Antiretroviral Drugs and Liver Injury

Vincent Soriano; Massimo Puoti; Pilar Garcia-Gascó; Juergen K Rockstroh; Yves Benhamou; Pablo Barreiro; Barbara McGovern


AIDS. 2008;22(1):1-13. 

In This Article


Antiretroviral drug-related liver injury (ARLI) is a common cause of morbidity, mortality and treatment discontinuation in HIV-infected patients.[1] Prevention and management of ARLI have emerged as major issues among HIV-infected patients in the era of HAART.[2] Virtually every licensed antiretroviral medication has been associated with liver enzyme elevations, although certain drugs may cause liver injury more frequently than others. In addition, certain comorbidities, such as chronic hepatitis B (HBV) or hepatitis C (HCV) infection, may predispose patients to ARLI.[3] Several major mechanisms of ARLI have been described, including metabolic host-mediated injury, hypersensitivity reactions, mitochondrial toxicity, and immune reconstitution phenomena. The management of ARLI should be based on its clinical severity and underlying pathogenic mechanism. Herein, we propose use of a universal standard for defining liver injury to enable comparisons between future studies. Novel mechanisms for hepatotoxicity are also discussed along with preventive measures to avoid the onset of ARLI.

ARLI is defined by elevations in liver enzymes in serum, with alanine aminotransferase (ALT) characteristically greater than aspartate aminotransferase (AST). To date, there has been broad variability in the criteria used in clinical studies to categorize the severity of hepatotoxicity. Some studies have utilized ALT parameters as minimal as two times the upper limits of normal[4] while others have employed an absolute threshold (e.g., >100 IU/ml), regardless of baseline liver function tests.[5] The clinical relevance of these elevations is uncertain.

More recently, the AIDS Clinical Trials Group (ACTG) has defined a grading scheme against the patient's baseline serum aminotransferase concentrations. For example, in patients with a normal pretherapy ALT or AST, hepatic injury is graded as moderate or severe based on a 5-fold or 10-fold increase in aminotransferases, respectively.[6] In patients with abnormal liver enzymes prior to therapy, a >3.5-fold or a 5-fold increase in ALT or AST is considered indicative of moderate or severe hepatotoxicity, respectively.[7]

Liver function test abnormalities require careful interpretation. On the one hand, some drugs (e.g., nevirapine and less frequently efavirenz) increase γ-glutamyl transpeptidase serum levels. This laboratory result is often misinterpreted as a marker of liver damage, when isolated elevation of this enzyme actually reflects enzyme induction. Similarly, hyperbilirubinaemia alone should not be equated with liver injury, since indirect hyperbilirubinaemia may be related to medications, such as indinavir or atazanavir[8,9,10]; this risk is increased in patients with underlying Gilbert's syndrome, a genetic disorder (Figure 1). On the other hand, drug-induced liver injury that is associated with an elevated direct bilirubin and clinical jaundice portends a poor clinical outcome. A cholestatic profile should only be considered when there is an associated increase in serum alkaline phosphatase as well as bilirubin.

Interplay between bilirubin metabolism, Gilbert's genotype and atazanavir (ATV) or indinavir (IDV) use. UDT, UDP glucuronosyltransferase.

Elevated aminotransferases also need to be interpreted within their clinical context. For example, increased liver enzymes in a patient with chronic HBV infection do not necessarily imply drug injury but may reflect HBV-related hepatic flares, which often occur during the natural course of the disease.

With the widespread use of HAART and the availability of new antiretroviral medications, ARLI has gained prominent attention owing to its negative impact on clinical outcomes. Drug-associated hepatotoxicity also creates an economic burden on already strained medical budgets, since additional visits and hospital admissions are often required for appropriate patient care and management.[1] Furthermore, antiretroviral drug discontinuation hampers maintenance of HIV suppression.

The severity of ARLI may range from the absence of symptoms to liver decompensation, and the outcome can range from spontaneous resolution to liver failure and death.[11,12] In one study, severe hepatotoxicity with acute hepatic necrosis was recognized in 2% of HIV-infected patients dying from liver disease. Furthermore, in a large ACTG cohort of nearly 3000 patients initiating HAART, the most common grade 4 adverse events were liver related; this risk was increased in patients with underlying chronic viral hepatitis.[13]

Fortunately, the vast majority of episodes of ARLI are asymptomatic, and most ALT elevations resolve spontaneously, as described for many other medications, probably through a process called 'adaptation'.[14] However, in a minority, drug-induced liver injury can be overt and have serious consequences. Therefore, it is critically important for the clinician to understand risk factors associated with poor outcomes and the pathogenic mechanisms of disease.


Comments on Medscape are moderated and should be professional in tone and on topic. You must declare any conflicts of interest related to your comments and responses. Please see our Commenting Guide for further information. We reserve the right to remove posts at our sole discretion.