Drug-Induced Liver Injury: Highlights from a Review of the 2015 Literature

Philip Sarges, Joshua M Steinberg, James H Lewis

Numerous publications contributed to the expanding knowledge base about drug-induced liver injury (DILI) in 2015. New findings from the US Drug Induced Liver Injury Network (DILIN) in their most recently updated registry include a 1- to 3-week delay in the appearance of acute DILI from short-course antibiotics such as cefazolin. They corroborated the finding that acute DILI in patients with underlying liver disease was far more severe and potentially fatal than in patients without liver disease. The only drug that seemed to have an increased risk of hepatotoxicity in these patients was azithromycin. While nearly one in six patients with acute DILI had persistently elevated liver tests at 6 months, and results for 75 % of these patients continued to be abnormal at 12 months, most of these "chronic" injury cases were relatively minor and the result of cholestatic hepatotoxins. Newly described DILI agents include tolvaptan, as well as some new direct-acting antiviral protease inhibitors for chronic hepatitis C. The latter have been associated with serious acute hepatitis, hyperbilirubinemia, and decompensation. Herbal hepatotoxicity continues to be increasingly reported, although applying causality assessment to these cases can, in fact, be more challenging than with prescription drugs. As important as cases with DILI, the class of PCSK9 inhibitors used to lower low-density lipoprotein (LDL) cholesterol have not been associated with significant liver injury, in contrast with other lipid-lowering agents. With respect to pharmacologic DILI risk factors, new data show that drugs metabolized by cytochrome P450 enzymes had a nearly four times higher likelihood of causing DILI. Interestingly, high lipophilicity, which was previously felt to be a risk factor for DILI, was not found to be associated, although more study is needed to confirm this observation. While human leukocyte antigen (HLA) genotypes have been linked to several specific agents, the role of such testing in the general population remains undefined due to the currently low positive and negative predictive values of the available tests. New DILI biomarkers, specifically microRNA-122 and keratin-18, among others, appear to have the necessary predictive value to determine the prognosis and outcome of patients with paracetamol (acetaminophen [AAP])-induced acute liver failure (ALF), and may be of great benefit in deciding who requires N-acetylcysteine (NAC), and for what duration. Treatment options for other forms of DILI remain limited; no firm conclusions can currently be drawn for the use of NAC in non-AAP ALF.