Avoiding drug-induced liver injury

Abstract

Liver injury is estimated to occur in up to 19 cases per 100,000 persons, with up to 5% of hospital admissions for liver injury attributed to medication use. Despite the low proportion of hospitalizations, more than one-half of acute liver failures are drug induced. Diagnosing drug-induced liver injury is done by excluding other causes. However, considering the timing of injury as it relates to new drug use, what happens when the suspected drug is stopped or restarted, as well as the incidence of known liver injury with that drug and the overall presentation of the patient can be useful when performing this diagnosis of exclusion. In a recent article published in the New England Journal of Medicine, Hoofnagle and Björnsson recognized three types of drug-induced liver injury. The first, direct hepatotoxicity, occurs when a drug intrinsically toxic to the liver is taken, and the injury results in 5 days or less of starting the drug. These cases often occur as a result of drug overdoses, with acetaminophen overdose a common example. Idiosyncratic hepatotoxicity is when a drug causes liver injury in rare cases and is unpredictable because of the low incidence of injury (1 in 2,000 or greater patient exposures). Indirect hepatotoxicity occurs when a drug promotes liver injury as part of its mechanism of action instead of a direct insult to the liver. An example is when a drug prompts the patient's immune system to attack the liver. These three definitions of drug-induced liver injury have different clinical patterns of presentation called phenotypes, which are summarized by Hoofnagle and Björnsson in their article. In direct hepatotoxicity, the injury presents with elevated ALT or alkaline phosphatase (ALP) levels, but without hyperbilirubinemia. Presentation occurs within days of a new drug or a dose increase to an existing drug. These elevations either spontaneously resolve (as a result of the patient's liver adapting to the presence of the drug) or worsen and progress to jaundice or worse. With idiosyncratic hepatotoxicity, the injury presents with elevated ALT, but only slightly elevated ALP. Presentation occurs within 90 days of a drug addition or dose change to an existing drug. This phenotype can be further classified into hepatocellular injury, chronic hepatitis, cholestatic hepatitis, and drug-induced mixed hepatitis. All of these subclassifications can also present with rash, fever, and eosinophilia. With indirect hepatotoxicity, the phenotype can include fatty liver disease, which is caused by drug-induced weight gain (as is the case with many antipsychotic medications). Acute hepatitis can be induced by reactivation of a latent infection, such as hepatitis B, which can occur with some chemotherapy agents. These cases typically present within 2 to 12 weeks after starting the drug. This phenotype definition is new and somewhat controversial. A review of 889 cases of drug-induced liver injury found that 9 of the top 10 drugs implicated are antimicrobials, with amoxicillin/clavulanate at the top of the list. It is curious that only 3 of the top 25 of these drugs were approved by FDA since 2000. Herbal and nutritional supplements are implicated in up to 20% of liver injury cases in the United States. Most of these cases involve products that contain multiple ingredients, making identifying the exact agent that caused the injury impossible. Further muddying the waters is that many multiagent dietary supplements are mislabeled and/or adulterated. Alarmingly, the most common injury associated with supplement use is acute hepatocellular hepatitis, which can necessitate a liver transplant. Pharmacists should be vigilant when monitoring patients with liver injury for drug-related causes. They should also strongly discourage the use of herbal and nutritional supplements in all patients, especially those who are at risk of developing liver injury.