Chapter 422. Drug, Toxin, and TPN Liver Disease

Drug-Induced Liver Injury

Adverse drug reactions in children are uncommon. Nevertheless, drug-induced hepatotoxicity, when it occurs, must be promptly recognized, and the offending agent discontinued, although cessation does not always result in rapid recovery. Delays in recognizing hepatic injury may significantly contribute to morbidity, resulting in a need for liver transplantation or in death.1-5

The role of the liver in the processing or biotransformation of xenobiotics (foreign substances) is discussed in Chapter 418. Mechanisms of hepatoxicity vary, as they depend on the drug, dosage, and patient factors such as age, gender, nutrition, and genetic predisposition. In general, medicinal and environmental agents known to cause hepatotoxicity have been characterized as predictable (intrinsic) or unpredictable (idiosyncratic) hepatotoxins. The patterns of liver injuries are clinically and histopathologically diverse (Table 422-1).

Specific hepatotoxins that are commonly prescribed for the pediatric population include analgesics (acetaminophen), anticonvulsants, and antibiotics. Acetaminophen is a predictable or intrinsic hepatotoxicant and acetaminophen overdose has been recognized as one of the most common causes of liver failure in the United Kingdom and the United States. Specific therapy for acetaminophen overdose is available. N-Acetylcysteine, when provided in the first hours or days after overdose, replenishes glutathione stores and enables the liver to metabolize acetaminophen without generating toxic metabolites (see Chapter 120).

Examples of idiosyncratic hepatotoxic reactions are those associated with phenytoin, an anticonvulsant, and sulfasalazine, used in the treatment of inflammatory bowel disease. These drugs may cause an illness that resembles a hypersensitivity reaction, with lymphadenopathy, fever, sore throat, and peripheral as well as tissue eosinophilia. Other idiosyncratic reactions may depend on drug metabolism and are not necessarily associated with this clinical picture.

Commonly prescribed antibiotics such as erythromycin have been associated with hepatic injury, although antibiotic-associated hepatotoxicity appears to be more common in adults. Nevertheless, fulminant hepatic failure has been reported in a child receiving trimethoprim-sulfamethoxazole. Minocycline, frequently used in the treatment of adolescent acne, has been associated with the development of autoimmune hepatitis. The loss of intrahepatic bile ducts or “vanishing bile duct syndrome” has been associated with synthetic penicillins and the anticonvulsant carbamazepine. Pemoline, a stimulant used in the treatment of attention deficit disorder, has been reported to cause hepatic necrosis. Oral contraceptives have been associated with hepatic vein thrombosis (Budd-Chiari syndrome) and liver tumors. Recreational drugs of abuse are increasingly reported as causes of hepatic injury in adolescents.

It is mandatory that a history of prescribed, over-the-counter, or illicit drug usage be sought in any child who presents with evidence of hepatic dysfunction. The treatment of drug-induced liver injury depends largely on timely recognition, which may be evident by increased serum aminotransferase values, conjugated hyperbilirubinemia, coagulopathy, jaundice, or more systemic side effects such as fever, lymphadenopathy, and rash. Once a drug-induced liver injury is identified, therapy is mainly supportive, but withdrawal of the offending agent is critical to minimize hepatotoxicity. With the exception of acetaminophen overdose, no specific therapies exist.1-5

Toxin-Induced Liver Injury

There are several types of environmental hepatotoxins. These include herbal preparations used as nutritional or health aids, contaminated food (fungicide-treated wheat), and household items such as pesticides or cleaning products. Occasionally, liver toxicity results when these toxins are ingested either as a form of drug abuse (inhalants) or as a suicide attempt.

As discussed for pharmaceutical agents, the mechanisms of hepatotoxicity of environmental toxins vary. Ingestion of poisonous mushrooms such as Amanita phalloides, results in fatty liver (steatosis) and hepatocyte necrosis that is often severe, rapidly progressive, and fatal.6 Venoocclusive liver disease has been reported with certain herbal teas, such as comfrey, taken as a nutritional supplement.7

Treatment of liver toxicity secondary to environmental agents is comparable to that for drug-induced liver injury. Avoidance or withdrawal of the offending agent is crucial, and this usually depends on an accurate historical account. Therapy is usually supportive, but there are a few exceptions. Toxicity from iron overdose is treated with chelation therapy.

Although not usually classified as a drug or toxin, administration of total parenteral nutrition (TPN) is associated with liver disease. TPN-associated cholestasis is defined as a conjugated bilirubin greater than 2 mg/dL in an individual receiving TPN who has no evidence of another underlying primary liver disease (eg, hepatitis, metabolic liver disease, biliary atresia, etc) to explain the cholestasis. The initial biochemical evidence of TPN-associated cholestasis is an elevated serum bile acid level, although this is rarely obtained in routine clinical practice. More typical laboratory evidence of TPN-induced liver disease includes elevated transaminases, GGT (gamma glutamyl transferase, a biliary tract enzyme), and conjugated hyperbilirubinemia. Hepatosplenomegaly is a common clinical exam finding of TPN-associated cholestasis.

Both the biochemical and exam findings associated with TPN-induced liver dysfunction may occur after as little as 2 weeks of TPN administration. Liver failure associated with TPN presents clinically with jaundice, massive/firm hepatosplenomegaly, ascites, and bleeding. Laboratory evidence of portal hypertension such as thrombocytopenia, anemia, and evidence of liver dysfunction with hypoalbuminemia, coagulopathy, and hyperammonemia may ensue.

Epidemiology

Risk factors for the development of TPN-associated liver disease have long been known to be greater in infants, particularly premature infants, than in older children or adults. Factors increasing the likelihood of developing parenteral nutrition-associated cholestasis include a birth weight of less than 500 g (Odds ratio, OR, of 30.7), a birth weight 500 to 749 g (OR, 13.1), gastroschisis (OR 20.3), and jejunal atresia (OR 24.0).8-12 TPN related deaths are much more common in those infants with TPN-associated cholestasis with higher direct bilirubin levels (for instance Å  ~ 4 mg/dL) or who have had at least one septic event.10,12,13 Cholestasis in infants correlates very closely with birth weight, degree of prematurity, and length of time receiving TPN.14 TPN-associated cholestasis has been reported to occur in as many as 90% of children who receive intravenous nutrition for more than 3 months.12 Liver failure occurs early on in infants on TPN and can be rapidly progressive, with a reported incidence of 17%.13

Pathophysiology

The precise etiology of TPN-induced liver disease remains debated, but most agree that this is a multifactorial condition.8-11 Infants and children with TPN-induced liver disease present histologically with cholestasis progressing toward fibrosis, and even biliary cirrhosis, over time. The pathology differs from the typical pathological picture seen with TPN-associated liver disease in adults who are more likely to develop steatosis or a fatty liver from TPN.11,15 (Fig. 422-1A). The fibrosis and progression to cirrhosis tends to be more prevalent and rapid in the smallest of the premature infants as compared to older children and adults (Figure 422-1B). This has led some to postulate that a toxic component of TPN or lipids (eg, phytosterols) may be accumulating in TPN recipients and altering critical bile acid transporter genes (eg, BSEP, or the bile salt export pump) in the livers of neonates, and that, perhaps, this effect is more profound in infants because of maturational differences in the expression of such transporters.11,16-18 Various components of TPN and lipids have been implicated in the pathogenenesis of TPN-induced liver disease including dextrose, manganese, lipids, phytosterols, and others.10,11

Treatment

Prevention of sepsis is particularly important in children at risk for TPN-induced liver disease. Bacteremia in chronic TPN recipients may occur from skin entry of bacteria into the bloodstream or from translocation of bacteria across the intestinal wall into the bloodstream. Any infection, whether viral or bacterial at any site in the body can exacerbate cholestasis associated with TPN, and so sterile practices and careful handling of central venous catheters is of utmost importance for all patients receiving long-term TPN. In addition, the lack of enterally-stimulated hormones such as cholecystokinin, glucagon, motilin, and others may decrease gallbladder or intestinal contractility and motility, thereby reducing bile flow, and predisposing to bacterial stasis and overgrowth of intraluminal bacteria that can translocate across the intestinal wall, resulting in sepsis-associated cholestasis.18-22 The goal in most infants with short bowel syndrome and intestinal failure is to advance the enteral feeds so that an intravenous catheter (with its inherent infectious risks) is no longer needed as described in Chapter 408. Enteral feeding, even if merely “trophic” in nature, may help stimulate bile flow and lessen cholestasis.

The dihydroxy bile acid, ursodeoxycholic acid (Actigall, Ursodiol), is a choleretic that has increasingly being administered to infants and children with TPN-associated cholestasis. Reports in both animal models of cholestasis and in small cohorts of infants and adults with TPN-associated cholestasis suggest that introduction of ursodeoxycholic acid is associated with improvement in liver biochemical parameters.22-24 Alternative medications for TPN-associated cholestasis (eg, CCK-cholecystokinin and tauroursodeoxycholic acid) have been tried, but not proven successful.25,26 Since clinical experience has shown that TPN-induced cholestasis persists despite ursodeoxycholic acid therapy, it appears that this agent alone is not sufficient to reverse the cholestasis associated with TPN.

Previously the fat component of TPN was provided by a soybean oil emulsion that is high in omega-6 fatty acids. Recent trials with an alternative lipid emulsions high in omega 3 fatty acids (fish-oil derived) have shown very promising results with a markedly reduced incidence of TPN cholestasis, and improvement in the clinical course of children with established TPN cholestasis.

Larger, well-designed, randomized, controlled clinical trials are needed to better evaluate safety and efficacy.27