Antidepressant drugs can cause drug-induced liver injury (DILI). The authors review clinical data relevant to antidepressant-induced liver injury and provide recommendations for clinical practice.
A PubMed search was conducted for publications from 1965 onward related to antidepressant-induced liver injury. The search terms were “liver injury,” “liver failure,” “DILI,” “hepatitis,” “hepatotoxicity,” “cholestasis,” and “aminotransferase,” cross-referenced with “antidepressant.”
Although data on antidepressant-induced liver injury are scarce, 0.5%−3% of patients treated with antidepressants may develop asymptomatic mild elevation of serum aminotransferase levels. All antidepressants can induce hepatotoxicity, especially in elderly patients and those with polypharmacy. Liver damage is in most cases idiosyncratic and unpredictable, and it is generally unrelated to drug dosage. The interval between treatment initiation and onset of liver injury is generally between several days and 6 months. Life-threatening antidepressant-induced liver injury has been described involving fulminant liver failure or death. The underlying lesions are often of the hepatocellular type and less frequently of the cholestatic and mixed types. The antidepressants associated with greater risks of hepatotoxicity are iproniazid, nefazodone, phenelzine, imipramine, amitriptyline, duloxetine, bupropion, trazodone, tianeptine, and agomelatine. The antidepressants that seem to have the least potential for hepatotoxicity are citalopram, escitalopram, paroxetine, and fluvoxamine. Cross-toxicity has been described, mainly for tricyclic and tetracyclic antidepressants.
Although an infrequent event, DILI from antidepressant drugs may be irreversible, and clinicians should be aware of it. Aminotransferase surveillance is the most useful tool for detecting DILI, and prompt discontinuation of the drug responsible is essential. The results of antidepressant liver toxicity in all phases of clinical trials should be available and published. Further research is needed before any new and rigorously founded recommendations can be made.