Introduction

Many drugs have been reported to cause hepatic disease in animals; the most common ones are listed in Figure 1. Hepatotoxic drugs can be classified into those causing predictable hepatic damage (intrinsic toxicoses) and those that are idiosyncratic in their potential to cause hepatic damage. Drugs causing intrinsic hepatic damage have a high incidence of hepatotoxicity, are usually dose-dependent, predictable and can be reproduced in experimental animals. Idiosyncratic reactions occur in a small percentage of animals, occur randomly, are usually not dose or duration dependent and are difficult to reproduce experimentally. Idiosyncratic toxicosis is the result of an unusual susceptibility of an affected animal to an adverse reaction resulting from metabolic aberration, hypersensitivity, or immune-mediated events. Specific mechanisms of injury are usually unknown when idiosyncratic toxicosis occurs. In general, treatment of drug-induced hepatic disease involves withdrawal of the drug and supportive care. This discussion concentrates on the most important drug-induced hepatopathies.

Figure 1. Drugs known to cause hepatic disease.

 Acetaminophen

 Anabolic steroids

 Anticonvulsant drugs

 Antineoplastic drugs

 Arsenicals

 Azathioprine (?)

 Carprofen

 Diazepam

 Diethylcarbamazine

 Furosemide

 Glucocorticoids

 Griseofulvin

 Halothane

 Itraconazole

 Ketoconazole

 Mebendazole

 Methoxyflurane

 Mitotane (op'-DDD)

 Sulphonamides

 Tetracycline

 Trimethoprim-sulphonamides

Phenobarbital-Induced Hepatic Injury

Hepatobiliary disease associated with the administration of many anticonvulsant drugs, including primidone, phenobarbital and phenytoin (either alone or in combination), has been described. Laboratory abnormalities include variable increases in serum alanine aminotransferase (ALT), aspartate aminotransferase (AST), alkaline phosphatase (ALP) and gamma-glutamyl transpeptidase (GGT) activities. These changes are most marked in dogs receiving primidone and/or phenobarbital. Because these enzyme abnormalities occur in many asymptomatic dogs without significant hepatic injury, other tests of hepatic function should be used to determine whether there is impending hepatic failure, including serum bile acid and plasma ammonia concentrations. Increased total bilirubin and decreased albumin, blood urea nitrogen (BUN), glucose and cholesterol concentrations, although not specific, are common indicators of hepatic failure. If these tests are abnormal, hepatic biopsy or a decrease in the dosage of the anticonvulsant may be warranted.

The prognosis is poor when histological lesions are severe and hepatic failure has occurred. Treatment involves withdrawal of anticonvulsant drugs if possible or use of alternative anticonvulsant drugs such a potassium bromide. There is no indication for the use of glucocorticoids unless there is an active inflammatory component. The use of ursodeoxycholic acid may also be indicated due to the presence of a significant cholestatic component in most cases. The use of S-adenosylmethionine (SAMe) and/or vitamin E is warranted due to oxidative injury occurring with the necroinflammatory process.

Carprofen Toxicity

Carprofen is a non-steroidal antiinflammatory drug (NSAID) commonly used to treat canine osteoarthritis. In a recent Pfizer Animal Health Technical Bulletin, the report rate for adverse reactions involving the liver was 5.2 cases per 10,000 dogs treated (0.052%), of which 1.4 cases per 10,000 dogs had clinical or laboratory evidence of hepatic failure. At the time of this Bulletin, 4 million dogs had received carprofen in the US. In one report, 21 dogs were described to have hepatocellular toxicosis associated with administration of carprofen. At the time of the report, over 500,000 dogs had received carprofen. This report provides valuable information about carprofen toxicosis.

The results of this study suggest that the drug reaction to carprofen is idiosyncratic and host-dependent in nature. Progression of the condition did not appear to correlate with the dose of carprofen, magnitude of hepatic enzyme activities, or histopathological severity of hepatic lesions. It is also noteworthy that renal lesions were detected in a number of dogs, a well documented side effect of other NSAIDs. Although prescreening haematological and serum biochemical analyses may not yield results that can be used to predict dogs that will have adverse reactions to carprofen, evaluation of renal and hepatic function prior to administration of the drug is recommended. Dogs with renal and hepatobiliary abnormalities may be poor candidates to receive this drug, or extra caution should be used if carprofen is to be used in these dogs. In addition, serum biochemistry analysis should be obtained approximately 2-3 weeks after starting carprofen to detect patients with developing hepatic or renal disease. Clients should be informed of the clinical signs of drug intolerance and instructed to immediately discontinue the drug if these signs develop. There is no need to obtain a yearly serum chemistry panel to look for carprofen hepatoxicity since the idiosyncratic reaction is acute in most dogs.

Steroid Hepatopathy

Steroid hepatopathy can result from excessive endogenous (hyperadrenocorticism) or administration of exogenous glucocorticoids. It represents one of the most common causes of increased serum hepatic enzyme activities and is the most common diagnosis on hepatic biopsy specimen analysis in dogs in one study. Steroid hepatopathy occurs only rarely in the cat. The likelihood of steroid hepatopathy developing in individual animals following glucocorticoid administration is variable, and depends most importantly on individual sensitivity to glucocorticoids. In addition, the development of steroid hepatopathy also depends on the type of, route of and duration of administration of the glucocorticoid. Changes can persist for several months after a single injection of a long-acting glucocorticoid or after long-term administration of oral glucocorticoids. Changes can also occur after topical, aural or ocular administration of glucocorticoids.

Clinical Signs

Clinical signs of steroid hepatopathy range from asymptomatic to those associated with glucocorticoid excess. There are usually no signs specifically related to hepatic failure, with the exception of lethargy in severe cases. Severe hepatic failure rarely, if ever, occurs with steroid hepatopathy.

Laboratory Findings

There are usually mild to moderate increases in serum ALT and AST activities and marked increases in ALP and GGT activities in dogs with steroid hepatopathy. The hepatic enzyme elevations are variable. Occasionally the magnitude of elevation in activities of the transaminases (ALT and AST) exceeds the magnitude of elevations of serum ALP and GGT activities. Albumin and serum bilirubin concentrations are almost always normal (when these are abnormal, other causes should be looked for). Often the laboratory abnormalities seen with primary non-hyperbilirubinaemic hepatobiliary disease and with steroid hepatopathy are similar. The presence of increased serum bilirubin concentration eliminates steroid hepatopathy from the differential diagnosis of hepatobiliary disease in almost all cases.

The increase in ALP activity is attributed to an isoenzyme that is different from that induced by biliary stasis. This isoenzyme, also produced in the liver, is referred to as the steroid-induced isoenzyme of ALP. Laboratory methods are available to distinguish the steroid-induced isoenzyme from that induced by biliary stasis. However, the steroid-induced isoenzyme of ALP is variably elevated with many primary hepatobiliary diseases, and therefore measurement of its activity is not a useful test to determine the presence of steroid hepatopathy. Although there is not a specific isoenzyme of GGT induced by glucocorticoids, it is important to note that the magnitude of increased serum GGT activity induced by glucocorticoids often parallels that of ALP, and that serum GGT activity cannot be used to distinguish steroid hepatopathy from primary hepatobiliary disease.

Results of hepatic function tests, including serum bile acid and blood ammonia concentrations, are variable with steroid hepatopathy. Serum bile acids can be normal or slightly to moderately elevated (less than 75-100 µmol/l). Marked increases in serum bile acids (>100 µmol/l) are unlikely to result from steroid hepatopathy, and other hepatobiliary diseases should be considered.

Diagnosis

If laboratory and clinical signs are compatible with steroid hepatopathy or the animal is asymptomatic, steroid hepatopathy should be ruled out prior to hepatic biopsy. This can be done with a history of glucocorticoid administration or appropriate laboratory tests such as the adrenocorticotrophic hormone (ACTH) stimulation or low-dose dexamethasone suppression tests. If the diagnosis is still uncertain, hepatic and adrenal imaging with ultrasound and hepatic biopsy (ideally via laparoscopy) are warranted. Biopsy specimen analysis usually readily distinguishes steroid hepatopathy from other hepatic diseases. Hepatic biopsy specimen analysis is also another method of diagnosing hyperadrenocorticism when other laboratory tests are inconclusive.

Treatment

Treatment involves elimination of the source of excess glucocorticoids if clinically appropriate. If the source is exogenous administration, corticosteroids should be discontinued if their administration is not necessary. If the use of corticosteroids is considered an important component of the patient's treatment and there are no clinical signs specifically referable to steroid hepatopathy, the dose of the corticosteroid should not be altered. In these cases, the biochemical changes should be considered a 'laboratory phenomenon' rather than a clinically relevant problem. If there are potential clinical signs referable to the use of corticosteroids and the patient's disease is controlled, the dosage should be decreased by at least 50% if their ongoing use is considered important. If clinical signs such as lethargy are present and persist, alternative immune-modulating drugs may be substituted if appropriate. The laboratory and morphological changes in the liver seen with steroid hepatopathy are completely reversible when the source of excess glucocorticoids is removed.

References

1.  MacPhail CM, Lappin MR, et al. Hepatocellular toxicosis associated with administration of carprofen in 21 dogs. Journal of the American Veterinary Medical Association 1998: 212: 1895-1901.