Autoimmune-like drug-induced liver injury: a review and update for the clinician

Jonathan G. Stine, Patrick G. Northup

Research output: Contribution to journalReview articlepeer-review

21 Scopus citations


Introduction: Autoimmune-like drug-induced liver injury (DI-AIH) is a rare but serious event with a growing body of scientific evidence and a fair degree of uncertainty. Areas covered: This review covers the definition, pathophysiology, treatment and patient-centered outcomes of DI-AIH and presents up-to-date information on the most commonly implicated drugs. Expert opinion: A high degree of clinical suspicion is required for the diagnosis of DI-AIH. This diagnosis should be considered in any patient with either acute or chronic elevations in liver-associated enzymes. Prevalence rates exceed 15% based on large international registry data. Autoantibodies, while common, are neither specific nor diagnostic of DI-AIH. Histology may be helpful in describing subtle differences between DI-AIH and de novo idiopathic autoimmune hepatitis (iAIH), but oftentimes the two are indistinguishable histologically. Alpha-methyldopa, fibrates, hydralazine, minocycline, nitrofurantoin, HMG-CoA reductase inhibitors (statins), iplimumab and tumor necrosis factor alpha antagonists are the most commonly associated drugs with DI-AIH. Complete recovery of liver injury is most often seen with DI-AIH, however, cases of prolonged injury may occur and may require treatment with immunosuppressive therapy. Relapse following cessation of corticosteroids for suspected DI-AIH should prompt reconsideration of the diagnosis and further exploration into possible iAIH.

Original languageEnglish (US)
Pages (from-to)1291-1301
Number of pages11
JournalExpert Opinion on Drug Metabolism and Toxicology
Issue number11
StatePublished - Nov 1 2016

All Science Journal Classification (ASJC) codes

  • Toxicology
  • Pharmacology


Dive into the research topics of 'Autoimmune-like drug-induced liver injury: a review and update for the clinician'. Together they form a unique fingerprint.

Cite this