 |
The mechanisms by which drugs damage the liver are variable, complex, and often poorly understood. Some drugs are directly toxic: with these, injury is generally characteristic for the drug, begins within hours of exposure, and is dose-related. Other drugs produce damage only rarely and only in susceptible people; the injury generally first occurs within a few weeks but occasionally may be delayed for several months after drug exposure. This injury is not dose-related. These reactions are rarely allergic; they are more accurately described as idiosyncratic. The distinction between direct toxicity and idiosyncrasy may not always be clear; eg, some drugs whose injury appears idiosyncratic probably damage cell membranes directly with toxic intermediate metabolites.
Although there is no perfect system for classifying liver damage caused by drugs, damage can be categorized as acute reactions (which consist of hepatocellular necrosis), cholestasis (with or without inflammation), and miscellaneous reactions (see Table 1: Drugs and the Liver: Common Hepatotoxic Drug Reactions ). Some drugs can cause chronic damage, which rarely leads to tumor growth.
|
Table 1
|
 |  |  |
|
Common Hepatotoxic Drug
Reactions
|
|
Drug
|
Reaction
|
|
Acetaminophen
|
Acute, direct hepatocellular toxicity; chronic toxicity
|
|
Allopurinol
|
Miscellaneous acute reactions
|
|
Amanita mushrooms
|
Acute, direct hepatocellular toxicity
|
|
Aminosalicylic acid
|
Miscellaneous acute reactions
|
|
Amiodarone
|
Chronic toxicity
|
|
Antibiotics, various
|
Miscellaneous acute reactions
|
|
Antineoplastics, various
|
Miscellaneous acute reactions
|
|
Arsenic compounds
|
Chronic toxicity
|
|
Aspirin
|
Miscellaneous acute reactions
|
|
C‑17 alkylated steroids
|
Acute cholestasis, steroid type
|
|
Chlorpropamide
|
Acute cholestasis, phenothiazine type
|
|
Diclofenac
|
Acute, idiosyncratic hepatocellular toxicity
|
|
Erythromycin estolate
|
Acute cholestasis, phenothiazine type
|
|
Halothane-related anesthetics
|
Acute, idiosyncratic hepatocellular toxicity
|
|
Hepatic intra-arterial antineoplastics
|
Chronic toxicity
|
|
HMG‑CoA reductase inhibitors (statins)
|
Miscellaneous acute reactions
|
|
Hydrocarbons
|
Acute, direct hepatocellular toxicity
|
|
Indomethacin
|
Acute, idiosyncratic hepatocellular toxicity
|
|
Iron
|
Acute, direct hepatocellular toxicity
|
|
Isoniazid
|
Acute, idiosyncratic hepatocellular toxicity; chronic toxicity
|
|
Methotrexate
|
Chronic toxicity
|
|
Methyldopa
|
Acute, idiosyncratic hepatocellular toxicity; chronic toxicity
|
|
Methyltestosterone
|
Acute cholestasis, steroid type
|
|
Monoamine oxidase inhibitors
|
Acute, idiosyncratic hepatocellular toxicity
|
|
Niacin
|
Chronic toxicity
|
|
Nitrofurantoin
|
Chronic toxicity
|
|
Oral contraceptives
|
Acute cholestasis, steroid type
|
|
Phenothiazines (eg, chlorpromazine )
|
Acute cholestasis, phenothiazine type; chronic toxicity
|
|
Phenylbutazone
|
Acute cholestasis, phenothiazine type
|
|
Phenytoin
|
Acute, idiosyncratic hepatocellular toxicity
|
|
Phosphorus
|
Acute, direct hepatocellular toxicity
|
|
Propylthiouracil
|
Acute, idiosyncratic hepatocellular toxicity
|
|
Quinidine
|
Miscellaneous acute reactions
|
|
Sulfonamides
|
Miscellaneous acute reactions
|
|
Tetracycline , high-dose IV
|
Acute, direct hepatocellular toxicity
|
|
Tricyclic antidepressants
|
Acute cholestasis, phenothiazine type
|
|
Valproate
|
Miscellaneous acute reactions
|
|
Vitamin A
|
Chronic toxicity
|
|
Diagnosis
and Treatment
Drug-induced hepatotoxicity is suspected when patients have unusual patterns of liver disease (eg, mixed or atypical patterns of cholestasis and hepatitis); in hepatitis or cholestasis for which common causes have been excluded; during administration of a drug with known hepatotoxicity (see Table 1: Drugs and the Liver: Common Hepatotoxic Drug Reactions ), even in the absence of symptoms or signs; or if a liver biopsy reveals histologic features suggesting a drug etiology. Jaundice due to drug-induced hemolysis may at first suggest hepatotoxicity, but in such cases, bilirubin is unconjugated and other liver function test results are normal.
No diagnostic tests can confirm that a drug caused hepatotoxicity. Diagnosis requires exclusion of other possible causes (eg, imaging tests to exclude obstruction if cholestasis is present; viral serology if hepatitis is present) and a temporal relationship between the drug and hepatotoxicity. A pattern of repeated, reversible hepatotoxicity after repeated doses is the most conclusive evidence, but because of the risk of serious liver damage, rechallenging a patient with a suspected hepatotoxic drug generally is not done. Biopsy is sometimes necessary, generally to exclude other treatable conditions. If the diagnosis is still unclear after testing, a trial of drug withdrawal may be indicated for diagnosis as well as treatment.
For a few drugs that cause direct hepatotoxicity (eg, acetaminophen ), blood levels can be used to assess the probability of liver damage. However, drug levels may fall if tests are delayed. Many nonprescription herbal products cause liver toxicity; patients with unexplained liver injury should be asked whether they are taking such products.
Treatment for drug-induced hepatotoxicity generally consists of withdrawing the drug and providing supportive therapy.
Hepatocellular
Necrosis
Hepatocellular necrosis is conceptually divided into direct toxicity and idiosyncrasy, although this distinction may be artificial. The hallmark is elevated aminotransferase levels, often to a striking degree. Patients with mild or moderate hepatocellular necrosis may develop manifestations of hepatitis (eg, jaundice, malaise). Patients with severe necrosis may develop manifestations of fulminant hepatitis (eg, hepatic insufficiency, portal-systemic encephalopathy).
Direct
toxicity:
Most direct hepatotoxins produce dose-related hepatic necrosis and often affect other organs (eg, kidneys).
Direct hepatotoxic damage from prescribed drugs can generally be prevented or minimized by following recommendations regarding maximum drug dosing and patient monitoring. Poisoning with direct hepatotoxins (eg, acetaminophen , iron, Amanita mushrooms) often produces gastroenteritis within hours. However, manifestations of liver damage may develop after only 1 to 4 days. Cocaine use occasionally causes acute hepatocellular necrosis, perhaps by inducing hepatocellular ischemia.
Idiosyncrasy:
Drugs can produce acute hepatocellular necrosis that is indistinguishable, even histologically, from viral hepatitis. The mechanisms are uncertain and probably vary with individual drugs. Isoniazid and halothane have been most thoroughly studied.
The mechanism of the rare halothane-related hepatitis is unclear but may include formation of reactive intermediates, cellular hypoxia, lipid peroxidation, and autoimmune-mediated damage. Risk factors include obesity (possibly because halothane metabolites are stored in adipose tissue) and repeated exposures to the anesthetic at relatively short intervals. Hepatitis typically develops within a few days to 2 wk after exposure, is heralded by fever, and is often severe. Occasionally, eosinophilia or a skin rash develops. Mortality is 20 to 40% if severe jaundice is present, but survivors usually recover completely. Methoxyflurane and enflurane, which are related anesthetics, can produce the same syndrome.
Cholestasis
Many drugs can produce a primarily cholestatic reaction. Usually the pathogenesis is poorly understood, but at least two forms of cholestatic injury—phenothiazine- and steroid-type—are clinically and histologically distinct. Diagnostic testing often includes noninvasive imaging to exclude biliary obstruction. Further testing (eg, magnetic resonance cholangiopancreatography, ERCP, liver biopsy) is necessary only if cholestasis persists after the drug is stopped.
Phenothiazine-type
cholestasis is a periportal inflammatory reaction. Immunologic mechanisms are suggested by some evidence, such as occasional eosinophilia or other signs of hypersensitivity reactions, but direct toxicity to hepatic canaliculi is also possible. This type of cholestasis occurs in about 1% of patients given chlorpromazine and less often in those given other phenothiazines. Cholestasis is often acute and is accompanied by fever and high levels of aminotransferases and alkaline phosphatase. Differentiation from extrahepatic obstruction may be difficult, even by liver biopsy. If the drug is stopped, complete resolution is typical, although progression to chronic cholestasis with fibrosis occurs rarely. Cholestasis produced by tricyclic antidepressants, chlorpropamide , phenylbutazone , erythromycin estolate, and many other drugs is clinically similar; however, progression to chronic liver damage from these drugs has not been clearly established.
Steroid-type
cholestasis appears to be an exaggeration of the physiologic effect of sex hormones on bile formation rather than an immunologic sensitivity or membrane cytotoxicity. Impaired canalicular water flow, microfilament dysfunction, altered membrane fluidity, and genetic factors may be responsible. Little or no hepatocellular inflammation exists. Although the incidence varies worldwide, it occurs in 1 to 2% of women taking oral contraceptives. Gradual onset of cholestasis without systemic symptoms is characteristic. Alkaline phosphatase is elevated, but aminotransferase levels are usually not very high, and liver biopsy shows only centrizonal bile stasis with little portal or hepatocellular damage. Complete resolution follows drug withdrawal in most cases but may be prolonged.
Cholestasis of pregnancy (see Pregnancy Complicated by Disease: Cholestasis (pruritus) of pregnancy) is closely related to steroid-related cholestasis. Women with cholestasis of pregnancy may develop cholestasis with subsequent oral contraceptive use and vice versa.
Miscellaneous
Acute Reactions
Some drugs cause mixed forms of hepatic dysfunction, granulomatous reactions (eg, quinidine , allopurinol , sulfonamides), or variants of liver injury that are difficult to classify. HMG-CoA reductase inhibitors (statins) produce subclinical aminotransferase elevations in 1 to 2% of patients, although clinically important liver injury is infrequent. Many antineoplastic drugs also cause liver damage; the mechanisms vary.
Chronic Liver
Disease
Certain drugs can cause chronic liver disease. Isoniazid , methyldopa , and nitrofurantoin can produce chronic hepatitis. Resolution usually occurs if fibrosis is not present. The illness may begin acutely or insidiously. Progression to cirrhosis may occur. Chronic hepatitis-like histology with scarring occurs rarely in patients using acetaminophen long-term in doses as low as 3 g/day, although higher doses are usually required. Alcoholics appear to be more susceptible, and the disorder is suspected in alcoholics found incidentally to have unusually high aminotransferase levels, especially AST (values rarely exceed 300 IU in alcoholic hepatitis alone). Amiodarone occasionally produces chronic liver injury with Mallory bodies and histologic features otherwise similar to alcoholic liver disease; membrane phospholipidosis is a factor in pathogenesis.
A sclerosing cholangitis–like syndrome can develop from hepatic intra-arterial chemotherapy, especially with floxuridine . Patients receiving methotrexate long-term (usually for psoriasis or RA) can develop insidiously progressive hepatic fibrosis, particularly in the case of alcoholics or if the drug is given daily; liver function tests are often unremarkable, and liver biopsy is needed. Although fibrosis caused by methotrexate is rarely clinically important, most authorities recommend biopsy when the cumulative drug dose reaches 1.5 to 2 g and occasionally thereafter. Noncirrhotic hepatic fibrosis that can produce portal hypertension can result from use of arsenical compounds or excessive amounts of vitamin A (eg, > 15,000 U/day for months) or niacin (see Vitamin Deficiency, Dependency, and Toxicity: Niacin Toxicity). In many tropical and subtropical countries, chronic liver disease and hepatocellular carcinoma are believed to result from ingesting foods containing fungal aflatoxins.
Besides causing cholestasis, oral contraceptives may also occasionally cause benign hepatic adenomas and, very rarely, hepatocellular carcinoma. Adenomas are usually subclinical but may present with sudden intraperitoneal rupture and hemorrhage, requiring emergency laparotomy. Most adenomas do not cause symptoms and are found incidentally during imaging tests. Because oral contraceptives increase clotting generally, they increase the risk of hepatic vein thrombosis (Budd-Chiari syndrome). Use of these drugs also increases the risk of gallstones because they enhance bile lithogenicity.
Last full review/revision November 2005
Content last modified November 2005
|  |