nutrition infusates, parenteral amino acid dosing and sources have been analyzed for their potential role in IFALD pathogenesis. There is reasonable evidence to suggest that a higher cumulative administered dose of parenteral amino acids is associated with higher risk of IFALD in infants. Steinbach et al. [28] found that the development of cholestasis in 122 patients on PN was more likely in neonates with a higher cumulative amino acid dose and longer duration of exposure to PN, factors which are essentially dependent on one another. Although the cumulative duration and dose of parenteral amino acid exposure may be a relevant risk factor, there is no evidence that the initial dose of amino acid or rate of advancement play a significant role in IFALD development [19].

There is speculation that the formulation of parenteral amino acid solution may affect the development of IFALD. In comparing two standard amino acid formulations (Aminosyn PF and TrophAmine), Forchielli et al. [29] found no difference in development of cholestasis, whereas several years later, Wright et al. [30] suggested that Aminosyn PF was more likely to lead to cholestasis in a retrospective analysis of neonates on PN≥21 days. Thus, there is inconclusive evidence regarding the formulation of amino acid solution in IFALD pathogenesis.

A deficiency in conditional amino acids in the neonate may also predispose to development of cholestasis. Premature neonates have decreased expression of cystathionase, which synthesizes taurine (and cysteine) from methionine [1]. Taurine is essential to the conjugation of bile acids, and therefore taurine deficiency might lead to reduced bile acid secretion and subsequent accumulation of toxic bile acids. In an animal model of parenteral nutrition associated cholestasis (PNAC), supplementation with taurine improved bile acid secretion [31].

Carnitine depletion has been demonstrated in both animal models and human studies after PN administration [32]. Carnitine plays a key role in mitochondrial fatty acid (FA) uptake and oxidation. Therefore, carnitine deficiency may impair energy utilization while increasing oxidative stress within the liver. Despite these findings, there has been no proven therapeutic benefit of carnitine supplementation on IFALD development [33].

Choline deficiency has been associated with the development of steatosis in patients on parenteral nutrition [34]. In children, choline deficiency has been associated with elevation of serum AST and ALT [35]. However, choline deficiency has not been shown to be associated with the development of cholestatic injury in pediatric patients.

Lipid Emulsions

Among the studies attempting to identify a single causative parenteral factor in the pathogenesis of IFALD, the most promising work has implicated intravenous lipid emulsions. The lipid component of PN in the USA has conventionally been provided by a soybean-based intravenous lipid emulsion. Consensus guidelines have recommended that lipid emulsions should account for 25–40% of non-protein calories administered in PN, translating to a daily dose of 2.0–3.5 g/kg/day [36]. Commonly used intravenous lipid emulsions are composed primarily of soybean or other plant-based oils. Until recent years, the US Food and Drug Administration (FDA) had only approved two lipid solutions for parenteral use, soybean-based Intralipid (Fresenius Kabi, Bad Homburg, Germany) and soybean and safflower-based Liposyn II (Hospira Inc, Lake Forest IL). These plant-derived lipid oils contain primarily omega-6 (n-6) polyunsaturated fatty acids (PUFA). In general, long chain PUFAs (LCPUFAs) can be categorized as omega-6 (n-6) and omega-3 (n-3) PUFAs. Omega-6 FAs, including linoleic acid (LA), are precursors of arachidonic acid (AA) which in turn forms the structural backbone of proinflammatory eicosanoids. In contrast, n-3 FAs (e.g., α-linolenic acid) are metabolized into anti-inflammatory derivatives of eicosapentanenoic acid (EPA) and docosahexaenoic acid (DHA). For a number of years, alternative lipid emulsions have been approved for use in Europe. These contain a higher ratio of n-3 to n-6 FA, and include emulsions based primarily on fish oil (Omegaven; Fresenius Kabi, Bad Homburg, Germany) and a mixed lipid emulsion containing soy oil, medium chain triglycerides, olive oil and fish oil (SMOFlipid; Fresenius Kabi, Bad Homburg, Germany). Table 15.2 contains details of the components of currently available parenteral lipid emulsions [37]. In 2018, Omegaven was approved in the USA for treatment and provision of FAs to infants with PNAC, and SMOFLipid for use in patients 18 years and older.

In 2000, Colomb et al. [38] reported a series of ten infants on PN that developed cholestasis in temporal correlation with increased dosage of soy-based lipid emulsion. More recently, Diamond et al. [39] reported that an important risk factor for development of advanced IFALD (serum conjugated bilirubin ≥5.9 mg/dL) was days of exposure to parenteral lipid dose of ≥2.5 g/kg/day. In the last decade, a paradigm shift in the clinical management of pediatric IF has entailed the reduction or elimination of soy-based lipid emulsions and replacement with alternative lipid emulsions, including products derived from omega-3 FA (fish oil emulsion) [40–42]. The pathophysiology of liver injury following soy lipid emulsion administration includes considerations of the biologic effects of FAs, as well as the potential role of plant-derived sterols (phytosterols).

In reviewing the role of intravenous lipids as the primary mediator of liver injury, supporting data have been obtained from animal models and human studies in which n-6 based FAs are reduced or replaced with n-3 based FAs. Studies reporting the effects of these therapeutic maneuvers in pediatric patients will be reviewed in the “Treatment” section of this chapter. In short, Gura et al. [43] reported two pediatric patients with IF in whom cholestasis biochemically normalized after switching from soy-based lipid emulsion to fish oil-based lipid emulsion. This initial report has led to a re-examination of the role of lipid emulsions both clinically and in animal models. A primary hypothesis supporting the potential benefit of n-3 vs. n-6 based lipid solutions is the downstream anti-inflammatory properties of the n-3 FAs as compared to the potentially proinflammatory n-6 FAs. Several studies have clarified that this change in FA profile may alter immunomodulatory pathways both within the liver and systemically [44, 45]. As an example of direct anti-inflammatory effects within the liver, Schmocker et al. [46] reported decreased markers of hepatocyte injury and inflammation (ALT and liver biopsy scores) and decreased proinflammatory cytokine expression (TNF-α, IFN-γ, IL-1β, and IL-6) in a mouse model of steatohepatitis following supplementation with n-3 PUFA compared to controls.

In the clinical setting, the provision of proinflammatory n-6 FA has potential implications in the child with intestinal failure, where there are recurring infectious and proinflammatory events including intestinal inflammation and peritonitis, catheter-associated bloodstream infections, and small intestinal bacterial overgrowth and translocation. These infants and children are at risk for sepsis, increased intestinal permeability to macromolecules derived from bacteria, and they may face subclinical events of bacteremia and endotoxinemia. Therefore, the administration of anti-inflammatory n-3 FAs may potentially help to blunt systemic inflammatory responses that mitigate end-organ pathways, including liver injury in IFALD.

Aside from alterations in proinflammatory/anti-inflammatory balance, n-6 FA administration may also favor the development of hepatic steatosis [47]. In contrast, in murine models of hepatic steatosis, reduced histologic fat accumulation was demonstrated after administration of n-3 FA [48]. This effect of n-3 FA may be mediated through induction of FA β-oxidation or reduced hepatic lipogenesis, as shown in experimental models [47].

Another important component of intravenous lipid emulsions is the phytosterols, plant-based naturally occurring plant sterols that resemble cholesterol but have an alkylated side chain. In 1993, Clayton et al. [49] reported elevated plasma phytosterol levels in five children with cholestatic liver disease associated with PN administration. In all patients, phytosterol levels and serum bilirubin improved following reduction in the amount of lipid emulsion administered. The role of phytosterols in the pathogenesis of cholestasis may be mediated through their effects on bile acid synthesis, transport and secretion into bile. For example, injection of phytosterols to neonatal piglets led to decreased bile flow and increased serum bilirubin levels [50]. Similarly, rat hepatocytes in culture exposed to phytosterols demonstrated decreased morphological canalicular expression and decreased bile acid secretion [50]. In another experimental model, neonatal piglets who received a fish oil-based IV lipid emulsion in PN had improved cholestasis and bile flow as compared to soybean oil-derived lipid [51]. Although the latter study did not measure phytosterols specifically, these clinical and laboratory data suggest that the development of cholestasis in IFALD may be, in part, mediated by phytosterols in soybean-based lipid solutions. More recent studies demonstrate a correlation between phytosterol levels and the degree of both biochemical and histological injury in IFALD [52].

Molecular mechanisms that underlie the role of phytosterols in cholestasis development may involve alterations in expression of canalicular phospholipid and bile acid transporters. Tazuke et al. [53] have evaluated the effects of TPN administration on Abcb4 mRNA expression in mice, which encodes for mdr2, a canalicular phospholipid flippase. Mice administered TPN had significantly decreased expression of mdr2 correlating with elevated serum bile acid levels. In further experiments, there was no significant difference in mdr2 expression in mice that received PN with and without lipid emulsion, although serum bile acid levels were normal in mice that did not receive lipid [54]. Canalicular bile acid transport may occur through several proteins, including BSEP encoded by ABCB11, which is regulated by the nuclear receptor FXR. Administration of phytosterol (stigmasterol) led to decreased FXR expression in murine hepatocytes in primary culture, and further ameliorated downstream BSEP expression, suggesting a direct role of phytosterols in the pathogenesis of cholestasis through altered canalicular BSEP expression and subsequent hepatocyte retention of toxic bile acids [55]. In recent investigations, El Kasmi and Sokol have further demonstrated that PN with soybean based-lipid administration led to liver injury and cholestasis (elevated AST, ALT, bile acid, and bilirubin) and decreased FXR and BSEP expression in a murine model combining intestinal injury with PN administration, and that these effects were blunted when the animal received either saline, PN without lipid, or PN with fish oil-based lipid. Furthermore, addition of stigmasterol to fish oil-based lipid led to similar measures of liver injury and FXR/BSEP expression as in the soybean oil group [56]. These data support the pathogenic role of phytosterols in IFALD development, specifically cholestasis, through inhibition of FXR-mediated BSEP pathways or other canalicular biliary transporters. Phytosterols were also shown to directly activate macrophages, eliciting a proinflammatory cytokine profile. Inherited polymorphisms of genes regulating bile acid synthesis, metabolism and transport may further explain clinical phenotypes that are more susceptible to lipid-mediated IFALD.

A second important synergistic factor involved in the pathogenesis of IFALD appears to be the role of intestinal injury, inflammation and altered barrier function. In a mouse model of IFALD, El Kasmi and Sokol demonstrated that LPS absorbed by injured intestine activates hepatic macrophages and Kupffer cells to produce IL1-β, which binds to its hepatocyte receptor and down-regulates expression of FXR and LXR pathways through NFκB [57]. This results in down-regulation of Abcg5/g8, the canalicular transporter of phytosterols and subsequent hepatocyte accumulation of infused phytosterols in the PN lipid emulsions, which then interfere with FXR signaling and BSEP expression causing cholestasis. Furthermore, it has been shown in infants with IFALD compared to those with SBS without IFALD, that the fecal microbiome is shifted to overabundance of proteobacteria, which favor expression of LPS setting the stage for enhanced intestinal absorption of LPS and macrophage activation.

The proposed pathogenesis of IFALD in relation to the interplay between soy lipid emulsion administration, infection, Kupffer cell activation, and ultimately, cholestasis is summarized in Figure 15.4 [58].

Only gold members can continue reading. Log In or Register to continue

Share this:

• Click to share on Twitter (Opens in new window)
• Click to share on Facebook (Opens in new window)

Related posts:

Chapter 5 – Cirrhosis and Chronic Liver Failure in Children Chapter 13 – Familial Hepatocellular Cholestasis Chapter 16 – Diseases of the Gallbladder in Infancy, Childhood, and Adolescence Chapter 26 – Cystic Fibrosis Liver Disease in Children Chapter 32 – Lysosomal Storage Disorders in Children Chapter 43 – Liver Transplantation in Children: Indications and Surgical Aspects

Stay updated, free articles. Join our Telegram channel

Join