Drug Induced Liver Injury: A Clinical Perspective. Robert J. Fontana, MD

Transcription

1 Drug Induced Liver Injury: A Clinical Perspective Robert J. Fontana, MD Drug induced liver injury (DILI) is an uncommon cause of acute and chronic liver injury of increasing importance to patients, clinicians, and regulators. The incidence of DILI due to an individual agent is not well defined but studies suggest that the overall incidence of DILI may be as high as 10 to 15 cases per 100,000 patient years. Bona fide risk factors for DILI are also not well established, but ongoing multicenter registry studies such as the Drug Induced Liver Injury Network are attempting to identify the role of genetic, environmental, and immunological factors in DILI pathogenesis and outcomes. Acute hepatocellular injury (~50%) is more common than mixed or cholestatic liver injury but jaundiced DILI subjects with either type of liver injury have a ~ 10% risk of short-term mortality. Antibiotics are the most commonly implicated agents associated with DILI, but there are emerging reports of liver injury associated with the use of a multitude of herbal and dietary supplements. Despite their widespread use, the HMG-CoA reductase inhibitors or statins are an uncommon cause of idiosyncratic DILI. Furthermore, recent studies have shown that statins are actually safe and efficacious to use in hyperlipidemic patients with chronic liver disease. Acetaminophen hepatotoxicity remains a leading cause of severe acute liver injury. Limiting the amount of acetaminophen in prescription narcotic products may help reduce the incidence of future non-intentional overdoses but educating patients of the multitude of over the counter products that contain acetaminophen is also recommended. Epidemiology The epidemiology of DILI has been difficult to study for several reasons. First, DILI accounts for < 1% of patients presenting with acute hepatocellular or cholestatic liver injury, with viral hepatitis, alcohol, autoimmune, and pancreaticobiliary diseases being much more commonly encountered in clinical practice. In addition, a lack of standardized diagnostic criteria and confirmatory objective biomarkers has precluded accurate and reliable identification of DILI patients (Table 1). The time to onset and laboratory presentation of DILI due to a specific agent also can be quite variable. Furthermore, the severity of DILI can vary from asymptomatic laboratory abnormalities to symptomatic cholestatic reactions and rare instances of ALF. Finally, since the diagnosis of DILI requires a high index of suspicion and the stepwise elimination of competing causes of liver injury, a diagnosis of DILI is often delayed or missed altogether. Several groups have initiated prospective multicenter registry studies to recruit and enroll patients with bona fide DILI for further assessment (Table 2). From 2003 to 2006, DILIN enrolled 300 consecutive DILI patients and noted that 73% of the cases were attributed to a single medication, 9% to herbal or dietary supplements (HDS), and 18% to multiple suspect agents. The most commonly implicated agents in this study include antimicrobials, central nervous system agents, and immunomodulatory medications. Interestingly, both the Spanish and DILIN registries identified amoxicillin-clavulanate as the most commonly implicated individual agent (59 and 23 cases, respectively). Of the initial 300 patients enrolled in DILIN, asymptomatic laboratory abnormalities, overall mortality, and liver-related mortality were reported in 14%, 8%, and 3%, respectively. Herbal and dietary supplements In recent years, there has been an increase in the use of over-the-counter herbal and dietary supplements due to the public perception that these natural products are efficacious and safe to take in large quantities. Common indications to use these products include intentional weight-loss, musclebuilding, and enhancement of general well being. However, regulation of the marketing and 1

2 manufacturing of HDS products is currently limited, and there are increasing reports of hepatotoxicity associated with some of these agents. Since under-reporting of over-the counter product use is common, it is difficult to accurately determine the rates of adverse events associated with their use. However, supplements containing Hydroxycut, Herbalife, green tea, black cohash, and anabolic steroids have all been implicated with severe and life-threatening DILI. Green tea is a common ingredient in many herbal preparations. It is derived from the leaves of Camellia sinensis and is a purported weight loss agent and energy enhancer. Multiple cases of hepatitis have been observed in patients consuming green tea extract. In 2003, Exolise, a product containing green tea extract, was withdrawn from the market in France and Spain due to reports of hepatotoxicity. In most cases of green tea hepatotoxicity, acute hepatocellular injury develops within 3 months of initial use, and the majority of patients will recover with discontinuation but rare cases of ALF have been reported. Catechins are polyphenols in green tea extract that are thought to lead to cellular injury through mitochondrial disruption and generation of reactive oxygen species, but the actual mechanism of liver injury remains unclear. Risk factors for DILI Many experts believe that certain patients may have increased susceptibility to DILI. While early retrospective studies suggested that women were at higher risk of developing DILI possibly due to reduced body weight and greater overall use of medications, subsequent studies have shown no difference between men and women. Nonetheless, two recent studies showed that women are more likely to present with hepatocellular liver injury, and that women with DILI may be at higher risk of progressing to ALF. The elderly (i.e. those over the age of 65) may be at increased risk for developing DILI, due to altered drug pharmacokinetics and inadvertent drug interactions associated with polypharmacy. However, there is no clear and convincing evidence from prospective studies that advancing age confers a higher susceptibility to DILI per se, although the elderly may be at higher risk of developing DILI with certain medications such as isoniazid. Elderly patients may also be more likely to present with a mixed or purely cholestatic liver injury pattern compared to younger patients. However, it should also be noted that children may be at increased risk of developing valproate hepatotoxicity as well as Reye s syndrome from aspirin. Underlying liver disease has long been considered a risk factor for the development of DILI. Both chronic hepatitis B virus (HBV) and hepatitis C virus (HCV) infection may predispose patients to developing DILI from several medications including anti-tuberculosis agents, methimazole, and ibuprofen. Although an attenuated intra-hepatic immune response and altered drug metabolism have been implicated, it can be very difficult to distinguish a chronic liver disease flare due to a surge in viral replication versus a true DILI episode. Nonetheless, multiple studies have shown that HBV or HCV coinfection appears to increase the risk of DILI in patients with human immunodeficiency virus (HIV) infection. In fact, treatment of HCV with interferon-based therapy may lower the risk of DILI due to highly active antiretroviral therapy (HAART) in HIV patients with HCV co-infection. In contrast to the above, most studies have not found non-alcoholic fatty liver disease (NAFLD) to be an independent risk factor for DILI. Furthermore, patients with NAFLD do not have an increased susceptibility to statin-induced DILI. Alcohol consumption was previously considered to be a risk factor for DILI. However, despite its presumed role in exacerbating acetaminophen-induced hepatotoxicity, alcohol has not been shown to be a risk factor for developing idiosyncratic DILI in several studies. Pathogenesis 2

3 The overwhelming majority of DILI cases are thought to be idiosyncratic (i.e. independent of drug dose and due to a unique mixture of one s own characteristics). However, a recent review of drugs that were withdrawn from the marketplace or received black-box warnings for hepatotoxicity showed that most were prescribed at doses exceeding 50 mg per day. Recent studies have also suggested that some DILI cases may, in part, be immune-mediated. For example, hepatotoxicity due to amoxicillin-clavulanate, ximelagatran, and lumiracoxcib were recently associated with specific class I and class II human leukocyte antigen (HLA) genotypes. The delayed onset of DILI in many cases further supports the notion that the adaptive immune response may be involved in DILI pathogenesis. However, most other idiosyncratic DILI cases have not been shown to correlate with HLA genotypes. Autoimmunity due to an altered host protein or drug-metabolite/host protein complex is another potential mechanism of DILI. For example, several medications such as hydralazine, diclofenac, nitrofurantoin, and minocycline, are associated with both hepatotoxicity and autoimmune syndromes. Although many of these DILI patients present with high titer autoantibodies, the fundamental pathogenesis of sporadic autoimmune hepatitis unfortunately remains illusive. A third potential mechanism of hepatotoxicity involves polymorphisms in metabolic and/or detoxification pathways leading to the development of toxic metabolites. However, it has been difficult to reliably identify reactive metabolites in humans as well as in in vitro and animal model test systems. Clarifying the role of host genetic factors in DILI pathogenesis has proven difficult due to the variability of presentation, the large number of different drugs causing injury, and until recently, a lack of reliable prospective data and samples to test. Previous genetic association studies have largely focused on evaluating candidate genes involved in the uptake, metabolism, transport, or detoxification of a drug.the recent advent of genome-wide association (GWA) methodologies has enhanced the ability of researchers to identify the genetic basis of several common and rare diseases. GWA studies involve comparing common single nucleotide polymorphisms (SNP) present in at least 1 to 5% of individuals across the entire genome of case patients to unselected population controls of the same ethnic background. A DNA sample extracted from whole blood is plated onto a gene chip which can assess up to 1 million SNP s simultaneously using high throughput methods. Based on the frequency of identifying associations by chance, GWA studies usually require a strict threshold for statistical significance that exceeds 1 x 10-6 or more. The first successful GWA study in DILI identified a strong association between flucloxacillin-induced liver injury and the HLA-B*5701 allele. In this study, 51 case patients who experienced a typical cholestatic hepatitis following the use of flucloxacillin in England were genotyped as well as 64 drug-exposed controls. In the initial cohort as well as a validation cohort, possession of the HLA-B*5701 allele was associated with an 80-fold increased risk of developing DILI (p=9 x ). Interestingly, this HLA allele has also been associated with abacavir-hypersensitivity in HIV patients. A subsequent analysis also revealed a second gene, ST6GAL1, to be associated with flucloxacillin hepatotoxicity and a follow-up study suggested an independent role for polymorphisms in the promoter region of the pregnane X receptor (PXR) which is activated by flucloxacillin. Another recent GWA study identified an association between the HLA class II allele, HLA-DRB1*07, and elevations in serum ALT levels in 74 subjects treated with ximelagatran, a thrombin inhibitor that was withdrawn, and 130 treated controls. Finally, a GWA study was undertaken in 41 patients with lumiracoxcib induced liver injury and in a validation cohort of 98 patients. Fine mapping demonstrated a strong link between lumiracoxcib hepatotoxicity and a commonly occurring HLA haplotype, HLA-DRB1*150, with an odds ratio of 5. Natural history and treatment of DILI 3

4 After discontinuation of the suspected drug, the majority of patients with DILI will achieve complete clinical and laboratory recovery. However, nearly 10% of patients with drug-induced jaundice will eventually progress to transplant or death. Patients who progress to ALF with concomitant coagulopathy and encephalopathy are at high risk of death, but independent predictors of poor outcome in DILI-induced ALF are yet to be elucidated. While overall survival rates are similar between patients with ALF due to acetaminophen overdose and idiosyncratic DILI, a greater percentage of patients with idiosyncratic DILI require transplantation for long-term survival (53% vs. 6%). Information regarding medication factors in the outcomes of patients with DILI is sparse. In general, a longer duration of therapy with an offending medication is associated with increasing severity of liver injury as well as an increased risk of developing chronic DILI. In addition, specific drugs such as halothane may be associated with a higher case fatality rate in comparison to other agents such as erythromycin and amoxicillin-clavulanate which are generally associated with self-limited disease. Patients with underlying diabetes mellitus are more likely to develop severe DILI and perhaps even ALF. The role of chronic alcohol consumption in DILI remains largely unclear and has not been associated with development of DILI or evolution into chronic liver injury. A past history of DILI due to any medication increases the risk of future DILI. However, the role of other factors such as diet, smoking, and underlying illnesses in the development and outcome of DILI are largely unknown. In regards to clinical presentation, early studies suggested that hepatocellular liver injury may portend a poorer prognosis, although recent prospective data do not support this observation. In patients with hepatocellular injury, the ratio of aspartate aminotransferase (AST) to ALT may correlate with progression to ALF, transplantation, and death. The combination of hepatocellular injury and jaundice has been associated with higher mortality (i.e. Hy s rule ), and several recent studies have confirmed higher rates of death and liver transplantation in patients with this clinical presentation. The vast majority of patients with DILI will achieve complete recovery after cessation of the suspect medication. However, approximately 5 to 14% of patients may develop chronic DILI which has been variably defined in different studies. DILIN defines chronic DILI as persistently elevated liver biochemistries on at least 2 occasions, abnormal liver histology, or radiological evidence of persistent liver injury at 6 months or more after DILI onset. Patients who present with a cholestatic injury pattern may be at increased risk of developing chronic DILI but confirmatory prospective studies are needed. To date, neither corticosteroids nor ursodeoxycholic acid have been convincingly shown to improve prognosis or outcomes or reduce the likelihood of developing chronic DILI. N-acetylcysteine (NAC) is an approved treatment for patients with acetaminophen overdose at risk for liver injury. In 2009, the ALFSG reported the results of a randomized, double-blind, placebo-controlled trial which demonstrated the benefit of a 72-hour infusion of intravenous NAC in patients with nonacetaminophen ALF. Specifically, transplant-free survival was significantly higher for patients receiving NAC as compared to placebo (40% vs. 27% respectively, p=0.043) and this benefit was primarily confined to patients with early stage ALF, defined as coma grade I-II. Interestingly, patients with idiosyncratic DILI appeared to experience the greatest benefit in transplant-free survival (58% vs. 27%). Therefore, additional studies of NAC or other anti-oxidants/hepatoprotective agents are needed in patients with severe DILI. However, it should be noted that a recent study of IV NAC in children with ALF of varying etiologies demonstrated no survival benefit. 4

5 Table 1. Minimal elements required to make a diagnosis of DILI Feature Presenting clinical features - Age - Gender - Indication for drug - Fever, rash, itching Medication history - Implicated drug or HDS with generic name, dose, and frequency of administration - Duration of therapy - Other drugs taken in 2 months prior to DILI onset - Prior implicated drug exposure Initial evaluation - Time from drug start to symptoms or jaundice (if present) - Time from drug start to first abnormal lab - Eosinophils (number/µl or %) at onset or early in course of injury - Initial and serial AST and ALT levels - Initial and serial alk phos or GGT levels - Initial and serial bilirubin levels - Initial and serial INR Potential risk factors/ competing causes - Alcohol use - Risk factors for acute and chronic liver disease including family history - Exclusion of heart failure, shock, sepsis, parenteral nutrition prior to DILI onset - Blood tests to exclude other causes of acute liver disease - IgM anti-hav (or negative anti-hav) - HBsAg and IgM anti-hbc (or negative anti-hbc) - Anti-HCV and HCV RNA - ANA, SmAb (and titer if positive) - Imaging of the liver (type and results) - Liver USN and abdominal CT/ MRI - Liver histology Comments - Immunoallergic features frequently absent - Difficult when taking multiple herbals - Difficult to determine in some cases - Rechallenge infrequently encountered and not always reliable - Some patients are asymptomatic - First lab value meeting diagnostic criteria is DILI onset date - Eosinophilia defined as Absolute Eos > 500/ml or > 5% on diff - Review available pretreatment, peak, and post-discontinuation lab values - Include direct bilirubin if available. - Amount and pattern of use in 2 months prior to drug start and thereafter - Include parenteral exposures - Based on medical history and cardiac imaging - Consider anti-hev IgM if possible - Quantitative immunoglobulins, anti-hiv, CPK, CMV DNA by PCR, and EBV-DNA by PCR of value in selected cases - MRCP and ERCP helpful in some - Useful to review injury pattern and severity if available Adapted from Fontana RJ, et al Hepatology August

6 Table 2. Baseline features and outcomes in selected cohorts of DILI patients Chalasani et al 13 Andrade et al 11 De Valle et al 19 Study type Prospective Prospective Retrospective Country U.S. Spain Sweden DILI patients (n) Mean age (yrs) % Female 60% 49% 56% Ethnicity (%) Caucasian African-American Other 79% 11% 10% 100% 100% 57% / 20% / 23% 58% / 22% / 20% 48% / 12% / 40% % Hepatocellular/ mixed/ cholestatic % Jaundice 69% 71% 38% % Liver biopsy 50% 25% 13% Most frequently implicated agents, (N) amoxicillinclavulanate (23) nitrofurantoin (13) Isoniazid (13) TMP/SMX (9) amoxicillinclavulanate (59) anti-tb meds (31) ebrotidine (22) ibuprofen (18) % Herbal/ dietary 11% 2% 5% supplements % Hospitalized 60% 53% 51% % Death/ transplant 8% / 2% 5% / 2% N/a % Chronic DILI * 14% 10% N/a diclofenac (14) flucloaxcillin (8) azathioprine (5) atorvastatin (4) * Definition varies by study The content of this handout was largely adapted from Rangnekar A, Fontana RJ. An Update on Drug Induced Liver Injury. Minerva Gastroenterologica 2011; 52: Other general DILI References Seminars in Liver Disease 29(4); This entire issue is devoted to reviews on various aspects of druginduced liver injury. Agarwal VK, McHutchison JG, Hoofnagle JH. Important elements for the diagnosis of drug-induced liver injury. Clin Gastroenterol Hepatol May;8(5):

7 Fontana RJ, Seeff LB, Andrade RJ, Bjornsson E, Day CP, Serrano J, et al. Standardization of nomenclature and causality assessment in drug-induced liver injury: summary of a clinical research workshop. Hepatology Aug;52(2): Rockey DC, Seeff LB, Rochon J, Freston J, Chalasani N, Bonacini M, et al. Causality assessment in druginduced liver injury using a structured expert opinion process: comparison to the Roussel-Uclaf causality assessment method. Hepatology Jun;51(6): Senior JR. Drug hepatotoxicity from a regulatory perspective. Clin Liver Dis Aug;11(3): Textbooks on Drug-Induced Liver Injury Zimmerman HJ. Hepatotoxicity: the adverse effects of drugs and other chemicals on the liver. 2nd ed. Philadelphia: Lippincott, Williams & Wilkins; The classic text of drug-induced liver disease. Kaplowitz N, DeLeve LD. Drug-Induced Liver Disease. 2nd ed. New York: Informa Healthcare; Multi-author text that is the intellectual successor to Dr. Zimmerman s classic work. 7