You are on page 1of 10

1118 THE NEW ENGLAND JOURNAL OF MEDICINE Oct.

26, 1995

REVIEW ARTICLES

zymes that make up the P450 system.7-12 Found primar-


ily in the liver but also in the gastrointestinal tract,
MEDICAL PROGRESS kidneys, brain, and other tissues, P450 enzymes are
composed of a unique apoprotein and a heme prosthet-
ic group, which binds oxygen after electron-transfer
DRUG-INDUCED HEPATOTOXICITY reactions from NADPH, resulting in aliphatic and aro-
matic hydroxylation; O -, N-, or S -dealkylation; or de-
WILLIAM M. LEE, M.D. halogenation. A typical reaction of this type generates
a hydroxyl group, which can then participate in the
phase 2 reactions. Each group of genes with 40 per-

D RUG-induced liver injury is a potential complica-


tion of nearly every medication that is prescribed,
because the liver is central to the metabolic disposition
cent amino acid homology composes a family whose
gene products (isozymes) may function in a similar
fashion. For example, CYP3 is a family that contains an
of virtually all drugs and foreign substances.1-3 Al- A subfamily and several genes, numbered 1, 2, and so
though drugs are usually metabolized without injury to forth. The primary enzyme for the metabolism of
the liver, many fatal and near-fatal drug reactions oc- erythromycin in humans is P450 3A4.
cur each year. A few compounds produce metabolites
that cause liver injury in a uniform, dose-dependent Phase 2 Reactions
fashion.4-6 Most agents form a toxic byproduct only in After a phase 1 reaction, most compounds are still
rare persons. Injury to hepatocytes results either di- not very water-soluble and require further metabolism.
rectly from the disruption of intracellular function or In a typical phase 2 reaction, a large, water-soluble po-
membrane integrity or indirectly from immune-mediat- lar group is attached to a hydroxyl oxygen by glucuron-
ed membrane damage. Factors promoting the accumu- idation or sulfation, forming ether or ester linkages.
lation of hepatocyte toxins include genetic alterations in These are the sole steps required for the hepatic metab-
enzymes that allow the formation of the harmful me- olism of some compounds, but for most, the phase 2 re-
tabolite, competition by other drugs, and depletion of action is preceded or followed by phase 1 oxidation.
the substrates required to detoxify the metabolite. Compounds requiring glucuronidation include aceta-
In this article I provide a theoretical background for minophen, morphine, and furosemide, as well as biliru-
understanding drug-induced liver injury; outline the bin. Sulfation is as important as glucuronidation, par-
most common types of injury; address hepatotoxicity ticularly for the metabolism of steroid compounds and
due to combination agents, new formulations, and alter- bile acids. There are several species of sulfotransferases
native treatment regimens (vitamins and herbal reme- with overlapping specificities, each employing 3-phos-
dies); and discuss diagnosis, treatment, and prevention. phoadenosine-5-phosphosulfate synthesized from ATP
and sulfate ions. Although phase 2 reactions are usually
BACKGROUND accomplished without a detrimental effect, they can oc-
Most drugs and toxins enter the body through the casionally lead to toxic or carcinogenic byproducts.
gastrointestinal tract, with a minority absorbed directly
through the lungs or skin or by a parenteral route. Each Glutathione Metabolism
foreign compound is eliminated unchanged or metabo- A third metabolic pathway for detoxifying many com-
lized by enzymes, undergoes a spontaneous chemical pounds involves glutathione, a thiol-containing tripep-
transformation, or is simply not eliminated. Most com- tide capable of binding to potentially harmful elec-
pounds are lipophilic, entering the body through the trophilic compounds through glutathione S -transferase.
gastrointestinal tract and hepatocyte membrane barri- Glutathione substrate is depleted in the process of detox-
ers. Biotransformation is the process by which thera- ification and must be replenished by sulfhydryl com-
peutic agents are rendered more hydrophilic so that they pounds from the diet or by cysteine-containing drugs
can be filtered by the glomerulus or excreted in bile. such as N-acetylcysteine. The glutathione S-transferase
Biotransformation from a nonpolar to a polar compound reaction is therefore central to the detoxification of a
takes place in several steps, grouped as phase 1 and number of compounds, including acetaminophen. Other
phase 2 reactions. enzymes, such as alcohol dehydrogenase, are important
for the elimination of a few compounds, but the princi-
Phase 1 Reactions pal metabolic pathways for most agents are those dis-
In the phase 1 reaction, oxidation or demethylation cussed above.
occurs, mediated by cytochrome P450, a gene super-
family (CYP) that has nearly 300 members. A variety PATHOGENESIS OF TOXIC REACTIONS
of oxidative phase 1 reactions are performed by the en- Since the hepatocyte is the main metabolic engine of
the liver, most adverse drug reactions result first in hep-
From the Liver Unit, Department of Internal Medicine, University of Texas atocyte necrosis. The most common reaction leading to
Southwestern Medical Center, 5323 Harry Hines Blvd., Dallas, TX 75235-8887,
where reprint requests should be addressed to Dr. Lee. cell necrosis is the formation of covalent bonds between
Supported by the Houghton Foundation and the estate of Alison B. Harwood. a reactive metabolite of the parent compound and cell
Downloaded from www.nejm.org on September 19, 2003. This article is being provided free of charge for use in Indonesia:NEJM Sponsored.
Copyright © 1995 Massachusetts Medical Society. All rights reserved.
Vol. 333 No. 17 MEDICAL PROGRESS 1119

proteins or DNA. Oxidation may go


awry if a reactive electrophilic com- Acetaminophen
pound accumulates or if oxygen in-
O O O
termediates (such as the superoxide
anion or free radicals) are formed, HN C CH3 HN C CH3 HN C CH3
which then react with cellular com-
ponents. Perhaps the best example is
acetaminophen.
Although used universally for
UDP – glucuronosyl- Sulfotransferase
nonnarcotic pain relief, acetamino- transferase
phen has a predictable toxic effect if
taken in quantities exceeding those Glucuronate OH SO4
recommended in the package insert,
causing dose-related centrilobular Cytochrome P450 2E1
necrosis in the liver. The metabolic
pathway for acetaminophen involves
O O O
phase 1 and 2 reactions, glutathione
detoxification, and the formation of HN C CH3 N C CH3 HN C CH3
reactive intermediates, which dis-
rupt cell macromolecules (Fig. 1).
As a general rule, the capacity for GSH Cell proteins
glucuronidation is much greater Glutathione-
than that typically required each S-transferase
day; even patients with advanced liv- SG S -Protein
er disease continue to have adequate OH O OH
glucuronidation. If glucuronyl trans-
ferase and sulfotransferase are avail- Mercapturic Acid NAPQI Covalent Binding
4
able, phase 2 reactions predomi-
nate, with only a small fraction of Figure 1. Metabolic Pathway of Acetaminophen.
acetaminophen metabolized directly Acetaminophen primarily undergoes sulfation and glucuronidation (phase 2 reac-
by cytochrome P450, unless the tions) but is metabolized by cytochrome P450 2E1 (in a phase 1 reaction) to
quantity of acetaminophen exceeds N-acetyl-p-benzoquinoneimine (NAPQI) if the capacity of the phase 2 reactions is
the capacity of these phase 2 en- exceeded or if cytochrome P450 2E1 synthesis is induced. Glutathione-S -trans-
ferase is capable of detoxifying NAPQI to yield mercapturic acid and its derivatives,
zymes. At this point, an avidly elec- if glutathione is available. In the absence of glutathione substrate, covalent binding
trophilic compound, N-acetyl-p -ben- to cell proteins occurs. N-acetylcysteine is an excellent source of glutathione sub-
zoquinoneimine (NAPQI), is formed strate. UDP denotes uridine diphosphate, and GSH reduced glutathione.
through cytochrome P450 and may
bind covalently to cell macromolecules, thereby disrupt- rare toxic events include genetically variant P450
ing mitochondrial and possibly nuclear function. The isozymes, which contribute either to lack of metabolism
formation of covalent bonds is prevented if NAPQI of a given precursor or excess formation of a toxic me-
can be detoxified by conjugation (through glutathione- tabolite. One example is debrisoquin, an antihyperten-
S-transferase) to generate, through a series of steps, sive compound marketed in Europe and studied exten-
mercapturic acid, a harmless, water-soluble product ex- sively, since its urinary metabolites can be readily
creted by the kidney. Depletion of glutathione reduces analyzed. Debrisoquin is hydroxylated by P450 2D6, as
this last defense against the formation of NAPQI-related are perhexiline maleate, propranolol, quinidine, and
intracellular adducts. Thus, any situation that leads to desipramine, among other drugs. Almost 10 percent of
the depletion of glutathione will increase toxicity, where- normal people lack detectable levels of P450 2D6. In
as an increase in available glutathione stores will dimin- such persons, any drug metabolized primarily by this
ish this effect. Starvation and alcohol deplete mitochon- enzyme will have a greatly prolonged half-life. This de-
drial glutathione, whereas N-acetylcysteine replenishes fect, which is inherited as an autosomal recessive trait,
glutathione stores and protects against acetaminophen- involves abnormal production of messenger RNA, so
induced injury.13-15 In a similar fashion, the P450 iso- that the appropriate apoprotein is not made.7,8 The
zyme (P450 2E1), which is responsible for the conver- studies of P450 2D6 suggest that genetic enzyme vari-
sion of acetaminophen to NAPQI, is induced by ethanol ants are one explanation for the occasional and isolated
and inhibited by cimetidine.11-13 Thus, at several meta- toxic reactions to substances that virtually everyone
bolic stages, ethanol increases toxicity, whereas cimeti- can metabolize.
dine may serve as an antidote.16 Advanced age and renal
insufficiency may have important adjunctive roles.17 Hepatocyte Necrosis
The actual cause of cell death remains unclear. One
Enzyme Polymorphism result of the covalent binding of substrate or lipid per-
Most drugs cause toxic reactions only rarely and oxidation within cells is an increase in levels of cytosol-
without a dose-related pattern. Explanations for these ic calcium. Calcium is important for the regulation of a
Downloaded from www.nejm.org on September 19, 2003. This article is being provided free of charge for use in Indonesia:NEJM Sponsored.
Copyright © 1995 Massachusetts Medical Society. All rights reserved.
1120 THE NEW ENGLAND JOURNAL OF MEDICINE Oct. 26, 1995

number of cell functions, including maintenance of the to all circumstances. The most practical way of catego-
cytoskeleton and membrane integrity. Actin depolymer- rizing drug reactions is according to the type of reac-
ization and polymerization are dependent on calcium- tion observed, which takes into account the histologic
ion fluxes within the cytosol. The results of studies changes and cell type involved, as well as the clinical
using NAPQI in isolated hepatocytes suggest that alter- picture (Table 2).
ations in calcium homeostasis occur with the influx of
calcium ions into the cytosol. Whether this is the cause Direct Toxic Reactions (Acetaminophen)
or the result of disordered membrane transport is un- Acetaminophen is an example of an agent that caus-
clear, but altered permeability may lead to blebs in the es a direct toxic reaction. Two clinical scenarios ac-
cell membrane and loss of membrane integrity. Other count for most cases of acetaminophen-related hepatic
mechanisms may also be at play; in each instance, the necrosis: the intentional suicidal overdose and the
covalent binding of reactive intermediates to cell pro- “therapeutic misadventure.” In the latter scenario, an
teins seems to be the initiating step. alcoholic takes acetaminophen for pain relief in doses
In addition to producing direct toxic effects, the for- that exceed those recommended in the package insert
mation of drug–protein adducts may lead to allergic re- (4 g per 24 hours). The result is a direct toxic reaction
actions, such as those observed with halothane. None- due to the enzyme-induction and glutathione-depletion
theless, the formation of antibodies to P450 enzyme mechanisms outlined above.26 Starvation may also play
species after hepatotoxic reactions does not necessar- a part, presumably because of glutathione depletion.27
ily indicate that these antibodies have a pathogenetic This alcohol–acetaminophen syndrome, which is often
role.18 unrecognized, may be the most common form of acute
liver failure in the United States and Australia.28,29 Ex-
Role of Physiologic Factors tremely elevated serum alanine and aspartate amino-
The metabolic fate of any compound is a complex transferase values (mean, approximately 9000 U per li-
process. Table 1 outlines the variables (other than the ter in one study) distinguish this condition from viral or
toxic potential of the compound itself) that may play a alcoholic hepatitis, but very high levels are also ob-
part in the metabolic outcome.17,19-25 Multiple factors served in patients who take an intentional overdose of
are often involved, the most frequent being enzyme in- acetaminophen. Even with the measurement of aceta-
duction.17 Common inducing agents include ethanol, minophen levels in the blood, it may be difficult to pre-
phenobarbital, and phenytoin, but cigarette smoke is dict the outcome in such patients.30 If there is uncer-
also a potent inducer of certain P450 enzyme species. tainty about the dose or time of ingestion or if the dose
appears to have been excessive regardless of the blood
TYPES OF DRUG REACTIONS acetaminophen level, N-acetylcysteine should be given
Although most hepatotoxic effects involve hepato- through a nasogastric tube immediately and for the en-
cyte necrosis, some drugs injure bile ducts or canaliculi, suing 48 hours, to provide glutathione substrate. The
causing cholestasis without marked damage of hepato- expected survival rate is higher than 80 percent, but
cytes. Other therapeutic agents affect sinusoidal or en- liver transplantation is occasionally necessary.
dothelial cells (resulting in veno-occlusive disease or
fibrosis) or fat-storing Ito cells (causing vitamin A tox- Idiosyncratic Reactions (Isoniazid)
icity, which leads to fibrosis) or cause a particular pat- Unlike acetaminophen, the majority of drug-related
tern of liver injury affecting multiple cell types. Drug reactions, such as those observed with isoniazid, are id-
reactions can be classified as hepatocellular, cholestatic, iosyncratic and unpredictable.31 Fifteen to 20 percent of
or mixed, but these are general terms and do not apply patients receiving isoniazid as a single agent for pro-
phylaxis against tuberculosis may have increased se-
Table 1. Variables Affecting Drug rum alanine and aspartate aminotransferase levels, but
Metabolism. only 1 percent have hepatic necrosis severe enough to
Age19,20
require the withdrawal of the drug. Several factors ex-
Sex20
plain the relatively common (albeit sporadic) toxic re-
Diet actions observed. First, the simultaneous use of alcohol
Micronutrients (calcium, iron, magnesium, or rifampin may augment the toxicity of isoniazid. Sec-
copper, and zinc) ond, elderly persons may be more likely to have toxic
Caffeine
Vegetable-enzyme inducers reactions than younger persons.32 Third, genetic differ-
Lipids21 ences are important, since persons who are capable of
Ethanol11,12
rapid acetylation of isoniazid have an increased likeli-
Pregnancy
Diabetes22
hood of toxic reactions resulting from the formation
Hepatic disease of acetylhydrazine, which is then transformed by cyto-
Renal disease23 chrome P450 into a reactive metabolite. Some studies
Immune stimuli suggest that persons with slow acetylation are at great-
Interferon er risk for a toxic reaction through a separate pathway
Interleukin-624
that leads to the formation of hydrazine, which itself
Enzyme polymorphism25
Drug–drug interference
may be toxic.32 In the case of isoniazid and perhaps of
Enzyme induction
other drugs causing idiosyncratic reactions, such reac-
tions are not truly idiosyncratic but occur when a series
Downloaded from www.nejm.org on September 19, 2003. This article is being provided free of charge for use in Indonesia:NEJM Sponsored.
Copyright © 1995 Massachusetts Medical Society. All rights reserved.
Vol. 333 No. 17 MEDICAL PROGRESS 1121

Table 2. Types of Toxic Reactions Occurring granulomas in liver-biopsy specimens. This allergic re-
in the Liver. action is accompanied by both hepatocyte necrosis and
TYPE OF REACTION EXAMPLES OF AGENTS
cholestasis.37,38 The mechanisms responsible for the com-
bined allergic and hepatotoxic reactions are unknown,
Direct reaction Acetaminophen, carbon tetra- but the slow resolution of the illness suggests that the
chloride, mushrooms, phos-
phorus allergen remains on the hepatocyte surface for weeks or
Idiosyncratic reaction Isoniazid, disulfiram, propyl- months.
thiouracil* This drug-induced hypersensitivity hepatitis syn-
Toxic–allergic reaction Halothane, isoflurane, ticrynafen
Allergic hepatitis Phenytoin, amoxicillin–clavulan- drome results in a mononucleosis-like illness that may
ic acid, sulfonamides be confused with a viral illness or streptococcal pharyn-
Cholestatic reaction Chlorpromazine, erythromycin
estolate, estradiol, captopril,
gitis, so that the agent is not withdrawn, despite signs
sulfonamides of developing hepatitis. The result is often a severe
Granulomatous reaction Diltiazem, quinidine, phenytoin, form of the Stevens–Johnson syndrome, with fever last-
procainamide
Chronic hepatitis Nitrofurantoin, methyldopa, iso- ing for weeks. The substitution of phenobarbital for
niazid, trazodone phenytoin occasionally results in cross-reactivity and a
Alcoholic hepatitis–like Amiodarone, perhexiline male- similar hypersensitivity reaction. As with any therapeu-
reaction ate, valproic acid
Microvesicular steatosis Tetracyclines, aspirin, zidovu- tic agent, rapid recognition of a possible toxic reaction
dine, didanosine, fialuridine and discontinuation of the drug are the keys to limiting
Fibrosis or cirrhosis alone Methotrexate, vitamin A, meth-
yldopa
hepatic damage. The features of the allergic reaction
Veno-occlusive disease Cyclophosphamide, other are sometimes not obvious, although eosinophilia or
chemotherapeutic agents, granulomas may be present in liver-biopsy specimens.
herbal teas
Ischemic damage Cocaine, sustained-release nico-
tinic acid, methylenedioxyam- Cholestatic Reactions (Estradiol)
phetamine The drugs that mainly affect bile flow, causing chole-
*There are hundreds of other agents that can cause idiosyncratic reac- static injury, include estradiol, chlorpromazine, tri-
tions. methoprim–sulfamethoxazole, rifampin, erythromycin
estolate, nafcillin,39 and captopril. Typically, jaundice
of genetic and environmental influences coincide to pro- appears early, with associated pruritus but little alter-
duce a sufficient quantity of one or more toxic metab- ation in the patient’s general well-being. A liver biopsy
olites. In most patients, there is no evidence of an al- reveals engorgement of the canaliculi with bile and
lergic reaction, and the histologic picture is virtually minimal hepatocellular injury (Fig. 2C). Eosinophils
indistinguishable from that of viral hepatitis (Fig. 2B). may be found in mildly inflamed portal tracts. The
Diclofenac is another example of a commonly used mechanism of cholestatic injury remains unclear. Estra-
agent that, like other nonsteroidal agents, occasionally diol and other estrogens have been shown to decrease
causes severe hepatotoxic reactions.33 bile flow and Na/K –ATPase, change tight junctions
between cells, and alter the fluidity of the hepatocyte
Combined Toxic and Allergic Reactions (Halothane) membrane. Given the large number of women (and
A seldom-used anesthetic agent that was very popu- men) taking estrogens, this form of cholestasis is re-
lar for a number of years, halothane can induce a com- markably rare.
bination of toxic and allergic reactions leading to liver
injury.34,35 Severe halothane-related hepatitis generally Granulomatous Reactions
develops after multiple exposures to the drug such as Noncaseating granulomas resembling sarcoidosis in
those that can occur on subspecialty surgical services. the liver are caused by a variety of drugs. The clinical
Although there is usually no rash, fever and eosino- picture is the same as that of other forms of granulo-
philia are commonly observed, and the histologic fea- matous hepatitis: low-grade fever and chronic fatigue,
tures of liver-biopsy specimens are similar to those with jaundice only in rare cases. The list of possible
seen with idiosyncratic reactions. The initial elevations agents is long (Table 3).
in serum alanine and aspartate aminotransferase levels
are delayed, but the interval between the administra- Drug-Induced Chronic Hepatitis (Methyldopa)
tion of halothane and toxic reactions becomes shorter Methyldopa and a number of other compounds have
with each exposure. Protein adducts formed from the been found to cause a more indolent form of liver dam-
initial toxic reaction provide the hapten for the forma- age that closely resembles autoimmune chronic active
tion of antibodies, so that with subsequent exposure, hepatitis (Fig. 2). Hyperglobulinemia may be present,
antibody and cellular recognition of the halothane–pro- with positive tests for antinuclear antibodies. The classic
tein-adduct antigen on the hepatocyte surface leads to agent producing this reaction is oxyphenisatin, a laxa-
cell lysis.35 A similar process occurs with other halo- tive that has been withdrawn from the market.40 Early
genated, volatile anesthetic agents.36 identification of drug-related chronic hepatitis is not
easy; cirrhosis may develop before the hepatitis is diag-
Allergic Hepatitis (Phenytoin) nosed. Identifying the drug or toxin that has caused the
Drugs such as phenytoin can cause a systemic aller- cirrhosis is difficult retrospectively if the patient has
gic reaction characterized by fever, rash, lymphadenop- been consuming alcohol or if unrecognized viral hepati-
athy, eosinophilia, and the presence of eosinophils or tis is present. Nevertheless, in addition to methyldopa,41
Downloaded from www.nejm.org on September 19, 2003. This article is being provided free of charge for use in Indonesia:NEJM Sponsored.
Copyright © 1995 Massachusetts Medical Society. All rights reserved.
1122 THE NEW ENGLAND JOURNAL OF MEDICINE Oct. 26, 1995

A B

C D
Figure 2. Photomicrographs of Normal Hepatic Tissue and Tissue with Various Types of Injury Due to Drugs.
Panel A shows a normal hepatic lobule, with a portal tract (P) and a hepatic venule (V) (hematoxylin and eosin, 145). Hepatic-
cell plates are typically one cell thick. In this specimen, there is slight sinusoidal dilatation but no inflammation, necrosis, or regen-
eration. Panel B shows a liver-biopsy specimen obtained at autopsy from a 24-year-old man in whom subacute fatal liver failure
developed after he had taken sulfasalazine for several months for ulcerative colitis (hematoxylin and eosin, 145). Several years
earlier, sulfasalazine had been discontinued after an episode of hepatitis thought to be caused by the drug. The doctor who pre-
scribed the second course of the drug was not aware of the previous episode. Moderately severe necrosis of hepatocytes is present
throughout the lobule, with mononuclear cells, eosinophils, and a few polymorphonuclear leukocytes. Signs of cellular regeneration
are indicated by the presence of binucleate and trinucleate cells (arrow). Panel C shows an example of isolated cholestatic liver
injury due to treatment with ethinyl estradiol (hematoxylin and eosin, 185). The specimen is from a 59-year-old woman in whom
pruritus and jaundice developed two months after the initiation of estrogen-replacement therapy. She recovered completely one
month after discontinuing the medication. Prominent casts of inspissated bile are visible as green, translucent deposits in the cana-
liculi, and there is bile staining of hepatocytes but no evidence of necrosis or inflammation. Panel D shows a specimen from a
patient with chronic hepatitis due to treatment with methyldopa, which the patient had taken for more than one year (hematoxylin
and eosin, 185). The classic histologic features of autoimmune hepatitis are present: intense infiltration of the portal triads with
mononuclear cells, including plasma cells (lower right), and piecemeal necrosis (the encroachment of these cells into the hepatic
lobule, with the ballooning of adjacent hepatocytes, upper left). Panel E shows a specimen from a patient with nonalcoholic steato-
hepatitis who had been treated with amiodarone (hematoxylin and eosin, 185). Marked Mallory’s bodies are present within hep-
atocytes (small arrow), as well as macrovesicular fat (large arrow) and polymorphonuclear leukocytes (P). Panel F shows a speci-
men from a patient with indolent hepatic injury due to treatment with methotrexate (Masson’s trichrome, 185). The patient had
taken methotrexate for seven years, receiving a total dose of more than 5 g. The drug was discontinued after this biopsy specimen
had been obtained. Little or no inflammation is present, but there is evidence of increased fibrosis and expansion of the portal triad
by fibrous tissue (upper left). Nuclear atypia, vacuolization, and mild fatty metamorphosis, which are also present, are typical find-
ings in patients treated with methotrexate. Panel G shows the histologic features of moderately severe vitamin A toxicity (Masson’s
trichrome, 185). Ingestion of excess vitamin A results in the engorgement of the sinusoidal fat-storing (or Ito) cells, which appear
here principally as vacuolated spaces because of the removal of the lipid (vitamin A) in the fixation process. The marked surrounding
fibrosis, indicated by blue-staining collagen, is derived from the Ito cells. Panel H shows the features of severe veno-occlusive
disease in a liver-biopsy specimen obtained at autopsy 31 days after allogeneic bone marrow transplantation following conditioning
with busulfan and cyclophosphamide (Masson’s trichrome, 185). The lumen of the small hepatic venule shown is completely oc-
cluded with loose extracellular material and entrapped red cells. No endothelium is visible. There is extensive necrosis of hepato-
cytes and sinusoidal congestion surrounding the venule, with only a few remaining hepatocytes (lower right). Panels E and G are
provided courtesy of Dr. K.G. Ishak, Armed Forces Institute of Pathology, Washington, D.C., and Dr. M.A. Rothschild, Manhattan
Veterans Affairs Medical Center, New York. Panel H is provided courtesy of Drs. H.M. Shulman and G.B. McDonald, Fred Hutchinson
Cancer Research Center, Seattle.

Downloaded from www.nejm.org on September 19, 2003. This article is being provided free of charge for use in Indonesia:NEJM Sponsored.
Copyright © 1995 Massachusetts Medical Society. All rights reserved.
Vol. 333 No. 17 MEDICAL PROGRESS 1123

E F

G H

acetaminophen,42 nitrofurantoin,43 trazodone, and phen- ciated with considerable cellular dysfunction but without
ytoin have been determined to cause this syndrome cell death. This is the characteristic lesion of fatty liver
(Fig. 2D). Because these drugs are used for the long- caused by pregnancy, high doses of tetracyclines, and
term treatment of relatively benign conditions, moni- Reye’s syndrome associated with aspirin.29 Macrovesic-
toring for untoward effects may be inadequate. Multi- ular and microvesicular steatosis occur in association
ple prescription renewals may be a problem in the case with the acquired immunodeficiency syndrome (AIDS)
of nitrofurantoin, which is used to control recurrent uri- and with the use of zidovudine.48-50 Such lesions have
nary tract infections. been reported in eight patients receiving zidovudine48
and in one patient treated with didanosine.51 These re-
Fatty Liver and Alcoholic Hepatitis–Like Reactions ports are of particular interest in the light of the recent
(Amiodarone)
tragic outcome with fialuridine, a new nucleoside ana-
Although fatty liver is most commonly related to logue for the treatment of hepatitis B.52-54 Like the pa-
obesity, diabetes, alcoholism, or corticosteroid therapy, tient treated with didanosine, several patients receiving
amiodarone and several other drugs can cause a disor- fialuridine had severe or fatal lactic acidosis in asso-
der similar to alcoholic hepatitis, termed nonalcoholic ciation with microvesicular steatosis after eight weeks
steatohepatitis. Amiodarone, which has a unique histo- of therapy. These changes were assumed to be due to
logic and clinical profile, is a potent antiarrhythmic a gradual uncoupling of mitochondrial oxidative me-
agent used to treat life-threatening ventricular tachy- tabolism.
cardia. This drug (and some related compounds) has
been shown to cause severe liver toxicity, in an acute or Indolent Cirrhosis (Methotrexate)
chronic form,44-47 as part of a multisystem syndrome. Of the several agents capable of causing a gradual
Patients typically have moderately elevated serum ala- progression to cirrhosis without any manifestation of
nine and aspartate aminotransferase levels, with the clinical illness, methotrexate is the most frequently cit-
characteristic lesion of steatohepatitis (Fig. 2E), and ed example. This agent is used in patients with severe
cirrhosis can develop in just a few months. psoriasis or rheumatoid arthritis, and toxicity may de-
The presence of microvesicular fat within hepato- velop over a period of several years without any symp-
cytes has a different meaning from that of the macrove- toms or evidence of hepatitis or other biochemical ab-
sicular steatosis discussed above. Fine vesicles are asso- normalities.55,56 A liver biopsy is the only sure way to
Downloaded from www.nejm.org on September 19, 2003. This article is being provided free of charge for use in Indonesia:NEJM Sponsored.
Copyright © 1995 Massachusetts Medical Society. All rights reserved.
1124 THE NEW ENGLAND JOURNAL OF MEDICINE Oct. 26, 1995

Table 3. Drugs Associated with Granuloma- well.61 A similar systemic reaction with rhabdomyolysis
tous Liver Disease.* has been observed with sustained-release nicotinic acid62
Allopurinol and methylenedioxyamphetamine (“ecstasy”).63
Aspirin
Carbamazepine
Nonprescription and Reformulated Agents
Cephalexin Over-the-counter preparations are assumed to be
Diazepam safer than prescription drugs, but this is not always the
Diltiazem case. Laxatives that contained oxyphenisatin are exam-
Halothane
ples of nonprescription drugs associated with liver
Hydralazine
Isoniazid
damage.40 Products sold in health food stores may be
Methyldopa
especially hazardous, since they are assumed to be at
Metolazone least nontoxic, even if not effective. Because of this im-
Nitrofurantoin plicit assumption, patients are more likely to induce
Oxyphenbutazone toxic reactions by exceeding the recommended doses.
Penicillin The list of alternative medicines that can cause toxic
Phenytoin reactions includes vitamin A,64 germander,65,66 chapar-
Procainamide ral leaf,67,68 comfrey,69 and jin bu huan, a Chinese herbal
Procarbazine product.70
Quinidine
Safety profiles may change when drugs are reformu-
Sulfonamides
lated. For example, nicotinic acid, a relatively safe
Sulfonylureas
Trichlormethiazide
agent, had greatly increased hepatotoxic effects when
issued in a sustained-release form. Such effects had
*Adapted from Zimmerman and Maddrey.1
been recognized rarely with higher doses, but the flush-
ing reaction that occurs in most patients limited exces-
establish the diagnosis of indolent cirrhosis caused by a sive doses. The use of a sustained-release formulation
drug reaction. Although guidelines vary, a pretreatment led to tolerance of higher doses, which in turn led to hy-
biopsy is not indicated unless the patient has abnormal potension, ischemic liver injury, and fulminant hepatic
liver-function values or there is a suspicion of alcohol- failure.62
ism. Many clinicians routinely perform a biopsy after
administering a total dose of 2500 mg of methotrexate Multidrug Regimens
(Fig. 2F). Methyldopa41 and vitamin A57 (Fig. 2G) have It should not be surprising that one drug can inter-
been reported to cause a similar syndrome. fere with the biotransformation of another drug. What
is surprising is that such interference does not occur
Veno-occlusive Disease more often. There are several circumstances in which
Intensive chemotherapy, usually including the agent drug combinations are associated with an increased
cyclophosphamide, is most closely associated with the risk of toxic reactions. First, drugs may be combined in
development of a rapidly progressive, occlusive disease a single formulation, such as trimethoprim–sulfameth-
of small hepatic venules due to endothelial-cell injury oxazole,71 amoxicillin–clavulanic acid,72,73 and isonia-
(Fig. 2H). The abrupt onset of painful hepatomegaly, zid–rifampin.74 With each of these combination agents,
ascites, jaundice, and other symptoms of hepatic insuf- there have been numerous reports of hepatotoxic reac-
ficiency heralds this disease, which is the most common tions that were more severe than those associated with
complication of bone marrow transplantation.58 A sim- one agent used alone. The mechanism of injury is be-
ilar syndrome is observed in persons who drink Jamai- lieved to involve the induction of cytochrome P450 by
can “bush tea.” one agent, which increases the quantity of the toxic me-
tabolite formed from the other.
OTHER FACTORS IN DRUG-INDUCED LIVER DISEASE
Isoniazid and rifampin may be used concurrently as
Cocaine Abuse single agents rather than as a combined formulation.
Ischemic liver damage is a well-known complication Either agent alone can be the cause of a hepatotoxic re-
of severe heart failure but may also be caused by hy- action, although rifampin generally impairs bilirubin
potensive reactions to drugs. Although cocaine abuse is uptake, resulting only in elevated bilirubin levels.74 In
a widespread problem, little has been written about he- addition, acetaminophen-related toxicity may be poten-
patic injury due to cocaine. After the ingestion of co- tiated by isoniazid. The more elaborate four- and five-
caine, shock and disseminated intravascular coagula- drug regimens used for resistant tuberculosis have sim-
tion may develop, with evidence of myonecrosis. The ilar hazards.75 For reasons that are unclear, patients
associated toxic effect on the liver is likely to be ische- with AIDS have an apparently increased susceptibility
mic in nature, the result of systemic hypotension in- to drug-related liver injury, particularly in association
duced by coronary (and systemic arterial) vasospasm with trimethoprim–sulfamethoxazole, pentamidine,76
with congestive heart failure.59 The task of sorting out and oxacillin.77
the more subtle forms of liver injury in cocaine abusers
is complicated by the concurrent abuse of other drugs, DIAGNOSIS, TREATMENT, AND PREVENTION
including alcohol,60 and by the presence of viral hepati- The diagnosis of drug-induced liver injury is often
tis, but cocaine appears to be directly hepatotoxic as obscured by difficulty in determining the precise timing
Downloaded from www.nejm.org on September 19, 2003. This article is being provided free of charge for use in Indonesia:NEJM Sponsored.
Copyright © 1995 Massachusetts Medical Society. All rights reserved.
Vol. 333 No. 17 MEDICAL PROGRESS 1125

of the drug ingestion from the patient’s history. Essen- Table 4. A Partial List of Newer Agents As-
tial to the diagnosis is evidence that the patient was not sociated with Hepatotoxic Reactions.
ill before ingesting the drug, became ill while taking it, Chlorzoxazone82
and, in most cases, had striking improvement after its Clozapine83
withdrawal. Since drug-related hepatitis can be fatal, it Diclofenac33,84
is vital to be aware of the possible severity of the hepat- Doxepin85
ic reaction and to discontinue any potentially toxic Etoposide86
medication immediately. The best way to identify the Etretinate87
drug causing the reaction is to make a careful time line Floxacillin88
Flutamide89
of all drugs ingested and to be particularly suspicious
Glyburide90
of any treatment with a potentially hepatotoxic drug
Ketoconazole91
begun during the three months before the onset of an Labetalol92
illness. Many drugs, including digoxin and theophyl- Lisinopril93
line, are virtually never implicated as causes of liver in- Lovastatin94
jury, whereas certain classes of drugs, such as non- Norfloxacin95
steroidal agents33,78 and some antibiotic agents,79 are Ofloxacin96
commonly implicated. Pentamidine97
The main treatment for drug-induced hepatotoxicity Piroxicam98
is the withdrawal of the agent, with careful observation Terbutaline99
Ticlopidine100
of the patient to make sure the expected improvement
Trazodone101
begins to occur within a few days. Certain agents, such
as amoxicillin–clavulanic acid and phenytoin, have
been associated with a syndrome in which the condi-
tion actually worsened for several weeks after the drug during the first year after their introduction, particular-
was withdrawn and took months to resolve complete- ly if they offer no advantages over accepted formula-
ly.73 If clinical or laboratory signs of hepatic failure ap- tions. In addition, physicians should caution their pa-
pear, hospitalization is mandatory.29 The prognosis for tients to be alert for signs of drug-induced liver injury,
patients with acute liver failure caused by idiosyncratic especially in the case of agents with well-recognized
drug reactions is poor, with a mortality rate higher hepatotoxic effects. The challenge for physicians and
than 80 percent in most series.80 Corticosteroid treat- pharmaceutical companies alike is to alert patients to
ment may be used in patients with evident hypersensi- the potential toxic effects of drugs without frightening
tivity, but controlled trials have not proved the efficacy them to such an extent that they avoid taking needed
of such treatment. Intentional overdoses must be treat- medication. For known hepatotoxins, such as isoniazid
ed like any poisoning, with appropriate emergency and diclofenac, monthly monitoring of serum alanine
measures. A suspected overdose with acetaminophen and aspartate aminotransferase levels is suggested dur-
is treated with N-acetylcysteine even when the drug ing the first six months of treatment. Since many drug
has been ingested 36 hours or more earlier.15 Hemodi- reactions develop quickly, monitoring is not a complete
alysis or hemofiltration is rarely indicated. Putative safeguard against toxicity. Many fatal drug reactions
toxic agents are generally not given again, since re- might have been prevented, however, had the agent
challenge may be associated with a more severe subse- been withdrawn at the first sign of illness. The educa-
quent reaction (Fig. 2B). A cautious rechallenge should tion of patients is therefore essential to the prevention
be considered only if the toxic reaction observed was of drug-induced hepatotoxicity. Patients who do not
highly questionable and if no other drug is available realize that drug-induced injury is possible and those
for the treatment of a potentially life-threatening dis- who are encouraged to continue taking a drug despite
order. initial signs of toxicity are at the highest risk for fatal
Each year, dozens of new pharmacologic agents ap- reactions.
pear on the market. The pressure from the public, as I am indebted to Drs. Burton Combes and Ronald W. Estabrook for
well as the pharmaceutical industry, to bring new their critical insights and to Dr. Steve Foster for assistance with the
agents to the marketplace is great, and cautionary tales photomicrographs.
of failed drugs, such as ticrynafen, are often forgotten.
REFERENCES
During the first nine months after its introduction, this
diuretic and uricosuric agent was involved in more than 1. Zimmerman HJ, Maddrey WC. Toxic and drug-induced hepatitis. In:
Schiff L, Schiff ER, eds. Diseases of the liver. Philadelphia: J.B. Lippin-
25 fatal hepatotoxic reactions.81 Although ticrynafen cott, 1993.
was removed from the market as soon as the problem 2. Kaplowitz N. Drug metabolism and hepatotoxicity. In: Kaplowitz N, ed.
was recognized, its toxicity was only fully realized after Liver and biliary diseases. Baltimore: Williams and Wilkins, 1991.
3. Farrell GC. Drug-induced liver disease. Edinburgh, Scotland: Churchill
its public release. Each new agent approved by the Livingstone, 1994.
Food and Drug Administration has undergone rigorous 4. Mitchell JR, Jollow DJ, Potter WZ, Gillette JR, Brodie BB. Acetamino-
phen-induced hepatic necrosis. IV. Protective role of glutathione. J Phar-
clinical trials, but there is no substitute for the wider macol Exp Ther 1973;187:211-7.
use that follows product licensing. Some of the newer 5. Baerg RD, Kimberg DV. Centrilobular hepatic necrosis and acute renal
agents associated with acute liver necrosis are listed in failure in “solvent sniffers.” Ann Intern Med 1970;73:713-20.
6. Klein AS, Hart J, Brems JJ, Goldstein L, Lewin K, Busuttil RW. Amanita
Table 4.33,82-101 poisoning: treatment and the role of liver transplantation. Am J Med 1989;
Physicians may wish to defer prescribing new drugs 86:187-93.
Downloaded from www.nejm.org on September 19, 2003. This article is being provided free of charge for use in Indonesia:NEJM Sponsored.
Copyright © 1995 Massachusetts Medical Society. All rights reserved.
1126 THE NEW ENGLAND JOURNAL OF MEDICINE Oct. 26, 1995

7. Gonzalez FJ. Human cytochromes P450: problems and prospects. Trends 40. Reynolds TB, Peters RL, Yamada S. Chronic active and lupoid hepatitis
Pharmacol Sci 1992;13:346-52. caused by a laxative, oxyphenisatin. N Engl J Med 1971;285:813-20.
8. Watkins PB. Drug metabolism by cytochromes P450 in the liver and small 41. Lee WM, Denton WT. Chronic hepatitis and indolent cirrhosis due to
bowel. Gastroenterol Clin North Am 1992;21:511-26. methyldopa: the bottom of the iceberg? J S C Med Assoc 1989;85:75-9.
9. Murray M. P450 enzymes: inhibition mechanisms, genetic regulation and 42. Barker JD Jr, de Carle DJ, Anuras S. Chronic excessive acetaminophen use
effects of liver disease. Clin Pharmacokinet 1992;23:132-46. and liver damage. Ann Intern Med 1977;87:299-301.
10. Guengerich FP. Human cytochrome P-450 enzymes. Life Sci 1992;50: 43. Reinhart HH, Reinhart E, Korlipara P, Peleman R. Combined nitrofuran-
1471-8. toin toxicity to liver and lung. Gastroenterology 1992;102:1396-9.
11. Takahashi T, Lasker JM, Rosman AS, Lieber CS. Induction of cytochrome 44. Rigas B, Rosenfeld LE, Barwick KW, et al. Amiodarone hepatotoxicity: a
P-4502E1 in the human liver by ethanol is caused by a corresponding in- clinicopathologic study of five patients. Ann Intern Med 1986;104:348-
crease in encoding messenger RNA. Hepatology 1993;17:236-45. 51.
12. French SW, Wong K, Jui L, Albano E, Hagbjork AL, Ingelman-Sundberg 45. Gehenot M, Horsmans Y, Rahier J, Geubel AP. Subfulminant hepatitis re-
M. Effect of ethanol on cytochrome P450 2E1 (CYP2E1), lipid peroxida- quiring liver transplantation after benzarone administration. J Hepatol
tion, and serum protein adduct formation in relation to liver pathology and 1994;20:842.
pathogenesis. Exp Mol Pathol 1993;58:61-75. 46. Morelli S, Guido V, De Marzio P, Aguglia F, Balsano F. Early hepatitis
13. Roe AL, Snawder JE, Benson RW, Roberts DW, Casciano DA. HepG2 during intravenous amiodarone administration. Cardiology 1991;78:291-4.
cells: an in vitro model for P450-dependent metabolism of acetaminophen. 47. Lewis JH, Ranard RC, Caruso A, et al. Amiodarone hepatotoxicity: prev-
Biochem Biophys Res Commun 1993;190:15-9. alence and clinicopathologic correlations among 104 patients. Hepatology
14. Prescott LF, Illingworth RN, Critchley JAH, Stewart MJ, Adam RD, 1989;9:679-85.
Proudfoot AT. Intravenous N-acetylcystine: the treatment of choice for 48. Gradon JD, Chapnick EK, Sepkowitz DV. Zidovudine-induced hepatitis.
paracetamol poisoning. BMJ 1979;2:1097-100. J Intern Med 1992;231:317-8.
15. Smilkstein MJ, Knapp GL, Kulig KW, Rumack BH. Efficacy of oral 49. Freiman JP, Helfert KE, Hamrell MR, Stein DS. Hepatomegaly with se-
N-acetylcysteine in the treatment of acetaminophen overdose: analysis of vere steatosis in HIV-seropositive patients. AIDS 1993;7:379-85.
the National Multicenter Study (1976 to 1985). N Engl J Med 1988;319: 50. Chattha G, Arieff AI, Cummings C, Tierney LM Jr. Lactic acidosis com-
1557-62. plicating the acquired immunodeficiency syndrome. Ann Intern Med
16. Wolff T, Strecker M. Endogenous and exogenous factors modifying the ac- 1993;118:37-9.
tivity of human liver cytochrome P-450 enzymes. Exp Toxicol Pathol 51. Lai KK, Gang DL, Zawacki JK, Cooley TP. Fulminant hepatic failure as-
1992;44:263-71. sociated with 2,3-dideoxyinosine (ddI). Ann Intern Med 1991;115:283-
17. Bonkovsky HL, Kane RE, Jones DP, Galinsky RE, Banner B. Acute hepat- 4.
ic and renal toxicity from low doses of acetaminophen in the absence of 52. Touchette N. HBV-drug deaths prompt restudy of similar antivirals. J NIH
alcohol abuse or malnutrition: evidence for increased susceptibility to drug Res 1993;5:33-5.
toxicity due to cardiopulmonary and renal insufficiency. Hepatology 1994; 53. Macilwain C. NIH, FDA seek lessons from hepatitis B drug trial deaths.
19:1141-8. Nature 1993;364:275.
18. Beaune PH, Bourdi M. Autoantibodies against cytochromes P-450 in drug- 54. McKenzie R, Fried MW, Sallie R, et al. Hepatic failure and lactic acidosis
induced autoimmune hepatitis. Ann N Y Acad Sci 1993;685:641-5. due to fialuridine (FIAU), an investigational nucleoside analogue for
19. Hunt CM, Westerkam WR, Stave GM, Wilson JA. Hepatic cytochrome chronic hepatitis B. N Engl J Med 1995;333:1099-105.
P-4503A (CYP3A) activity in the elderly. Mech Ageing Dev 1992;64:189-99. 55. Newman M, Auerbach R, Feiner H, et al. The role of liver biopsies in pso-
20. Hunt CM, Westerkam WR, Stave GM. Effect of age and gender on the ac- riatic patients receiving long-term methotrexate treatment: improvement
tivity of human hepatic CYP3A. Biochem Pharmacol 1992;44:275-83. in liver abnormalities after cessation of treatment. Arch Dermatol 1989;
21. Yoo JS, Smith TJ, Ning SM, Lee MJ, Thomas PE, Yang CS. Modulation of 125:1218-24.
the levels of cytochromes P450 in rat liver and lung by dietary lipid. Bio- 56. O’Connor GT, Olmstead EM, Zug K, et al. Detection of hepatoxicity as-
chem Pharmacol 1992;43:2535-42. sociated with methotrexate therapy for psoriasis. Arch Dermatol 1989;
22. Barnett CR, Abbott RA, Bailey CJ, Flatt PR, Ioannides C. Cytochrome 125:1209-17.
P-450-dependent mixed-function oxidase and glutathione S-transferase 57. Geubel AP, De Galocsy C, Alves N, Rahier J, Dive C. Liver damage caused
activities in spontaneous obesity-diabetes. Biochem Pharmacol 1992;43: by therapeutic vitamin A administration: estimate of dose-related toxicity
1868-71. in 41 cases. Gastroenterology 1991;100:1701-9.
23. Ikemoto S, Imaoka S, Hayahara N, Maekawa M, Funae Y. Expression of 58. McDonald GB, Hinds MS, Fisher LD, et al. Veno-occlusive disease of the
hepatic microsomal cytochrome P450s as altered by uremia. Biochem liver and multiorgan failure after bone marrow transplantation: a cohort
Pharmacol 1992;43:2407-12. study of 355 patients. Ann Intern Med 1993;118:255-67.
24. Kurokohchi K, Yoneyama H, Matsuo Y, Nishioka M, Ichikawa Y. Effects 59. Silva MO, Roth D, Reddy KR, Fernandez JA, Albores-Saavedra J, Schiff
of interleukin 1 alpha on the activities and gene expressions of the cyto- ER. Hepatic dysfunction accompanying acute cocaine intoxication.
chrome P450IID subfamily. Biochem Pharmacol 1992;44:1669-74. J Hepatol 1991;12:312-5.
25. Wrighton SA, Stevens JC. The human hepatic cytochromes P450 involved 60. Boyer CS, Peterson DR. Potentiation of cocaine-mediated hepatotoxicity
in drug metabolism. Crit Rev Toxicol 1992;22:1-21. by acute and chronic ethanol. Alcohol Clin Exp Res 1990;14:28-31.
26. Seeff LB, Cuccherini BA, Zimmerman HJ, Adler E, Benjamin SB. Aceta- 61. Van Thiel DH, Perper JA. Hepatotoxicity associated with cocaine abuse.
minophen hepatotoxicity in alcoholics: a therapeutic misadventure. Ann Recent Dev Alcohol 1992;10:335-41.
Intern Med 1986;104:399-404. 62. Dalton TA, Berry RS. Hepatotoxicity associated with sustained-release ni-
27. Whitcomb DC, Block GD. Association of acetaminophen hepatotoxicity acin. Am J Med 1992;93:102-4.
with fasting and alcohol use. JAMA 1994;272:1845-50. 63. Henry JA, Jeffreys KJ, Dawling S. Toxicity and deaths from 3,4-methyl-
28. Brotodihardjo AE, Batey RG, Farrell GC, Byth K. Hepatotoxicity from enedioxymethamphetamine (“ecstasy”). Lancet 1992;340:384-7.
paracetamol self-poisoning in western Sydney: a continuing challenge. 64. Fallon MB, Boyer JL. Hepatic toxicity of vitamin A and synthetic reti-
Med J Aust 1992;157:382-5. noids. J Gastroenterol Hepatol 1990;5:334-42.
29. Lee WM. Acute liver failure. N Engl J Med 1993;329:1862-72. 65. Larrey D, Vial T, Pauwels A, et al. Hepatitis after germander (Teucrium
30. Smilkstein MJ, Douglas DR, Daya MR, et al. Acetaminophen poisoning chamaedrys) administration: another instance of herbal medicine hepato-
and liver function. N Engl J Med 1994;331:1310-2. toxicity. Ann Intern Med 1992;117:129-32.
31. Garibaldi RA, Drusin RE, Ferebee SH, Gregg MB. Isoniazid-associated 66. Loeper J, Descatoire V, Letteron P, et al. Hepatotoxicity of germander in
hepatitis: report of an outbreak. Am Rev Respir Dis 1972;106:357-65. mice. Gastroenterology 1994;106:464-72.
32. Mitchell I, Wendon J, Fitt S, Williams R. Anti-tuberculous therapy and 67. Katz M, Saibil F. Herbal hepatitis: subacute hepatic necrosis secondary to
acute liver failure. Lancet 1995;345:555-6. chaparral leaf. J Clin Gastroenterol 1990;12:203-6.
33. Ramakrishna B, Viswanath N. Diclofenac-induced hepatitis: case report 68. Chaparral-induced toxic hepatitis — California and Texas, 1992. MMWR
and literature review. Liver 1994;14:83-4. Morb Mortal Wkly Rep 1992;41:812-4.
34. Lindenbaum J, Leifer E. Hepatic necrosis associated with halothane anes- 69. MacGregor FB, Abernethy VE, Dahabra S, Cobden I, Hayes PC. Hepato-
thesia. N Engl J Med 1963;268:525-30. toxicity of herbal remedies. BMJ 1989;299:1156-7.
35. Gut J, Christen U, Huwyler J. Mechanisms of halothane toxicity: novel in- 70. Woolf GM, Petrovic LM, Rojter SE, et al. Acute hepatitis associated with
sights. Pharmacol Ther 1993;58:133-55. the Chinese herbal product jin bu huan. Ann Intern Med 1994;121:729-
36. Scheider DM, Klygis LM, Tsang TK, Caughron MC. Hepatic dysfunction 35.
after repeated isoflurane administration. J Clin Gastroenterol 1993;17:168- 71. Alberti-Flor JJ, Hernandez ME, Ferrer JP, Howell S, Jeffers L. Fulminant
70. liver failure and pancreatitis associated with the use of sulfamethoxazole-
37. Kleckner HB, Yakulis V, Heller P. Severe hypersensitivity to diphenylhy- trimethoprim. Am J Gastroenterol 1989;84:1577-9.
dantoin with circulating antibodies to the drug. Ann Intern Med 1975;83: 72. Larrey D, Vial T, Micaleff A, et al. Hepatitis associated with amoxycillin-
522-3. clavulanic acid combination: report of 15 cases. Gut 1992;33:368-71.
38. Pohl LR. Drug-induced allergic hepatitis. Semin Liver Dis 1990;10:305- 73. Hebbard GC, Smith KG, Gibson PR, Bhathal PS. Augmentin-induced
15. jaundice with a fatal outcome. Med J Aust 1992;156:285-6.
39. Mazuryk H, Kastenberg D, Rubin R, Munoz SJ. Cholestatic hepatitis as- 74. Westphal JF, Vetter D, Brogard JM. Hepatic side-effects of antibiotics.
sociated with the use of nafcillin. Am J Gastroenterol 1993;88:1960-2. J Antimicrob Chemother 1994;33:387-401.
Downloaded from www.nejm.org on September 19, 2003. This article is being provided free of charge for use in Indonesia:NEJM Sponsored.
Copyright © 1995 Massachusetts Medical Society. All rights reserved.
Vol. 333 No. 17 MEDICAL PROGRESS 1127

75. Chiu J, Nussbaum J, Bozzette S, et al. Treatment of disseminated Myco- 88. Davies MH, Harrison RF, Elias E, Hubscher SG. Antibiotic-associated
bacterium avium complex infection in AIDS with amikacin, ethambutol, acute vanishing bile duct syndrome: a pattern associated with severe, pro-
rifampin, and ciprofloxacin. Ann Intern Med 1990;113:358-61. longed, intrahepatic cholestasis. J Hepatol 1994;20:112-6.
76. Medina I, Mills J, Leoung G, et al. Oral therapy for Pneumocystis carinii 89. Wysowski DK, Freiman JP, Tourtelot JB, Horton ML III. Fatal and nonfa-
pneumonia in the acquired immunodeficiency syndrome: a controlled trial tal hepatotoxicity associated with flutamide. Ann Intern Med 1993;118:
of trimethoprim–sulfamethoxazole versus trimethoprim–dapsone. N Engl 860-4.
J Med 1990;323:776-82. 90. Meadow P, Tullio CJ. Glyburide-induced hepatitis. Clin Pharm 1989;8:470.
77. Saliba B, Herbert PN. Oxacillin hepatotoxicity in HIV-infected patients. 91. Knight TE, Shikuma CY, Knight J. Ketoconazole-induced fulminant hepa-
Ann Intern Med 1994;120:1048. titis necessitating liver transplantation. J Am Acad Dermatol 1991;25:398-
78. Schiff ER, Maddrey WC. Can we prevent nonsteroidal anti-inflammatory 400.
drug-induced hepatic failure? Gastrointest Dis Today 1994;3:7-13. 92. Michelson EJ. Labetalol hepatotoxicity. Ann Intern Med 1991;114:341.
79. Carson JL, Strom BL, Duff A, et al. Acute liver disease associated with 93. Larrey D, Babany G, Bernuau J, et al. Fulminant hepatitis after lisinopril
erythromycins, sulfonamides, and tetracyclines. Ann Intern Med 1993; administration. Gastroenterology 1990;99:1832-3.
119:576-83. 94. Raveh D, Arnon R, Israeli A, Eisenberg S. Lovastatin-induced hepatitis. Isr
80. O’Grady JG, Alexander GJM, Hayllar KM, Williams R. Early indicators of J Med Sci 1992;28:101-2.
prognosis in fulminant hepatic failure. Gastroenterology 1989;97:439-45. 95. Lopez-Navidad A, Domingo P, Cadafalch J, Farrerons J. Norfloxacin-
81. Zimmerman HJ, Lewis JH, Ishak KG, Maddrey WC. Ticrynafen-associat- induced hepatotoxicity. J Hepatol 1990;11:277-8.
ed hepatic injury: analysis of 340 cases. Hepatology 1984;4:315-23. 96. Blum A. Ofloxacin-induced acute severe hepatitis. South Med J 1991;84:
82. Powers BJ, Cattau EL Jr, Zimmerman HJ. Chlorzoxazone hepatotoxic re- 1158.
actions: an analysis of 21 identified or presumed cases. Arch Intern Med 97. Picon M, Causse X, Gelas P, Retornaz G, Trépo C, Bouletreau P. Hépatite
1986;146:1183-6. aiguë sévère à la pentamidine au cours du traitement d’une pneumocystose
83. Kellner M, Wiedemann K, Krieg JC, Berg PA. Toxic hepatitis by clozapine liée au SIDA. Gastroenterol Clin Biol 1991;15:463-4.
treatment. Am J Psychiatry 1993;150:985-6. 98. Planas R, De Leon R, Quer JC, Barranco C, Bruguera M, Gassull MA.
84. Purcell P, Henry D, Melville G. Diclofenac hepatitis. Gut 1991;32:1381- Fatal submassive necrosis of the liver associated with piroxicam. Am J
5. Gastroenterol 1990;85:468-70.
85. Keegan AD. Doxepin-induced recurrent acute hepatitis. Aust N Z J Med 99. Quinn PG, Sherman BS, Tavill AS, Gibas AL. Terbutaline hepatitis in preg-
1993;23:523. nancy: report of two cases and literature review. Am J Gastroenterol 1994;
86. Tran A, Housset C, Boboc B, Tourani JM, Carnot F, Berthelot P. Etopo- 89:781-4.
side (VP 16-213) induced hepatitis: report of three cases following stand- 100. Grimm IS, Litynski JJ. Severe cholestasis associated with ticlopidine. Am
ard-dose treatments. J Hepatol 1991;12:36-9. J Gastroenterol 1994;89:279-80.
87. Kano Y, Fukuda M, Shiohara T, Nagashima M. Cholestatic hepatitis occur- 101. Hull M, Jones R, Bendall M. Fatal hepatic necrosis associated with trazo-
ring shortly after etretinate therapy. J Am Acad Dermatol 1994;31:133-4. done and neuroleptic drugs. BMJ 1994;309:378.

Downloaded from www.nejm.org on September 19, 2003. This article is being provided free of charge for use in Indonesia:NEJM Sponsored.
Copyright © 1995 Massachusetts Medical Society. All rights reserved.

You might also like