Drug induced liver injury and its relationship to autoimmune hepatitis

Department of Medicine, University Medical Centre Hamburg-Eppendorf, Martinistrasse 52, 20246 Hamburg, GermanySee Article, pages 820–827Drug-induced liver injury (DILI) is a leading cause of acute liverfailure [1] and is of major importance to the pharmaceuticalindustry, as it is the most frequent reason for withdrawal of sub-stances from the market. More than 1000 different drugs and her-bal remedies have been described to cause DILI. Some drugscause a dose-dependent toxicity which can be anticipated – awell known example is acetaminophen. The vast majority of DILIcases cannot be foreseen and are, therefore, termed idiosyncratic.Oxidative stress, reactive metabolites, mitochondrial toxicity,modulation of drug metabolizing transporters, induction of apop-tosis or necrosis, as well as immunoallergic reactions to proteinadducts may all be involved in DILI, but for most of the drugsthe specific mechanisms contributing to DILI are unknown [2].A classification of hepatocellular, mixed or cholestatic liver injuryis used in the clinic to describe damage and severity of liverinjury but its value in predicting the outcome of DILI is unclear.DILI can range from asymptomatic elevation of liver enzymesto fulminant hepatic failure [1]. Its over-all mortality dependson the drug used, the timely detection of DILI including theappropriate action taken, as well as individual cofactors andcomorbidities such as the presence of diabetes mellitus [3]. Mor-tality rates of approximately 10% [1] have been reported but maybe highly overestimated since most mild drug reactions will notbe registered. In most patients, DILI leads to a complete recovery.However, up to 1% of all patients with DILI may develop liver cir-rhosis subsequently [4].DILI is a rare event with an estimated incidence of 1 per10,000 to 1 per 100,000 treated patients [5]. Some risk factorshave been identified: the daily amount of the drug causing thereaction (intake of >50 mg daily increases risk), a predominantlyhepatic metabolism of the drug (>50%) [6], and mitochondrialdysfunction that may be the underlying cause of the increasedrisk of DILI in diabetes patients [7]. Genetic polymorphismswithin genes relevant to drug metabolism [8] are being identifiedand may be used to stratify the risk of DILI to certain drugs in thefuture.A question relevant to daily clinical practice is, if re-exposureto the same or to other drugs may cause a second DILI episode.Re-exposure to the same drug generally should be avoided dueto the considerable mortality rate reported after re-exposure todrugs such as halothane (up to 50% due to a combination of mito-chondrial and immune-mediated injury) [9]. Nonetheless, fordrugs that are urgently needed in case of life threatening disease,such as in tuberculosis, careful re-exposure can be undertakenafter the resolution of the first DILI episode with acceptable risk[10]. The risk of re-exposure to a different drug that may belongto the same class of drugs or be completely unrelated is not wellstudied. To this end, the study presented by Lucena et al. in thisissue adds valuable and reassuring information[11]. In this study,the Spanish DILI registry of patients with probable or highlyprobable idiosyncratic DILI was searched for patients who suf-fered from a second DILI episode. Only a small number of patients(9 patients, 1.21%) from the cohort of patients with prior DILIdeveloped a second episode of liver injury and none of thepatients’ required liver transplantation or had a fatal outcome.Of particular interest, the majority of patients (8 out of 9) expe-riencing a second DILI episode had a similar (hepatocellular)damage pattern. The drugs responsible for the second DILI epi-sode were structurally related to the drug of the first DILI episodein 4 out of 9 cases and, in 2 cases, the target of the drug was thesame. This may point toward an immune-mediated drug injurydirected against similar protein adducts forming haptens. Indeed,4 out of the 9 patients (44%) were diagnosed with so called ‘‘auto-immune (AIH) DILI’’ since these patients fulfilled the criteria forprobable or definite AIH, according to the revised and simplifiedscore of the international autoimmune hepatitis study group[12,13]. This number was considerably higher than in patientswith a first DILI episode, where only 12/733 (1.6%) patients werediagnosed with ‘‘AIH DILI’’ [11], suggesting that immune-medi-ated reactions become the main mechanisms leading to DILI afterre-exposure to different drugs.There is no consensus on the nomenclature of immune-med-iated DILI and we will propose some definitions and respectivediagnoses at the end of this article. Until then, the diagnosesgiven in the publications cited will be used.In clinical practice, the differentiation between immune-med-iated DILI and AIH may be challenging. There are no pathogno-monic features of AIH and the diagnosis is made according to aclinical, biochemical, serological, and histological pattern whichhas to be confirmedby the response to immunosuppressive treat-ment [14]. This difficulty in differentiating between AIH and drugJournal of Hepatology 2011 vol. 55