Drug-Induced Hepatotoxicity

ACE Inhibitors

Hepatic injury occurs occasionally with ACE inhibitors. Captopril and enalapril are implicated in most reported cases, but other ACE inhibitors likely have similar hepatotoxic potential. Most cases show cholestatic injury, but mixed and hepatocellular damage also are reported.1,2

Alcohol

Fatty infiltration of the liver occurs in 70–100% of alcoholics. Fatty liver is generally without clinical manifestation, but 30% of alcoholics develop alcoholic hepatitis and about 10% develop cirrhosis. Malnutrition can potentiate alcoholic liver disease, and alcohol can enhance the hepatotoxicity of other drugs.1 Aldesleukin Increases in serum bilirubin, alkaline phosphatase, and transaminases occur frequently. These primarily cholestatic changes are rapidly reversible after drug discontinuation.

Allopurinol

Hepatic granulomas, hepatitis and hepatic necrosis can accompany other symptoms (especially rash, fever, eosinophilia, and vasculitis) of allopurinol hypersensitivity. Damage is usually focal, but widespread damage also is reported. This reaction is rare but serious when it occurs. Onset is usually after 3–6 weeks of treatment. Renal impairment might be a predisposing factor for allopurinol-induced hepatitis. Cholestasis also has been attributed to allopurinol.

Aminoglutethimide

Laboratory evidence of cholestasis is common, but clinical evidence is rare.1,8 Aminosalicylic Acid Up to 5% of patients develop a generalized hypersensitivity reaction. About 25% of these patients have evidence of mixed cholestatic and hepatocellular injuries as part of their hypersensitivity reactions. Fatalities have been reported.

Acetaminophen

Centrilobular hepatic necrosis can follow acute overdose with ≥140 mg/kg in children or ≥6 g in adults. These doses saturate the normal metabolic pathways, producing large quantities of a hepatotoxic metabolite. Children appear to have a lower risk than adults of developing acetaminophen-induced hepatitis.

Laboratory evidence of hepatotoxicity peaks 3–4 days after the acute exposure.

Therapy with acetylcysteine to bind the metabolite is indicated when the 4-hr postingestion serum acetaminophen level is >150 mg/L. Even without acetylcysteine, fatalities are uncommon after acetaminophen overdose. Nonfatal cases usually recover fully in a few weeks. Chronic alcohol ingestion increases acetaminophen toxicity, as does recent fasting.

Acute alcohol ingestion is thought by some to have a protective action. Less destructive, but still detectable, hepatitis is reported in patients taking large acetaminophen doses for therapeutic purposes.1,3–5

Amiodarone

Mild increases in transaminases and LDH levels occur in up to one-half of patients, whereas phospholipidosis occurs in virtually all; normal values often return despite continued therapy. Symptoms (eg, jaundice, nausea and vomiting, hepatomegaly, or weight loss) occur in 1–4% of patients. Onset is typically after 2–4 months of therapy but can be delayed for ≥1 yr. Recovery after drug discontinuation can take from several months to ≥1 yr. The dose-related hepatotoxicity of amiodarone is reminiscent of alcoholic hepatitis. Cirrhosis and fatalities are also reported.

Amoxicillin and Clavulanic Acid

Based on an extensive review of medical records, the frequency of acute hepatic injury with amoxicillin and clavulanic acid is 1.7 cases/10,000 prescriptions (compared with 0.3 for amoxicillin alone). In most cases, the hepatic injury is cholestatic. The risk of hepatic injury is increased by repeated prescriptions for amoxicillin and clavulanic acid and by advancing age.

Androgens

(See Steroids, C-17--Alkyl.)

Antidepressants,

Heterocyclic The prevalence of hepatic injury is estimated at about 1%, with most of the cases presenting as cholestasis. This idiosyncratic reaction resembles the cholestasis associated with phenothiazines.

Asparaginase

Slowly reversible steatosis occurs in 50–90% of patients, apparently due to the drug’s influence on protein synthesis. Daily administration might be more hepatotoxic than weekly administration.

Azathioprine

This drug is less hepatotoxic than its metabolite, mercaptopurine. Azathioprine’s hepatotoxicity is predominantly cholestatic rather than hepatocellular. Vascular lesions, including venous occlusion and peliosis hepatis, have been reported, but their prevalence is unknown. Nodular regenerative hyperplasia has followed use of this drug in kidney and liver transplantations.

Busulfan

Use in bone marrow transplant patients is associated with apparently dose-related veno-occlusive disease of the liver. Although the exact contribution of the drug is difficult to discern, this syndrome occurs in 20% of adults and 5% of children with total doses ≥16 mg/kg.

Carbamazepine

Mild changes in liver function tests occur frequently. Hepatic necrosis, granulomas, and cholestasis have occurred, with some cases showing signs of hypersensitivity. Onset is most often in the first 4 weeks of therapy. Fatalities have been reported.

Carmustine

Changes in liver function tests in 20–30% of patients, from a few days to several weeks after drug administration. Changes are usually mild and resolve quickly with drug discontinuation.

Cephalosporins

Transient minor increases in AST, ALT, and alkaline phosphatase occur frequently. Ceftriaxone use is associated with development of “gallbladder sludge” in up to 25% of patients.

Chlorpropamide

Most hepatotoxic reactions are cholestatic and probably are caused by an immune mechanism. Prevalence is estimated at 0.5–1.5%, with onset usually within the first 2 months of therapy.

Chlorzoxazone

Idiosyncratic hepatocellular damage occurs rarely, but fatalities have been reported. Discontinue the drug if elevated levels of transaminases or bilirubin are detected.

Cisplatin

Transient, dose-related elevations of hepatic enzymes occur frequently.

Clozapine

Transient elevations of hepatic enzymes occur frequently during the first 3 months of clozapine use. Although several cases of fulminant hepatitis have been reported, the risk of serious clozapineinduced hepatotoxicity remains small and some investigators recommend against routine testing.

Cocaine

Hepatic necrosis has been reported in cases of cocaine abuse, including at least one fatality. The prevalence of this reaction is not known.

Contraceptives

, Oral Data from two large, long-term cohort studies (about 33,000 users) did not detect any association between oral contraceptive use and serious liver disease. One study detected an increase in the frequency of mild liver disease among users of older, high-estrogen (>50 g) products. Older combination oral contraceptives were associated with an increase in the annual incidence of hepatic adenomas (3.4/100,000 vs 1.3/100,000 in nonusers), especially after ≥5 yr of use. The frequency of gallbladder disease also was increased by older oral contraceptives.

Cyclosporine

Elevated serum levels of alkaline phosphatase and conjugated bilirubin consistent with cholestasis occur in 50–60% of patients. These changes are usually mild and pose little threat.

Dantrolene

At least 1.8% of patients develop laboratory evidence of hepatic dysfunction, with symptomatic hepatitis in about 0.6%; the fatality rate among jaundiced patients is about 25%. Predisposing factors seem to include dosage (>300 mg/day), sex (women more than men), age (>30 yr), and duration of therapy (≥2 months).

Dapsone

Hepatitis can occur as part of the “dapsone syndrome,” a generalized hypersensitivity reaction that includes rash, fever, and lymphadenopathy. The true prevalence is unknown but might be as high as 5%. The onset is usually during the first 2 months of therapy. Although most dapsone-associated liver injury is hepatocellular, some cases of cholestasis have occurred.

Disulfiram

Small increases in serum transaminase levels occur frequently. Hepatitis is reported occasionally, which can be caused by hypersensitivity. Most cases develop during the first few months of treatment. The best estimate of the incidence of fatal hepatitis is about 1/30,000 users/year.

Erythromycin

Erythromycin was thought to be a frequent cause of jaundice, but recent studies indicate that jaundice occurs only occasionally. Cholestasis apparently results from hypersensitivity (60% have eosinophilia and 50% have fever), appearing after 10–14 days of initial therapy or after 1–2 days in patients with a history of erythromycin exposure. Despite extensive use in children, most cases are reported in adults. Rapid reversal of symptoms follows drug discontinuation, but laboratory (

changes can persist for up to 6 months. Although most cases involve the estolate salt, hepatotoxicity has occurred with the ethylsuccinate, stearate, and propionate salts and with erythromycin base.

Ethionamide

Hepatitis can occur in 3–5% of patients, and serum enzyme elevations can occur in ≥30%. Onset of hepatitis is usually after several months of therapy.

Felbamate

Although the prevalence of hepatocellular destruction is unclear, it is of sufficient concern to limit the use of felbamate to carefully selected patients. At least 6 cases of fatal felbamate-induced hepatic necrosis have been reported.

Ferrous Salts

Hepatic necrosis can appear within 1–3 days of an acute overdose. The fatality rate is high if the patient is not treated promptly.

Floxuridine

Hepatic arterial infusion of floxuridine results in 9% sclerosing cholangitis at 9 months and 26% after 1 yr. Elevations of liver enzyme levels are common but not predictive of greater hepatotoxicity.

Flutamide

Through 1994, there were as least 20 reported deaths reasonably attributed to flutamide-induced hepatotoxicity. Those deaths, typically the result of massive hepatic necrosis, occurred between 5 days and 9 months (mean 3 months) after initiation of flutamide therapy. Further, the hospitalization rate for noninfectious liver disease in flutamide-treated patients was 10 times the expected rate. Monthly liver function testing is recommended for the first 4 months.

Gold Salts

Cholestasis occurs occasionally with normal doses of parenteral gold salts; hypersensitivity is the suspected mechanism. Onset is commonly within the first few weeks of therapy, and recovery usually occurs within 3 months after drug discontinuation. Lipogranulomas are frequently found in liver biopsies of parenteral gold-treated patients. These can persist long after drug withdrawal but do not seem to impair liver function. Hepatic necrosis can result from overdose.

Halothane

As many as 30% of patients have increased serum transaminases or other evidence of mild hepatic impairment. Despite extensive publicity, the actual frequency of severe halothane hepatitis is low, ranging from 1/3500 to 1/35,000, with reported case fatality rates of 14–67%. Susceptibility is greatest in adults, women, obese patients, and especially in patients with prior exposure to halothane. The mechanism of hepatitis is poorly understood, but hypersensitivity is most likely. Fever precedes jaundice in most patients. The onset of jaundice is usually 5–8.5 days after exposure but can occur 1–26 days after exposure; shorter latent periods are associated with prior halothane exposure.

Methoxyflurane and enflurane produce similar hepatotoxic reactions, although less frequently.1,48,49 Histamine H2-Receptor Antagonists Cimetidine and ranitidine are associated with increased liver enzymes. The risk of acute liver injury with cimetidine is about 1/5000, with most cases occurring during the first 2 months of use.

Isoniazid

Elevated serum transaminase levels occur frequently, are presumed to be associated with subclinical hepatitis, resolve rapidly after drug discontinuation, and can resolve despite continued isoniazid therapy. A syndrome resembling viral hepatitis occurs in 1–2% of patients, with the onset usually during the first 20 weeks of therapy.

Alcohol

consumption increases the risk of hepatotoxicity; the contribution of concomitant rifampin is poorly defined. The role of acetylator phenotype remains unclear, but a case-control study found that patients admitted to the hospital for suspected isoniazid-induced hepatotoxicity were significantly more likely to be slow acetylators than those who completed their courses of therapy without hepatotoxicity.

Itraconazole

The FDA has received reports of liver failure and death apparently associated with itraconazole use, included some cases without predisposing risk factors.98 Ketoconazole Elevated hepatic enzyme levels occur in about 20% of ketoconazole-treated patients, with overt hepatitis in 3%. The typical onset for overt hepatitis is 30–60 days after initiation of ketoconazole therapy. There have been a few deaths attributed to ketoconazole hepatotoxicity.

Lamotrigine

At least 9 cases of lamotrigine-associated hepatotoxicity have been published, including at least 1 case of severe hepatic failure. Most of these cases were complicated by multiple-drug therapy.

Mercaptopurine

Jaundice associated with cholestasis, hepatic necrosis, and mixed reactions occurs in 6–40% of patients, with the highest prevalence associated with doses ≥2 mg/kg/day. Onset is usually during the first 2 months of therapy.

Methotrexate

Hepatic injury (macrovesicular steatosis, necrosis, and bridging fibrosis) occurs frequently, depends on dose and duration of therapy, and can progress to cirrhosis if the drug is not stopped. Intermittent high doses pose less risk than daily low doses. Cirrhosis is reported in up to 24% of patients receiving long-term daily doses; other contributing factors are alcoholism and pre-existing liver or kidney disease. Hepatic fibrosis is not detected by standard liver function tests and is best detected by biopsy. Biopsy has been recommended at intervals of up to 36 months, after every 1.5 g of methotrexate, if 6 of 12 monthly transaminase levels are elevated, or if the serum albumin level drops below normal. Isolated elevations of transaminase levels do not preclude continued methotrexate therapy.

Methyldopa

Mild changes in liver function tests occur in up to 35% of patients taking methyldopa, but the prevalence of clinical hepatitis is probably <1%. Most cases occur during the first 3 months of therapy. Hepatitis is more common in women, and most patients have rapid recovery after drug discontinuation. The fatality rate is <10% among patients who develop hepatitis. There is evidence to support a hypersensitivity mechanism in some patient

Minocycline

The long-term use of minocycline for acne or arthritis has resulted in at least 65 reported cases of minocycline-induced hepatitis. Autoimmune hepatitis associated with lupus-like symptoms occurs with a median onset of 1 yr, and an apparent hypersensitivity mechanism is responsible for other cases occurring during the first month of minocycline therapy.

Nevirapine

Severe, life-threatening hepatotoxicity has been reported in patients taking nevirapine for HIV infection and health care workers taking the drug for postexposure prophylaxis. Fatalities have occurred in HIV-infected patients.

Niacin

Elevations of hepatic enzyme and bilirubin levels occur in 30–50% of patients taking sustainedrelease niacin in therapeutic doses, with jaundice in 3% of patients taking 3 g/day for >1 yr. Symptomatic hepatic dysfunction occurs frequently and limits the use of the sustained-release product. Immediate-release niacin also is hepatotoxic but to a lesser extent than sustained-release.

Nitrofurantoin

Hepatic damage occurs occasionally, usually during the first month of therapy. Cholestasis is the most common presentation; hepatic necrosis also is reported. Hypersensitivity is the suspected mechanism, and the onset is frequently associated with fever, rash, and eosinophilia.

Nonsteroidal Anti-inflammatory Drugs

The incidence of clinically apparent hepatic injury from nonsalicylate NSAIDs is estimated to be about 1/10,000 patient–years. The incidence for sulindac may be 5–10 times higher than for the other nonsalicylate NSAIDs. Half of the reactions to sulindac are cholestatic and 25% are hepatocellular. Despite previous reports to the contrary, current data analysis does not support a higher incidence of hepatotoxicity with diclofenac

Octreotide

Most patients on long-term therapy develop cholelithiasis and/or gallbladder sludge; some require cholecystectomy. The prevalence and speed of onset of symptoms might be dosage related.

Papaverine

Numerous reports of hepatocellular injury and elevated liver enzymes in 27–43% of patients indicate a marked hepatotoxic potential.

Pemoline

Pemoline occasionally causes elevated liver enzymes. The prescribing information for pemoline includes a boxed warning describing 15 cases of acute hepatic failure reported to the FDA between 1975 and 1998; 12 cases resulted in death or required liver transplantation. The earliest onset of hepatic abnormalities in these cases occurred 6 months after the start of pemoline therapy. The few published reports of possible pemoline-induced fulminant hepatic failure do not hold up well under close scrutiny.

Penicillamine

Cholestasis resulting from a hypersensitivity reaction occurs occasionally.

Penicillins Cloxacillin and flucloxacillin

are rarely associated with cholestatic hepatitis. The effect is reversible but can persist for months after drug discontinuation.

Phenothiazines

Most reports of liver damage involve chlorpromazine. The prevalence of hepatic enzyme elevation with this drug has been estimated to be as high as 42%, although 10% is probably more realistic. Similarly, cholestatic jaundice has been projected to occur in up to 5% of patients receiving chlorpromazine, but the actual prevalence is closer to 1%. The onset of cholestasis is generally in the first month of therapy and usually follows a prodrome of GI or influenza-like symptoms.

Phenytoin

Hepatocellular necrosis is occasionally associated with phenytoin therapy, usually accompanied by other signs of hypersensitivity (eg, eosinophilia, fever, rash, and lymphadenopathy).

Plicamycin

Laboratory evidence of dose-related hepatotoxicity occurs in virtually all patients. A common lesion is perivenous necrosis.

Progestins

(See Steroids, C-17--Alkyl.)

Propoxyphene

A small number of cases of propoxyphene-induced cholestasis have been reported;

Propylthiouracil

Increased ALT levels occur in up to 30% of patients. Onset is usually within the first 2 months of therapy, and ALT levels commonly return to normal with dosage reduction.

Pyrazinamide

Pyrazinamide-induced hepatitis depends on dose and duration of therapy. Daily administration appears to present a greater risk than weekly administration

Riluzole

Elevated hepatic enzymes occur frequently; the prevalence appears to be dosage related.

Ritonavir

Elevations of serum AST and ALT greater than 3.6 times base line occur in 30% of patients treated with ritonavir.

Quinidine

Hepatic damage is rare and usually accompanied by other signs of hypersensitivity, especially fever. Most reactions occur in the first month of therapy. The pathology is usually a mixture of hepatocellular necrosis and cholestasis; granulomas also have been reported.

Salicylates

Up to 50% of patients taking antiarthritic dosages have laboratory evidence of liver damage. The risk of liver damage is greatest in patients with connective tissue disorders such as SLE or juvenile rheumatoid arthritis.

Sulfasalazine

A small number of cases of sulfasalazine-associated hepatic damage, including fatalities, have been reported in children and adults. Hepatic necrosis is apparently part of a generalized hypersensitivity reaction that includes rash, fever, and lymphadenopathy.

Tacrine

. Most increases were detected in the first week of therapy. Most patients’ ALT levels returned to no more than twice the ULN within 1 month after drug discontinuation, and no patients developed jaundice.

Terbinafine

The FDA has received reports of liver failure and death apparently associated with oral terbinafine use, including some cases without predisposing risk factors.

Tetracycline

Microvesicular steatosis can occur in patients receiving large doses of tetracycline IV, usually >1.5 g/day. Contributing factors include pregnancy, malnutrition, and impaired renal function, but hepatotoxicity has been reported in patients with none of these factors.

Tolcapone

ALT levels increase to >3 times the upper limit of normal in about 8% of tolcapone-treated patients. These elevations usually develop 6–12 weeks after the start of tolcapone use and can resolve despite continued therapy.

Trimethoprim-Sulfamethoxazole

“Clinically important” liver disease occurs in at least 5.2/100,000 patients (3.8/100,000 with trimethoprim alone).

Troleandomycin

From 30 to 50% of patients receiving the drug show some laboratory evidence of abnormal liver function, and up to 4% develop jaundice.1 Valproic Acid Hepatic enzyme elevations occur in 7–44% of patients, with clinically apparent liver disease in 0.05–1%.

Vitamin A

Hepatomegaly, portal hypertension, and mild increases in liver enzyme levels are common features of chronic vitamin A toxicity.

Zafirlukast

Asymptomatic hepatic enzyme elevations occur frequently. At least three cases of severe hepatitis have been reported including one that resulted in liver ttransplantation.

This table includes only those drugs with well-established records of hepatotoxicity. A drug not listed in the table does not mean it cannot produce liver damage because virtually all drugs have been reported to elevate serum liver enzymes. Combining drugs that have hepatotoxic potential commonly results in greater than additive liver damage. In general, drug-induced hepatotoxicity is most prevalent in older patients, women, and those with pre-existing hepatic impairment.

Reference

1-Zimmerman HJ. Hepatotoxicity: the adverse effects of drugs and other chemicals on the liver. 2nd ed. Philadelphia: Lippincott Williams & Wilkins; 1999. 2. Hagley MT et al. Hepatotoxicity associated with angiotensin-converting enzyme inhibitors. Ann Pharmacother 1993;27:228–31. 3. Whitcomb DC, Block GD. Association of acetaminophen hepatotoxicity with fasting and ethanol use. JAMA 1994;272:1845–50. 4. Heubi JE et al. Therapeutic misadventures with acetaminophen: hepatotoxicity after multiple doses in children. J Pediatr 1998;132:22–7. 5. Tolman KG. Hepatotoxicity of non-narcotic analgesics. Am J Med 1998;105(1B):13S–19S. 6. Fisher B et al. Interleukin-2 induces profound reversible cholestasis: a detailed analysis in treated cancer patients. J Clin Oncol 1989;7:1852–62. 7. Arellano F, Sacristán JA. Allopurinol hypersensitivity syndrome: a review. Ann Pharmacother 1993;27:337–43. 8. Nagel GA et al. Phase II study of aminoglutethimide and medroxyprogesterone acetate in the treatment of patients with advanced breast cancer. Cancer Res 1982;42(suppl):3442S–4S. 9. Simpson DG, Walker JH. Hypersensitivity to para-aminosalicylic acid. Am J Med 1960;29:297–306. 10. Lee WM. Drug-induced hepatotoxicity. N Engl J Med 1995;333:1118–27. 11. Guigui B et al. Amiodarone-induced hepatic phospholipidosis: a morphological alteration independent of pseudoalcoholic liver disease. Hepatology 1988;8:1063–8. 12. Richer M, Robert S. Fatal hepatotoxicity following oral administration of amiodarone. Ann Pharmacother 1995;29:582–6. 13. Garcia Rodriguez LA et al. Risk of acute liver injury associated with combination of amoxicillin and clavulanic acid. Arch Intern Med 1996;156:1327–32. 14. Perry MC. Chemotherapeutic agents and hepatotoxicity. Semin Oncol 1992;19:551–65. 15. Pratt CB et al. Comparison of daily versus weekly L-asparaginase for the treatment of childhood acute leukemia. J Pediatr 1970;77:474–83. 16. Pratt CB, Johnson WW. Duration and severity of fatty metamorphosis of the liver following L-asparaginase therapy. Cancer 1971;28:361–4. 17. Gane E et al. Nodular regenerative hyperplasia of the liver graft after liver transplantation. Hepatology 1994; 20:88–94. 18. Grochow LB et al. Pharmacokinetics of busulfan: correlation with veno-occlusive disease in patients undergoing bone marrow transplantation. Cancer Chemother Pharmacol 1989;25:55–61. 19. Vassal G et al. Busulfan and veno-occlusive disease of the liver. Ann Intern Med 1990;112:881. Letter. 20. Horowitz S et al. Hepatotoxic reactions associated with carbamazepine therapy. Epilepsia 1988;29:149–54. 21. Thompson JW, Jacobs RF. Adverse effects of newer cephalosporins. An update. Drug Saf 1993;9:132–42. 22. Schoenfield LJ et al. Chenodiol (chenodeoxycholic acid) for dissolution of gallstones: the National Cooperative Gallstone Study. A controlled trial of efficacy and safety. Ann Intern Med 1981;95:257–82. 23. Labeling change. FDA Med Bull 1996;26(1):3. 24. MacFarlane B et al. Fatal fulminant liver failure due to clozapine: a case report and review of clozapine-induced hepatotoxicity. Gastroenterology 1997;112:1707–9. 25. Hummer M et al. Hepatotoxicity of clozapine. J Clin Psychopharmacol 1997;17:314–7. 26. Wanless IR et al. Histopathology of cocaine hepatotoxicity. Report of four patients. Gastroenterology 1990;98:497–501. 27. Lindberg MC. Hepatobiliary complications of oral contraceptives. J Gen Intern Med 1992;7:199–209. 28. Hannaford PC et al. Combined oral contraceptives and liver disease. Contraception 1997;55:145–51. 29. Atkinson K et al. Cyclosporine-associated hepatotoxicity after allogeneic marrow transplantation in man: differentiation from other causes of posttransplant liver disease. Transplant Proc 1983;15(suppl 1):2761–7. 30. Kassianides C et al. Liver injury from cyclosporine A. Dig Dis Sci 1990;35:693–7. 31. Utili R et al. Dantrolene-associated hepatic injury. Incidence and character. Gastroenterology 1977;72:610–6.