Chronic Liver Injury Induced by Drugs

A Systematic Review

Jonathan G. Stine; Naga Chalasani


Liver International. 2015;35(11):2343-2353. 

In This Article

Abstract and Introduction


To examine the available literature and summarize what is known about chronic drug-induced liver injury. We reviewed PubMed/MEDLINE through March 2015. We developed a MEDLINE search strategy using PubMed medical subject heading terms chronic liver injury, hepatotoxicity, drug-induced liver injury, cirrhosis and chronic liver disease. We reviewed the reference list of included articles to identify articles missed in the database search. Chronic liver injury from drugs is more common than once thought with prevalence as high as 18% based on large national registries. Patients with cholestatic injury, age ≤65 years, and a long latency period (>365 days) are at increased risk. Of the most common drugs associated with drug-induced liver injury, antibiotics (amoxicillin-clavulanic acid, trimethoprim-sulfamethoxazole, azithromycin) are most likely to cause chronic injury. The presence of autoantibodies is common with chronic DILI, however, it is not diagnostic nor is it specific to autoimmune-like drug-induced liver injury. Immunosuppressive therapy may be necessary for individual cases of autoimmune-like drug-induced liver injury where cessation of the drug alone does not result in resolution of injury, however, the lowest dose should be used for the shortest duration with careful attention to the development of side effects. The effectiveness of treament of cholestatic liver injury with corticosteroids or ursodiol remains unclear. Cases of drug-induced fatty liver, nodular regenerative hyperplasia and peliosis hepatitis are less common subtypes of chronic drug-induced liver injury that deserve special consideration. A high degree of clinical suspicion is required for the diagnosis of chronic drug-induced liver injury and should be suspected in any patient with liver associated enzyme abnormalities that persist out past 6 months of initial presentation. Treatment with drug removal and/or immunosuppressive therapy appears to be effective for the majority of cases. More study into pharmacogenomics and personalized medicine may aid in predicting which patients will go on to develop chronic drug-induced liver injury.


Drug-induced liver injury (DILI) can produce all forms of acute and chronic liver disease and while the majority of DILI episodes are self-limited with resolution following cessation of the offending agent,[1–5] a significant number of patients will progress to chronic DILI, which has been defined by DILIN network as continued injury 6 months after the initial diagnosis.[6] This is characterized by continued elevated in alanine aminotransferase (ALT), often >2× the upper limit of normal (ULN).[6] Despite the consenus opinion of the DILIN network, there exists some controversy of the true definition of chronic DILI as other large registry networks have provided different definitions. Andrade et al.[7] and the Spanish registry have defined chronic DILI to be persistent laboratory abnormalities in liver associated enzymes in a hepatocellular pattern of damage more than 3 months after drug withdrawal or more than 6 months after a cholestatic/mixed injury. An international DILI Expert Working Group from the United Kingdom also agreed on a 3 month cutoff for continued injury attributable to chronic DILI.[8]

Regardless of the differing opinions on the time criteria for chronic injury, chronic DILI may present as autoimmune-like DILI (AI-DILI) with or without autoantibodies,[9,10] vanishing bile duct syndrome (VBDS),[11–14] hepatic steatosis,[15] nodular regenerative hyperplasia (NRH), peliosis hepatitis[16] or even cirrhosis which typically occurs due to long-term prolonged intrahepatic cholestasis and bile duct injury.[12,17,18] A recent update from the Drug-Induced Liver Injury Network (DILIN) on their prospective cohort study of 899 patients found that 18% of patients had evidence of chronic DILI.[6] Patients with cholestatic liver injury, those aged ≤65 years, and a long latency period (>365 days) were at increased risk of progressing to chronic DILI.[6] A number of drugs have been implicated in the development of chronic DILI, including various antimicrobials, cardiovascular agents, central nervous system drugs, antineoplastics and analgesics and may occur in the setting of cutaneous manifestations including Stevens-Johnson Syndrome.[6] Of the top ten agents implicated by the DILIN network, the top three medications leading to chronic DILI (rate >20%) were amoxicillin-clavulanic acid, trimethroprim-sulfamethaxazole (TMP-SMZ) and azithromycin, the latter of which was unexpectedly more common in patients with pre-existing chronic liver disease.[6] The presence of autoantibodies is frequent with both acute and chronic DILI but is not specific to AI-DILI.[19] Antibody clearance typically lags behind resolution of chronic liver injury following cessation of the causative medication.[20] Cases of drug-induced fatty liver, nodular regenerative hyperplasia and peliosis hepatitis are less common subtypes of chronic DILI that deserve special consideration.

In general, DILI remains a challenge to the clinician as a diagnosis of exclusion based on both a history of known medication exposure and a thorough evaluation for alternative etiologies. While much is published about acute liver failure across several international registries,[1,2,4,21,22] less is known about chronic DILI and the diagnosis can be confounded by underlying chronic liver disease, including non-alcoholic fatty liver disease and chronic hepatitis C, the latter of which may confound cases of chronic DILI reported before the late 1980s. DI-AIH can be extremely difficult to distinguish from idiopathic AIH as Bjornnsson et al.[9] demonstrated a similar prevalence of autoantibodies, and corticosteroid responsiveness when comparing the two groups of patients. However, the presence of cirrhosis is much more common in denovo AIH as is a relapsing course or flare of liver associated enzymes with corticosteroid withdrawal.[9] Histology can also be useful in differentiating these two similar phenotypes as patients with DI-AIH are more likely to have portal inflammation and intracellular cholestasis and the presence of these in addition to fibrosis and plasma cell infiltrate is highly sensitive and specific.[10]

To date there has been no published systematic review on this topic. Thus, our objective was to critically evaluate and summarize the available literature for chronic DILI in a systematic fashion.