Human & Experimental Toxicology http://het.sagepub.com/
Female gender as a susceptibility factor for drug-induced liver injury David E. Amacher Hum Exp Toxicol published online 3 December 2013 DOI: 10.1177/0960327113512860 The online version of this article can be found at: http://het.sagepub.com/content/early/2013/12/02/0960327113512860
Published by: http://www.sagepublications.com
Additional services and information for Human & Experimental Toxicology can be found at: Email Alerts: http://het.sagepub.com/cgi/alerts Subscriptions: http://het.sagepub.com/subscriptions Reprints: http://www.sagepub.com/journalsReprints.nav Permissions: http://www.sagepub.com/journalsPermissions.nav
>> OnlineFirst Version of Record - Dec 3, 2013 What is This?
Downloaded from het.sagepub.com at UNIVERSITE LAVAL on July 6, 2014
Article
Female gender as a susceptibility factor for drug-induced liver injury
Human and Experimental Toxicology 201X, Vol XX(X) 1–12 ª The Author(s) 2013 Reprints and permission: sagepub.co.uk/journalsPermissions.nav DOI: 10.1177/0960327113512860 het.sagepub.com
David E. Amacher
Abstract Adverse drug reactions (ADRs) can involve all tissues and organs, but liver injuries are considered among the most serious. A number of prospective, multicenter studies have confirmed a higher risk of ADRs in general among female subjects compared to a male cohort. Although drug-induced liver injury (DILI) is infrequently encountered, the preponderance of evidence suggests that women appear to be more susceptible than men to fulminate hepatic/acute liver failure especially in response to some anti-infective drugs and to autoimmunelike hepatitis following exposure to certain other therapeutic drugs. A number of hypotheses have been proposed to explain this sex difference in susceptibility to DILI. Collectively, these hypotheses suggest three basic sex-dependent mechanisms that include differences in various aspects of drug pharmacokinetics (PK) or pharmacodynamics following the administration of certain drugs; specific hormonal effects or interactions with immunomodulating agents or signaling molecules; and differences in the adverse response of the immune system to some drugs, reactive drug metabolites, or drug-protein adducts. At the preclinical drug safety stage, there is a need for more research on hormonal effects on drug PK and for additional research on gender differences in aberrant immune responses that may lead to idiosyncratic DILI in some female patients. Because the detection of rare but serious hepatic ADRs requires the exposure of very large patient populations, pharmacovigilance networks will continue to play a key role in the postmarketing surveillance for their detection and reporting. Keywords Drug-induced liver injury, female gender, adverse drug response
Introduction Adverse drug reactions (ADRs) are harmful and unintended reactions to medicines that occur at doses normally used for treatment. One type of ADR is caused by an extension or exaggeration of the intended pharmacological action of the drug, and a second type results from toxic effects unrelated to the intended pharmacological action of the drug but dependent upon unique host factors. ADRs vary in severity, and the most serious ADRs are those that require hospitalization, are permanently disabling, or result in death. It has been estimated that ADRs are the fourth to sixth largest cause of mortality in hospitalized patients in the United States.1 ADRs can involve all tissues and organs, but liver injuries are considered among the most serious. A recent systemic review of 106,586 hospitalized patients has shown that approximately 5.3% of the hospital admissions were associated with ADRs.2 A prospective, multicenter study based on intensive pharmacovigilance involving 2,371 patients
for whom 25,532 drugs had been prescribed has confirmed a higher risk of ADR among female subjects than a male cohort.3 Additional general examples will be described in the section on Pharmacovigilance studies that demonstrate sex differences in events diagnosed as ADR and specific liver examples in the section on Pharmacovigilance studies that suggest gender influences the susceptibility to drug-induced liver injury. Overall, compared with male patients, female patients have a 1.5- to 1.7-fold greater risk of developing an ADR.4 Results from a survey conducted between January 1997 and January 2001 involving 10 prescription drugs that had been withdrawn from the US market indicated that 8 posed
Toxadvisor Toxicology Consulting, USA Corresponding author: David E. Amacher, Toxadvisor Toxicology Consulting, P.O. Box 254, Hadlyme, Connecticut, CT 06439, USA.
Downloaded from het.sagepub.com at UNIVERSITE LAVAL on July 6, 2014
2
Human and Experimental Toxicology XX(X)
greater health risks for women than men, and 2 of these resulted in liver failure.5 Four of these drugs may have led to more adverse events in women because they were prescribed more frequently to women than men. The remaining four were associated with more adverse events in women even though they were widely prescribed to both women and men. The term ‘‘gender difference’’ has been used broadly to include differences due to sex, other biological processes, and nonbiological influences.6 The term ‘‘sex differences’’ has been defined as the purely biologic or physiologic differences of females and males.7 In this review, these terms will be used interchangeably to refer to biological processes leading to adverse drug reactions in adult men and women. The purpose of this article is to review the current status of published studies on sex/gender differences in druginduced liver injury (DILI) responses to pharmaceutical agents. The topics include a summarization of general biological differences that can affect drug efficacy and toxicity, the status of clinical testing of new chemical entities in women, established gender differences in drug pharmacokinetics, pharmacovigilance studies that demonstrate sex differences in the diagnosis of ADR, an overview of the etiology of DILI, and a review of recent, specific pharmacovigilance studies that suggest links between gender influences and DILI.
effects on drug metabolism.9 The generally greater fat stores in women may account for larger volumes of distribution (Vd) for lipophilic drugs.10 Differences in the innate and adaptive immune systems between men and women have also been documented,11 and these can result in immune-mediated hepatotoxicity. Other suggested hypotheses include inadvertent drug overdosing on a milligram per kilogram basis, because women generally have a lower body weight than men, that women are more likely to report adverse events than men, and that women take more medications than men and therefore are more likely to experience adverse effects due to drug–drug interactions.6 There also are some differences in the medications prescribed for men and women. For example, women are more often prescribed psychotropic medications and drugs for the cardiovascular system, alimentary tract, and the musculoskeletal system than men.12,13 There can be differences in health issues, for example, there is a generally higher incidence of urinary tract infections in women than in men. Excluding behavioral or dosing differences, these hypotheses collectively suggest three basic mechanisms that include: (1) gender-dependent differences in pharmacokinetics following the administration of certain drugs, (2) gender-specific hormonal effects or interaction with signaling molecules that can affect drug safety, and (3) differences in an aberrant immune response that targets the liver following drug exposure that can result in adverse drug reactions.
General biological differences between men and women that can affect drug The clinical testing of new molecular efficacy and toxicity The current literature suggests a number of possible entities in women susceptibility mechanisms or factors that might be responsible for any differences in the observed incidence of adverse effects between men and women following exposure to therapeutic levels of drugs. Gender-based variations that can influence drug pharmacokinetics and subsequent toxicity include differences in gastric acid secretion, gastrointestinal blood flow, the relative amount of drug-binding proteins, the relative proportions of muscular and adipose tissue, renal blood flow, gender-specific expression of cytochrome P450 (CYP450) isozymes, and physiologic and hormonal changes during the menstrual cycle.8 Lower gastric acid secretion, lower gastric emptying time, and intestinal transit rate in women can affect the absorption and subsequent bioavailability of orally administered drugs. Menstruation, pregnancy, the use of oral contraceptives, and menopause all can have profound
Pre-approval clinical drug development through phases I–III can involve as few as 500 subjects and rarely more than 5,000 subjects.14 Historically, females of all ages have been underrepresented in these clinical trials, and a gender-specific analysis of the data are usually not included in the evaluation of the results. For a number of years, the US Food and Drug Administration (FDA) guidelines specifically excluded participation of females in many clinical drug studies due to the potential risks involved during pregnancy in women of child-bearing age and due to co-existing illnesses in elderly women.15 This followed the belated discovery of the human teratogens thalidomide in 1961 and diethylstilbestrol in 1971.16 In 1993, the FDA published the Guideline for the Study and evaluation of Gender Differences in the Clinical Evaluation of Drugs which lifted this
Downloaded from het.sagepub.com at UNIVERSITE LAVAL on July 6, 2014
Amacher
3
restriction for most women of childbearing potential and encouraged sponsors to collect gender-related data.17 A recent review of clinical trial data submitted to the FDA for New Molecular Entities for adult, nonsex specific indications between January 2006 and December 2007 indicated women were still underrepresented; however the number of trials enrolling women and the number of women participating in phase I trials has increased since 2001.18 As amended in 2001, the National Institute of Health (NIH) Policy and Guidelines on The Inclusion of Women and Minorities as Subjects in Clinical Research now ensure that all NIH-funded clinical research will be carried out in a manner sufficient to elicit information about individuals of both sex/genders and diverse racial and ethnic groups, particularly in NIH-defined phase III clinical trials, in order to examine differential effects on such groups.19 It should be noted, however, that the detection of at least 1 ADR with an incidence of 1 in 10,000 exposed individuals would require the exposure of at least 30,000 people to the drug,20 which can only be accomplished in the phase IV postapproval period.
Gender differences in drug pharmacokinetics The pharmacological as well as any toxicological interaction of a drug with the body proceeds through the sequential steps of pharmacokinetics (PK) that include drug absorption, distribution, metabolism, and elimination (ADME); any of which can be affected by gender differences. Studies in murine liver models have identified more than 1,000 genes whose expression is sex dependent, and their gene products include many drug metabolizing enzymes (DMEs).21 Although no genderspecific DMEs have been identified in humans, the analysis of gene expression data in 427 liver samples by Yang et al. have identified 77 drug metabolizing or transporter genes that demonstrate differential expression based on sex.22 Any sex differences in the gene expression of DMEs and transporters can affect drug efficacy or result in ADR by altering ADME in organs such as the liver. Changes in pharmacodynamics can affect both the desired therapeutic effect of a drug as well as its adverse effect profile.23 In the liver, drugs metabolized by phase I metabolism (oxidation, reduction, and hydrolysis via CYP450s), phase II conjugative metabolism (glucuronidation, conjugation, glucuronyltransferases, methyltransferases, dehydrogenases) and by the combined oxidative and conjugation processes are usually cleared
faster in men compared to women on a milligram per kilogram basis.24 The human CYP superfamily consists of 57 functional genes and 58 pseudogenes.25 Of these, members of the CYP1, 2, and 3 families play key roles in the metabolism of therapeutic drugs. Some DME genes such as CYP3A4, CYP2B6, and GSTA1 display large and measurable inter-individual variability.26 CYP1A2 is a major CYP450 in human liver that metabolizes a variety of clinically important drugs and several endogenous compounds such as arachidonic acids and steroids.27 Microsomal CYP1A2 activity measured by 7-ethoxyresorufin Odealkylase activity has been reported to be greater in males than females, which is consistent with clinical observation.28 In addition to CYP1A2, men potentially have a higher activity relative to women for CYP2E1, for the drug efflux transporter Pglycoprotein (P-gp/MDR1), and for some isoforms of glucuronosyltransferases and sulfotransferases.29 Human CYP3A4 is the most abundant hepatic phase I enzyme that metabolizes approximately 50% of the marketed drugs.30 In cryopreserved primary human hepatocytes, Parkinson et al. found that the gender difference in CYP3A4 activity (females ¼ twice males) as measured by testosterone 6-betahydroxylase was statistically significant, which supports the clinical observation that females metabolize certain CYP3A4 substrates faster than males do,28 and therefore the total clearance of CYP3A substrates appears to be mildly or moderately faster on a milligram per kilogram basis in women compared to men.24 The majority of all published studies indicate that CYP3A4 activity is generally higher in women than in men which is remarkable considering the aforementioned extensive individual variability, whereas the activity of many other systems involved in drug metabolism may be higher in men than in women. Gender-based differences in the expression levels of hepatic or renal metabolic enzymes and transporters can result in longer drug half-life, as for example, with anti-cancer drugs in female patients. In an FDA review of 300 new drug applications between 1995 and 2000, for the 163 that included a sex analysis, 11 of those drugs demonstrated >40% difference in PK between males and females, and even though this was listed on the product label, no dosing recommendations were made based on sex.31 Physiological factors that may result in genderrelated PK differences include the generally lower bodyweight and organ size, a higher percentage of
Downloaded from het.sagepub.com at UNIVERSITE LAVAL on July 6, 2014
4
Human and Experimental Toxicology XX(X)
body fat, a lower glomerular filtration rate, and different gastric motility in women compared with men.29 Men and women also show different pharmacodynamic responses to a variety of drugs. In women, issues such as menopause, pregnancy, oral contraceptive use, and menstruation can also have profound effects on drug metabolism which can be clinically important. Pregnancy may increase the elimination of antiepileptic agents, reducing their efficacy, while oral contraceptive use can interfere with the metabolism of many coadministered drugs.9 Drug metabolism by CYP2C9, CYP2C19, and Nacetyltransferase, appear to be similar in men and women (mg/kg), and the clearance of P-gp substrates also appear to be similar in men and women. Some previous authors have concluded that sex differences in drug metabolism and elimination are in fact primarily related to steroid hormone levels.32 In women, drug metabolism can be affected by the aforementioned sex-specific factors: menopause, pregnancy, and menstruation. For example, recent studies have demonstrated that pregnancy, depending upon the stage, alters the activity of CYP2D6, CYP3A4 (increased), and CYP1A2 (decreased), and increases the activity of the renal transporters organic cation transporter and P-gp.33 Studies in rat and mouse models have demonstrated that growth hormone (GH) is the key hormonal factor that controls sex differences in the expression of a large number or hepatic gene products that include P450 isozymes and other DMEs, and human similarly display plasma GH patterns that are sex dependent and regulated by gonadal hormones.21 Based on clinical data, during adolescence, GH is known to regulate sex-specific DME expression, and it appears that sex hormones can significantly influence DME activity in an isoform-specific manner.34 Recent results from an in vitro system designed to elucidate biological mechanisms for human sex differences in susceptibility to DILI have suggested that interaction between sexhormones and pro-inflammatory cytokines might alter liver metabolism and thus contribute to marked sex differences in susceptibility to DILI.35
Pharmacovigilance studies that demonstrate sex differences in events diagnosed as ADR Pharmacovigilance is based on information gathered from individual case safety reports and other pharmacoepidemiological data. Notably, some large drug
surveillance studies in hospital patients have demonstrated significant differences between men and women for the risk of ADR-related hospitalization. A summary of some recently published studies follows. In a study of ADR as the cause for hospital admission that involved the prospective analysis of 12,249 hospitalized patients, older age, and female gender were associated with ADR-related hospitalizations.36 The prospective analysis of 907 patients in 2 German hospitals,37 an event monitoring of 4,331 hospitalized patients in Switzerland,38 an investigation of 513,608 general practice patients using data collected from 48 national cohort studies in England,39 a nation-wide study of ADR-related hospitalizations between 2000 and 2005 in the Netherlands,40 a prospective drug surveillance study of 1,920 patients hospitalized in Chile,41 and a prospective multicenter study of 2371 in-patients in Germany and Israel,13,42 all concluded there was a higher risk of ADRs among female subjects. More specifically, a study of hospital admission data from the Dutch National Register between 2000 and 2005 has indicated clear sex differences in ADRs requiring hospital admission for cardiovascular drug groups.43 Conversely, a study of ADRs recorded in the French MidiPyre´ne´es Pharmacovigilance Centre in 1998, there was no statistically significance in the incidence of ADRs reported in males versus females for the total population.44 But serious ADRs were more frequently reported in females between 20 and 29 years and in males 0-9 and 60-69 years. ADRs were more frequently reported in females for genitourinary, sex hormones, antineoplastic, respiratory, and antiparasitic drugs. The International Classification of Diseases (ICD) is a standard diagnostic tool used to monitor the incidence and prevalence of diseases and other health problems. The adverse reactions monitored in these studies generally depend upon the drug class and cover a wide range of clinical endpoints, often using the ICD coding system,45 but excluding such things as errors in administration, preexisting disease, and intended forms of overdose. Some examples of diagnoses codes associated with high prevalence of ADR include drug-induced neutropenia, hypoglycemia, hepatitis unspecified, anaphylactic shock due to drugs, shock due to anesthesia,46 and gastrointestinal bleeding.47 In addition to the general consensus that ADRs are more frequently reported in women, the severity of ADRs, particularly liver toxicity, generally appears to be greater in women than men.
Downloaded from het.sagepub.com at UNIVERSITE LAVAL on July 6, 2014
Amacher
5
Drug-induced liver injury DILI refers to a specific liver injury or toxicity caused by drugs or other xenobiotics, and comprises a specific and important category of ADR. The basic mechanisms proposed for DILI are based on the production of reactive metabolites, immune-mediated hepatotoxicity, and a ‘danger signal’ hypothesis and/or alterations in mitochondrial function.48 The danger hypothesis suggests that the primary determinant for the mounting of an immune response against a drug is whether the drug or more likely, a reactive metabolite, cause cellular damage.49 The danger signal molecules are thought to act through the same receptors as pathogen-associated molecules to subsequently stimulate the immune system. In what has been referred to as the pharmacological interaction with immune receptors (p-i) concept, some drugs may bind directly and reversibly to immune receptors which stimulate those cells.50 For the mitochondrial dysfunction model, a 3-step mechanistic working model of drug-induced hepatotoxicity has been described by Russmann et al.51 These steps include: (1) An initial injury that results from direct cellular stress, direct mitochondrial inhibition, and/ or specific immune reactions, (2) the initial injury can then lead to direct mitochondrial impairment via mitochondrial permeability transition (MPT), and (3) the MPT subsequently leads to either necrosis or apoptosis depending on the availability of ATP. The liver contains components of both the innate and adaptive immune system, and is the major organ for the drug metabolism. In the United States, the most common drugs leading to DILI are antibiotics, central nervous system agents, herbal/dietary supplements, and immunomodulatory agents.52 The associations for human liver and skin ADR are known to involve human leukocyte antigen (HLA) genes for drugs such as abacavir and carbamazepine.53 In addition to those two drugs, strong HLA associations indicating the involvement of the immune system have been demonstrated with a number of drugs that cause liver injury including flucloxacillin, lumiracoxib, lapatinib, and ximelagatran.54 Further, drug-induced immune-mediated hepatic injury also has been implicated for drugs such as halothane, tienilic acid, dihydralazine, chlorpromazine, erythromycins, amoxicillin–calvulanic acid, sulfonamides and sulindac, methyldopa, minocycline, and nitrofurantoin.55 The latter authors have delineated the formation autoantibodies directed toward the microsomal cytochromes that participate in the metabolism of some
of those drugs, although it is not known if they participate in the destruction of hepatocytes. Minocyclineassociated hepatotoxicity is more common in women. Fortunately, the majority of cases of DILI are acute illnesses that are quickly resolved after the offending medication is stopped. In a 10-year study of 1,100 cases reported to the Danish Committee on ADR, hepatic injuries accounted for 5.9% of all ADRs reported, and 14.7% of the lethal adverse drug reactions.56 Between January 2001 and June 2007, retrospective analysis of 170 patients who had been diagnosed with DILI indicated that a total of 83 different drugs were associated with the hepatotoxicity with a single drug suspected in approximately 58% of the individuals.57 The main causative group was antibiotics. Of the 62 patients hospitalized, acute liver failure (ALF) developed in 14 (8.2%), chronicity occurred in 19 (11.2%), and death occurred in7 (4.1%). Clinically, DILI is a diagnosis of exclusion of other causes and the identification of a signature pattern of disease manifestation coordinated with the initiation and discontinuation of drug treatment.58 An estimated annual incidence rate of DILI worldwide is 13.9–24.0/ 100,000 inhabitants.59 Nonetheless, the diagnostic approach to study DILI remains rudimentary and inaccurate due to the lack of reliable markers for use in general practice,60 and the incidence of DILI in the general population is thought to be largely underestimated.61 Several multicenter networks have been established world-wide for the assessment of DILI. In the United States, The Drug-Induced Liver Injury Network (DILIN) was established in 2003 to advance understanding and research into DILI as a collaborative effort among the NIH, five clinical sites, and a data coordinating center.62 DILI is estimated to occur in 1–2 cases per 100,000 person years and represents 1.2–6.6% of cases of acute liver disease reported by tertiary referral centers.63 In an extensive international study to identify causes of DILI at three major DILI registries, a total of 319 pharmaceuticals were identified as causes for liver injury.58 In descending order, drugs that were frequently identified in the combined list (greater than 30 cases reported) included amoxicillin–clavulanic acid, diclofenac, cotrimoxazole, isoniazid, and disulfiram. It should be noted that there is marked geographic variation in the identification of drugs linked to DILI,59,60 and a single drug is not clearly involved in all cases.57 There are also differences in prescribing certain drugs in men and women.40 In a screening of 6,383 hospitalized patients using CIOMS (Council for International Organizations of Medical Sciences) criteria to access the probability of DILI and to exclude any
Downloaded from het.sagepub.com at UNIVERSITE LAVAL on July 6, 2014
6
Human and Experimental Toxicology XX(X)
disease-related causes for liver injury, the incidence of DILI was 1.4%, and the drug classes most frequently causing DILI were heparins, antibacterials, tuberculostatics, and antineoplastic agents.64
Alcohol-induced liver injury In women and some female animals, most studies reveal a propensity toward greater alcohol-induced liver injury (ALI) due to female gender.65 Because gastric alcohol dehydrogenase is lower in females than males, women tend to have a lower alcohol toxic threshold than men.66 Moreover, it has been established clinically that women develop ALI more rapidly and to greater extent than men while consuming fewer grams of alcohol.67 This occurs even though a study of alcohol metabolism has demonstrated that the mean alcohol elimination rate and mean computed liver volume are not significantly different in men and women.68 In studies with rats fed an ethanolcontaining diet, Kono et al. have concluded that elevated endotoxin after ethanol administration triggers more Kupffer cell activation via enhanced CD14 expression in females. This leads to increases in tumor necrosis factor-a messenger RNA (mRNA) expression in the female liver resulting in more severe ALI.69 Because estrogen has major influence on the susceptibility of Kupffer cells to gut-derived lipopolysaccharide resulting in increased proinflammatory cytokine production, it has been suggested that this could be a major contributing factor to the increased risk of women to alcohol-induced liver disease.70 In summary, women develop ALI more rapidly and have a lower alcohol toxic threshold than men.
Acetaminophen overdose Administered at normal therapeutic doses, the analgesic acetaminophen (APAP, paracetamol) is safe, but it can be intrinsically hepatotoxic at high doses. During normal hepatic metabolism, a major portion of APAP is conjugated with either sulfate or glucuronic acid to form water-soluble metabolites that are readily excreted, and only small amounts of the reactive intermediate, N-acetyl-p-benzoquinon imine (NAPQI), are formed by the CYP450 enzymes71,72 which include CYP2E1, CYP1A2, and CYP3A4.73 Further studies have revealed that activation of APAP to this active metabolite is primarily carried out by CYP2E1.74 APAP clearance has been shown to be 22% greater in males than females due to the increased activity of the glucuronidation pathway in males,75 which occurs
mainly by UDP-glucuronosyltransferases (UGT). UGT2B15, which has recently been demonstrated to play a role in APAP glucuronidation,76 shows higher activity in males compared to females.77 When an adequate supply of cellular reduced glutathione (GSH) is available, most of the reactive intermediate NAPQI can be detoxified. But if the hepatic concentration of GSH becomes depleted, as for example at high APAP doses, subsequent covalent binding to the sulfhydryl (–SH) groups of various cellular proteins increases, which ultimately can result in hepatic necrosis.71 A single 10–15 g APAP dose, or occasionally less, can produce clinical evidence of liver injury while fatal fulminant disease is usually, although not always, associated with ingestion of 25 g.78 Overdoses of APAP result in the generation of APAP-protein adducts which have been shown to provide an excellent correlate with the severity of the APAP toxicity.79 The innate immune system is thought to play a role in the exacerbation and progression of APAP-induced liver injury.80 APAP liver toxicity can result from either accidental overdose or intentional overdose. Published findings provide strong support for the hypothesis that mitochondrial dysfunction and DNA damage are critical events in the mechanism of cell necrosis after APAP overdose in patients.81 In a study that reviewed 352 poisoning episodes involving APAP overdoses, the accidental overdose category constituted 16 (9%) of 182 hospital index visits with no apparent gender differences noted. Within the intentional overdose group, 1.2% developed ALF versus 25% of the accidental group. Young females with intentional overdoses accounted for most of the cases.82 This pattern of mortality has been noted in other studies. In a study of 663 patients, unintentional APAP overdose was associated with increased mortality compared with intentional overdoses, despite lower APAP concentrations.83 APAP is considered a model intrinsic or predictable hepatotoxicant. However, in recent years, drug candidates that cause intrinsic liver injury such as APAP are now efficiently identified and discontinued from further development in the preclinical testing state. The preponderance of reported adverse liver reactions is idiosyncratic in nature, usually occurring 5–90 days after the causative medication was last taken.84 Based on their clinical characteristics, idiosyncratic adverse drug reactions in general appear to be immune-mediated, and a substantial body of evidence further suggests that most although not all are caused by chemically reactive species.85 Thus, DILI
Downloaded from het.sagepub.com at UNIVERSITE LAVAL on July 6, 2014
Amacher
7
associated with recently marketed drugs is predominantly if not exclusively idiosyncratic in nature.
Acute liver failure ALF is characterized by coagulopathy and encephalopathy. In a multicenter, prospective study involving 1198 subjects enrolled at 23 sites in the United States, Reuben et al. concluded that DILI is an uncommon cause of ALF that evolves slowly and affects a disproportionate number of women and minorities.86 In another study, the incidence of ALF, which again occurred more frequently in women, was about 2000 cases per year with greater than 50% due to therapeutic drug use.6 In yet another study of 308 consecutive patients with ALF admitted to 17 tertiary care centers in the US, 73% were women.87 APAP overdose (39% of cases) and idiosyncratic drug reactions (13% of cases) were the most frequent apparent causes of ALF. In a multicenter, prospective US study of APAP-induced ALF over a 6-year period, the medium dose ingested was 24 g, which is equivalent to 48 extra-strength tablets.88 In a study based on the UNOS liver transplant database from 1990 to 2002 to estimate a US population-based rate of liver transplantation due to ALF from drugs, liver transplantation due to drug hepatotoxicity accounted for 15% of transplants for ALF during that period.89 Seventy-six percent of the recipients were females with a high representation of African-American women undergoing transplantation for non-APAP DILI. Moreover, a study of 30 non-APAP druginduced fulminant liver failures among adults in Scotland has indicated it is more than twice as common in females as males.90 Together, these studies indicate a greater susceptibility in women for drug-induced ALF following either intrinsic or idiosyncratic toxicity.
Pharmacovigilance studies that suggest gender influences the susceptibility to drug-induced liver injury APAP is an intrinsic or predictable hepatotoxicant that causes dose-dependent liver injury in exposed individuals. Most other examples of human DILI involve idiosyncratic DILI that are peculiar to the individual and therefore are not readily predictable by the pharmacology of the drug or by preclinical testing. Idiosyncratic DILI is further defined as either immunologic (hypersensitivity) or metabolic idiosyncrasy.91 Histopathologically, the hepatocellular type
of DILI is more common in younger patients, whereas the cholestatic pattern increases with older age. Long-term follow up studies in a hepatotoxicity registry has indicated that cholestatic/mixed liver damage is more likely to become chronic while the severity is greater with the hepatocellular pattern.92 An analysis of a computerized diagnoses database in an out-patient hepatology clinic in a Swedish University Hospital during the period 1995–2005 revealed that DILI cases constituted 6% of all outpatients as well as 3% of referrals and occurred more frequently in women.93 In an analysis of 603 DILI cases submitted to the Spanish Registry between April 1994 and August 2007, fulminant liver failure/liver transplantation was encountered as an outcome more frequently in women (P < 0 0.01); however, it was concluded that female sex was not a predisposing factor to overall DILI.94 In a study that included data collected from April 1994 to August 2004, 461 out of 570 submitted cases involving 505 drugs were found to be associated with DILI. Anti-infective drugs were more frequently implicated, with amoxicillin–clavulanate accounting for the 12.8% of the total series. Although there was no difference in sex distribution for DILI, female sex was a significant factor associated with the development of fulminant hepatic failure.95 The susceptibility to amoxicillin–clavulanate-induced liver injury has been linked to the adaptive immune response and class I and II HLA genotypes specifically.96 The inflammatory hepatic disease autoimmune hepatitis (AIH) primarily affects women and is characterized by elevated transaminase levels and the presence of autoantibodies and elevated levels of immunoglobulin G.97 The frequency of druginduced autoimmune-like hepatitis (DIAIH) among patients with classical features of autoimmune hepatitis has been reported to be 9%.98 Specifically, minocycline and nitrofurantoin were implicated in 90% of cases. A search using the Mayo Clinic diagnostic medical index to identify and compare AIH and DIAIH patients indicated that a similar number of DIAIH cases have positive antinuclear antibodies and smooth muscle antibodies compared to the AIH patients.99 Acute onset, female predominance, and absence of cirrhosis at presentation were important clinical manifestations. It has been established that sexual dimorphism exits within the immune system, and females have higher levels of immunoglobulins, greater antibody response to antigens, and higher incidence of autoimmune diseases than males.100
Downloaded from het.sagepub.com at UNIVERSITE LAVAL on July 6, 2014
8
Human and Experimental Toxicology XX(X)
In an extensive analysis of gender and age differences in the reporting of drug-induced hepatic failure (HF) to the spontaneous ADR report system VigiBase™101 from 2000 to 2009, female predominance was found in the largest proportion of cases at age
Female gender as a susceptibility factor for drug-induced liver injury David E. Amacher Hum Exp Toxicol published online 3 December 2013 DOI: 10.1177/0960327113512860 The online version of this article can be found at: http://het.sagepub.com/content/early/2013/12/02/0960327113512860
Published by: http://www.sagepublications.com
Additional services and information for Human & Experimental Toxicology can be found at: Email Alerts: http://het.sagepub.com/cgi/alerts Subscriptions: http://het.sagepub.com/subscriptions Reprints: http://www.sagepub.com/journalsReprints.nav Permissions: http://www.sagepub.com/journalsPermissions.nav
>> OnlineFirst Version of Record - Dec 3, 2013 What is This?
Downloaded from het.sagepub.com at UNIVERSITE LAVAL on July 6, 2014
Article
Female gender as a susceptibility factor for drug-induced liver injury
Human and Experimental Toxicology 201X, Vol XX(X) 1–12 ª The Author(s) 2013 Reprints and permission: sagepub.co.uk/journalsPermissions.nav DOI: 10.1177/0960327113512860 het.sagepub.com
David E. Amacher
Abstract Adverse drug reactions (ADRs) can involve all tissues and organs, but liver injuries are considered among the most serious. A number of prospective, multicenter studies have confirmed a higher risk of ADRs in general among female subjects compared to a male cohort. Although drug-induced liver injury (DILI) is infrequently encountered, the preponderance of evidence suggests that women appear to be more susceptible than men to fulminate hepatic/acute liver failure especially in response to some anti-infective drugs and to autoimmunelike hepatitis following exposure to certain other therapeutic drugs. A number of hypotheses have been proposed to explain this sex difference in susceptibility to DILI. Collectively, these hypotheses suggest three basic sex-dependent mechanisms that include differences in various aspects of drug pharmacokinetics (PK) or pharmacodynamics following the administration of certain drugs; specific hormonal effects or interactions with immunomodulating agents or signaling molecules; and differences in the adverse response of the immune system to some drugs, reactive drug metabolites, or drug-protein adducts. At the preclinical drug safety stage, there is a need for more research on hormonal effects on drug PK and for additional research on gender differences in aberrant immune responses that may lead to idiosyncratic DILI in some female patients. Because the detection of rare but serious hepatic ADRs requires the exposure of very large patient populations, pharmacovigilance networks will continue to play a key role in the postmarketing surveillance for their detection and reporting. Keywords Drug-induced liver injury, female gender, adverse drug response
Introduction Adverse drug reactions (ADRs) are harmful and unintended reactions to medicines that occur at doses normally used for treatment. One type of ADR is caused by an extension or exaggeration of the intended pharmacological action of the drug, and a second type results from toxic effects unrelated to the intended pharmacological action of the drug but dependent upon unique host factors. ADRs vary in severity, and the most serious ADRs are those that require hospitalization, are permanently disabling, or result in death. It has been estimated that ADRs are the fourth to sixth largest cause of mortality in hospitalized patients in the United States.1 ADRs can involve all tissues and organs, but liver injuries are considered among the most serious. A recent systemic review of 106,586 hospitalized patients has shown that approximately 5.3% of the hospital admissions were associated with ADRs.2 A prospective, multicenter study based on intensive pharmacovigilance involving 2,371 patients
for whom 25,532 drugs had been prescribed has confirmed a higher risk of ADR among female subjects than a male cohort.3 Additional general examples will be described in the section on Pharmacovigilance studies that demonstrate sex differences in events diagnosed as ADR and specific liver examples in the section on Pharmacovigilance studies that suggest gender influences the susceptibility to drug-induced liver injury. Overall, compared with male patients, female patients have a 1.5- to 1.7-fold greater risk of developing an ADR.4 Results from a survey conducted between January 1997 and January 2001 involving 10 prescription drugs that had been withdrawn from the US market indicated that 8 posed
Toxadvisor Toxicology Consulting, USA Corresponding author: David E. Amacher, Toxadvisor Toxicology Consulting, P.O. Box 254, Hadlyme, Connecticut, CT 06439, USA.
Downloaded from het.sagepub.com at UNIVERSITE LAVAL on July 6, 2014
2
Human and Experimental Toxicology XX(X)
greater health risks for women than men, and 2 of these resulted in liver failure.5 Four of these drugs may have led to more adverse events in women because they were prescribed more frequently to women than men. The remaining four were associated with more adverse events in women even though they were widely prescribed to both women and men. The term ‘‘gender difference’’ has been used broadly to include differences due to sex, other biological processes, and nonbiological influences.6 The term ‘‘sex differences’’ has been defined as the purely biologic or physiologic differences of females and males.7 In this review, these terms will be used interchangeably to refer to biological processes leading to adverse drug reactions in adult men and women. The purpose of this article is to review the current status of published studies on sex/gender differences in druginduced liver injury (DILI) responses to pharmaceutical agents. The topics include a summarization of general biological differences that can affect drug efficacy and toxicity, the status of clinical testing of new chemical entities in women, established gender differences in drug pharmacokinetics, pharmacovigilance studies that demonstrate sex differences in the diagnosis of ADR, an overview of the etiology of DILI, and a review of recent, specific pharmacovigilance studies that suggest links between gender influences and DILI.
effects on drug metabolism.9 The generally greater fat stores in women may account for larger volumes of distribution (Vd) for lipophilic drugs.10 Differences in the innate and adaptive immune systems between men and women have also been documented,11 and these can result in immune-mediated hepatotoxicity. Other suggested hypotheses include inadvertent drug overdosing on a milligram per kilogram basis, because women generally have a lower body weight than men, that women are more likely to report adverse events than men, and that women take more medications than men and therefore are more likely to experience adverse effects due to drug–drug interactions.6 There also are some differences in the medications prescribed for men and women. For example, women are more often prescribed psychotropic medications and drugs for the cardiovascular system, alimentary tract, and the musculoskeletal system than men.12,13 There can be differences in health issues, for example, there is a generally higher incidence of urinary tract infections in women than in men. Excluding behavioral or dosing differences, these hypotheses collectively suggest three basic mechanisms that include: (1) gender-dependent differences in pharmacokinetics following the administration of certain drugs, (2) gender-specific hormonal effects or interaction with signaling molecules that can affect drug safety, and (3) differences in an aberrant immune response that targets the liver following drug exposure that can result in adverse drug reactions.
General biological differences between men and women that can affect drug The clinical testing of new molecular efficacy and toxicity The current literature suggests a number of possible entities in women susceptibility mechanisms or factors that might be responsible for any differences in the observed incidence of adverse effects between men and women following exposure to therapeutic levels of drugs. Gender-based variations that can influence drug pharmacokinetics and subsequent toxicity include differences in gastric acid secretion, gastrointestinal blood flow, the relative amount of drug-binding proteins, the relative proportions of muscular and adipose tissue, renal blood flow, gender-specific expression of cytochrome P450 (CYP450) isozymes, and physiologic and hormonal changes during the menstrual cycle.8 Lower gastric acid secretion, lower gastric emptying time, and intestinal transit rate in women can affect the absorption and subsequent bioavailability of orally administered drugs. Menstruation, pregnancy, the use of oral contraceptives, and menopause all can have profound
Pre-approval clinical drug development through phases I–III can involve as few as 500 subjects and rarely more than 5,000 subjects.14 Historically, females of all ages have been underrepresented in these clinical trials, and a gender-specific analysis of the data are usually not included in the evaluation of the results. For a number of years, the US Food and Drug Administration (FDA) guidelines specifically excluded participation of females in many clinical drug studies due to the potential risks involved during pregnancy in women of child-bearing age and due to co-existing illnesses in elderly women.15 This followed the belated discovery of the human teratogens thalidomide in 1961 and diethylstilbestrol in 1971.16 In 1993, the FDA published the Guideline for the Study and evaluation of Gender Differences in the Clinical Evaluation of Drugs which lifted this
Downloaded from het.sagepub.com at UNIVERSITE LAVAL on July 6, 2014
Amacher
3
restriction for most women of childbearing potential and encouraged sponsors to collect gender-related data.17 A recent review of clinical trial data submitted to the FDA for New Molecular Entities for adult, nonsex specific indications between January 2006 and December 2007 indicated women were still underrepresented; however the number of trials enrolling women and the number of women participating in phase I trials has increased since 2001.18 As amended in 2001, the National Institute of Health (NIH) Policy and Guidelines on The Inclusion of Women and Minorities as Subjects in Clinical Research now ensure that all NIH-funded clinical research will be carried out in a manner sufficient to elicit information about individuals of both sex/genders and diverse racial and ethnic groups, particularly in NIH-defined phase III clinical trials, in order to examine differential effects on such groups.19 It should be noted, however, that the detection of at least 1 ADR with an incidence of 1 in 10,000 exposed individuals would require the exposure of at least 30,000 people to the drug,20 which can only be accomplished in the phase IV postapproval period.
Gender differences in drug pharmacokinetics The pharmacological as well as any toxicological interaction of a drug with the body proceeds through the sequential steps of pharmacokinetics (PK) that include drug absorption, distribution, metabolism, and elimination (ADME); any of which can be affected by gender differences. Studies in murine liver models have identified more than 1,000 genes whose expression is sex dependent, and their gene products include many drug metabolizing enzymes (DMEs).21 Although no genderspecific DMEs have been identified in humans, the analysis of gene expression data in 427 liver samples by Yang et al. have identified 77 drug metabolizing or transporter genes that demonstrate differential expression based on sex.22 Any sex differences in the gene expression of DMEs and transporters can affect drug efficacy or result in ADR by altering ADME in organs such as the liver. Changes in pharmacodynamics can affect both the desired therapeutic effect of a drug as well as its adverse effect profile.23 In the liver, drugs metabolized by phase I metabolism (oxidation, reduction, and hydrolysis via CYP450s), phase II conjugative metabolism (glucuronidation, conjugation, glucuronyltransferases, methyltransferases, dehydrogenases) and by the combined oxidative and conjugation processes are usually cleared
faster in men compared to women on a milligram per kilogram basis.24 The human CYP superfamily consists of 57 functional genes and 58 pseudogenes.25 Of these, members of the CYP1, 2, and 3 families play key roles in the metabolism of therapeutic drugs. Some DME genes such as CYP3A4, CYP2B6, and GSTA1 display large and measurable inter-individual variability.26 CYP1A2 is a major CYP450 in human liver that metabolizes a variety of clinically important drugs and several endogenous compounds such as arachidonic acids and steroids.27 Microsomal CYP1A2 activity measured by 7-ethoxyresorufin Odealkylase activity has been reported to be greater in males than females, which is consistent with clinical observation.28 In addition to CYP1A2, men potentially have a higher activity relative to women for CYP2E1, for the drug efflux transporter Pglycoprotein (P-gp/MDR1), and for some isoforms of glucuronosyltransferases and sulfotransferases.29 Human CYP3A4 is the most abundant hepatic phase I enzyme that metabolizes approximately 50% of the marketed drugs.30 In cryopreserved primary human hepatocytes, Parkinson et al. found that the gender difference in CYP3A4 activity (females ¼ twice males) as measured by testosterone 6-betahydroxylase was statistically significant, which supports the clinical observation that females metabolize certain CYP3A4 substrates faster than males do,28 and therefore the total clearance of CYP3A substrates appears to be mildly or moderately faster on a milligram per kilogram basis in women compared to men.24 The majority of all published studies indicate that CYP3A4 activity is generally higher in women than in men which is remarkable considering the aforementioned extensive individual variability, whereas the activity of many other systems involved in drug metabolism may be higher in men than in women. Gender-based differences in the expression levels of hepatic or renal metabolic enzymes and transporters can result in longer drug half-life, as for example, with anti-cancer drugs in female patients. In an FDA review of 300 new drug applications between 1995 and 2000, for the 163 that included a sex analysis, 11 of those drugs demonstrated >40% difference in PK between males and females, and even though this was listed on the product label, no dosing recommendations were made based on sex.31 Physiological factors that may result in genderrelated PK differences include the generally lower bodyweight and organ size, a higher percentage of
Downloaded from het.sagepub.com at UNIVERSITE LAVAL on July 6, 2014
4
Human and Experimental Toxicology XX(X)
body fat, a lower glomerular filtration rate, and different gastric motility in women compared with men.29 Men and women also show different pharmacodynamic responses to a variety of drugs. In women, issues such as menopause, pregnancy, oral contraceptive use, and menstruation can also have profound effects on drug metabolism which can be clinically important. Pregnancy may increase the elimination of antiepileptic agents, reducing their efficacy, while oral contraceptive use can interfere with the metabolism of many coadministered drugs.9 Drug metabolism by CYP2C9, CYP2C19, and Nacetyltransferase, appear to be similar in men and women (mg/kg), and the clearance of P-gp substrates also appear to be similar in men and women. Some previous authors have concluded that sex differences in drug metabolism and elimination are in fact primarily related to steroid hormone levels.32 In women, drug metabolism can be affected by the aforementioned sex-specific factors: menopause, pregnancy, and menstruation. For example, recent studies have demonstrated that pregnancy, depending upon the stage, alters the activity of CYP2D6, CYP3A4 (increased), and CYP1A2 (decreased), and increases the activity of the renal transporters organic cation transporter and P-gp.33 Studies in rat and mouse models have demonstrated that growth hormone (GH) is the key hormonal factor that controls sex differences in the expression of a large number or hepatic gene products that include P450 isozymes and other DMEs, and human similarly display plasma GH patterns that are sex dependent and regulated by gonadal hormones.21 Based on clinical data, during adolescence, GH is known to regulate sex-specific DME expression, and it appears that sex hormones can significantly influence DME activity in an isoform-specific manner.34 Recent results from an in vitro system designed to elucidate biological mechanisms for human sex differences in susceptibility to DILI have suggested that interaction between sexhormones and pro-inflammatory cytokines might alter liver metabolism and thus contribute to marked sex differences in susceptibility to DILI.35
Pharmacovigilance studies that demonstrate sex differences in events diagnosed as ADR Pharmacovigilance is based on information gathered from individual case safety reports and other pharmacoepidemiological data. Notably, some large drug
surveillance studies in hospital patients have demonstrated significant differences between men and women for the risk of ADR-related hospitalization. A summary of some recently published studies follows. In a study of ADR as the cause for hospital admission that involved the prospective analysis of 12,249 hospitalized patients, older age, and female gender were associated with ADR-related hospitalizations.36 The prospective analysis of 907 patients in 2 German hospitals,37 an event monitoring of 4,331 hospitalized patients in Switzerland,38 an investigation of 513,608 general practice patients using data collected from 48 national cohort studies in England,39 a nation-wide study of ADR-related hospitalizations between 2000 and 2005 in the Netherlands,40 a prospective drug surveillance study of 1,920 patients hospitalized in Chile,41 and a prospective multicenter study of 2371 in-patients in Germany and Israel,13,42 all concluded there was a higher risk of ADRs among female subjects. More specifically, a study of hospital admission data from the Dutch National Register between 2000 and 2005 has indicated clear sex differences in ADRs requiring hospital admission for cardiovascular drug groups.43 Conversely, a study of ADRs recorded in the French MidiPyre´ne´es Pharmacovigilance Centre in 1998, there was no statistically significance in the incidence of ADRs reported in males versus females for the total population.44 But serious ADRs were more frequently reported in females between 20 and 29 years and in males 0-9 and 60-69 years. ADRs were more frequently reported in females for genitourinary, sex hormones, antineoplastic, respiratory, and antiparasitic drugs. The International Classification of Diseases (ICD) is a standard diagnostic tool used to monitor the incidence and prevalence of diseases and other health problems. The adverse reactions monitored in these studies generally depend upon the drug class and cover a wide range of clinical endpoints, often using the ICD coding system,45 but excluding such things as errors in administration, preexisting disease, and intended forms of overdose. Some examples of diagnoses codes associated with high prevalence of ADR include drug-induced neutropenia, hypoglycemia, hepatitis unspecified, anaphylactic shock due to drugs, shock due to anesthesia,46 and gastrointestinal bleeding.47 In addition to the general consensus that ADRs are more frequently reported in women, the severity of ADRs, particularly liver toxicity, generally appears to be greater in women than men.
Downloaded from het.sagepub.com at UNIVERSITE LAVAL on July 6, 2014
Amacher
5
Drug-induced liver injury DILI refers to a specific liver injury or toxicity caused by drugs or other xenobiotics, and comprises a specific and important category of ADR. The basic mechanisms proposed for DILI are based on the production of reactive metabolites, immune-mediated hepatotoxicity, and a ‘danger signal’ hypothesis and/or alterations in mitochondrial function.48 The danger hypothesis suggests that the primary determinant for the mounting of an immune response against a drug is whether the drug or more likely, a reactive metabolite, cause cellular damage.49 The danger signal molecules are thought to act through the same receptors as pathogen-associated molecules to subsequently stimulate the immune system. In what has been referred to as the pharmacological interaction with immune receptors (p-i) concept, some drugs may bind directly and reversibly to immune receptors which stimulate those cells.50 For the mitochondrial dysfunction model, a 3-step mechanistic working model of drug-induced hepatotoxicity has been described by Russmann et al.51 These steps include: (1) An initial injury that results from direct cellular stress, direct mitochondrial inhibition, and/ or specific immune reactions, (2) the initial injury can then lead to direct mitochondrial impairment via mitochondrial permeability transition (MPT), and (3) the MPT subsequently leads to either necrosis or apoptosis depending on the availability of ATP. The liver contains components of both the innate and adaptive immune system, and is the major organ for the drug metabolism. In the United States, the most common drugs leading to DILI are antibiotics, central nervous system agents, herbal/dietary supplements, and immunomodulatory agents.52 The associations for human liver and skin ADR are known to involve human leukocyte antigen (HLA) genes for drugs such as abacavir and carbamazepine.53 In addition to those two drugs, strong HLA associations indicating the involvement of the immune system have been demonstrated with a number of drugs that cause liver injury including flucloxacillin, lumiracoxib, lapatinib, and ximelagatran.54 Further, drug-induced immune-mediated hepatic injury also has been implicated for drugs such as halothane, tienilic acid, dihydralazine, chlorpromazine, erythromycins, amoxicillin–calvulanic acid, sulfonamides and sulindac, methyldopa, minocycline, and nitrofurantoin.55 The latter authors have delineated the formation autoantibodies directed toward the microsomal cytochromes that participate in the metabolism of some
of those drugs, although it is not known if they participate in the destruction of hepatocytes. Minocyclineassociated hepatotoxicity is more common in women. Fortunately, the majority of cases of DILI are acute illnesses that are quickly resolved after the offending medication is stopped. In a 10-year study of 1,100 cases reported to the Danish Committee on ADR, hepatic injuries accounted for 5.9% of all ADRs reported, and 14.7% of the lethal adverse drug reactions.56 Between January 2001 and June 2007, retrospective analysis of 170 patients who had been diagnosed with DILI indicated that a total of 83 different drugs were associated with the hepatotoxicity with a single drug suspected in approximately 58% of the individuals.57 The main causative group was antibiotics. Of the 62 patients hospitalized, acute liver failure (ALF) developed in 14 (8.2%), chronicity occurred in 19 (11.2%), and death occurred in7 (4.1%). Clinically, DILI is a diagnosis of exclusion of other causes and the identification of a signature pattern of disease manifestation coordinated with the initiation and discontinuation of drug treatment.58 An estimated annual incidence rate of DILI worldwide is 13.9–24.0/ 100,000 inhabitants.59 Nonetheless, the diagnostic approach to study DILI remains rudimentary and inaccurate due to the lack of reliable markers for use in general practice,60 and the incidence of DILI in the general population is thought to be largely underestimated.61 Several multicenter networks have been established world-wide for the assessment of DILI. In the United States, The Drug-Induced Liver Injury Network (DILIN) was established in 2003 to advance understanding and research into DILI as a collaborative effort among the NIH, five clinical sites, and a data coordinating center.62 DILI is estimated to occur in 1–2 cases per 100,000 person years and represents 1.2–6.6% of cases of acute liver disease reported by tertiary referral centers.63 In an extensive international study to identify causes of DILI at three major DILI registries, a total of 319 pharmaceuticals were identified as causes for liver injury.58 In descending order, drugs that were frequently identified in the combined list (greater than 30 cases reported) included amoxicillin–clavulanic acid, diclofenac, cotrimoxazole, isoniazid, and disulfiram. It should be noted that there is marked geographic variation in the identification of drugs linked to DILI,59,60 and a single drug is not clearly involved in all cases.57 There are also differences in prescribing certain drugs in men and women.40 In a screening of 6,383 hospitalized patients using CIOMS (Council for International Organizations of Medical Sciences) criteria to access the probability of DILI and to exclude any
Downloaded from het.sagepub.com at UNIVERSITE LAVAL on July 6, 2014
6
Human and Experimental Toxicology XX(X)
disease-related causes for liver injury, the incidence of DILI was 1.4%, and the drug classes most frequently causing DILI were heparins, antibacterials, tuberculostatics, and antineoplastic agents.64
Alcohol-induced liver injury In women and some female animals, most studies reveal a propensity toward greater alcohol-induced liver injury (ALI) due to female gender.65 Because gastric alcohol dehydrogenase is lower in females than males, women tend to have a lower alcohol toxic threshold than men.66 Moreover, it has been established clinically that women develop ALI more rapidly and to greater extent than men while consuming fewer grams of alcohol.67 This occurs even though a study of alcohol metabolism has demonstrated that the mean alcohol elimination rate and mean computed liver volume are not significantly different in men and women.68 In studies with rats fed an ethanolcontaining diet, Kono et al. have concluded that elevated endotoxin after ethanol administration triggers more Kupffer cell activation via enhanced CD14 expression in females. This leads to increases in tumor necrosis factor-a messenger RNA (mRNA) expression in the female liver resulting in more severe ALI.69 Because estrogen has major influence on the susceptibility of Kupffer cells to gut-derived lipopolysaccharide resulting in increased proinflammatory cytokine production, it has been suggested that this could be a major contributing factor to the increased risk of women to alcohol-induced liver disease.70 In summary, women develop ALI more rapidly and have a lower alcohol toxic threshold than men.
Acetaminophen overdose Administered at normal therapeutic doses, the analgesic acetaminophen (APAP, paracetamol) is safe, but it can be intrinsically hepatotoxic at high doses. During normal hepatic metabolism, a major portion of APAP is conjugated with either sulfate or glucuronic acid to form water-soluble metabolites that are readily excreted, and only small amounts of the reactive intermediate, N-acetyl-p-benzoquinon imine (NAPQI), are formed by the CYP450 enzymes71,72 which include CYP2E1, CYP1A2, and CYP3A4.73 Further studies have revealed that activation of APAP to this active metabolite is primarily carried out by CYP2E1.74 APAP clearance has been shown to be 22% greater in males than females due to the increased activity of the glucuronidation pathway in males,75 which occurs
mainly by UDP-glucuronosyltransferases (UGT). UGT2B15, which has recently been demonstrated to play a role in APAP glucuronidation,76 shows higher activity in males compared to females.77 When an adequate supply of cellular reduced glutathione (GSH) is available, most of the reactive intermediate NAPQI can be detoxified. But if the hepatic concentration of GSH becomes depleted, as for example at high APAP doses, subsequent covalent binding to the sulfhydryl (–SH) groups of various cellular proteins increases, which ultimately can result in hepatic necrosis.71 A single 10–15 g APAP dose, or occasionally less, can produce clinical evidence of liver injury while fatal fulminant disease is usually, although not always, associated with ingestion of 25 g.78 Overdoses of APAP result in the generation of APAP-protein adducts which have been shown to provide an excellent correlate with the severity of the APAP toxicity.79 The innate immune system is thought to play a role in the exacerbation and progression of APAP-induced liver injury.80 APAP liver toxicity can result from either accidental overdose or intentional overdose. Published findings provide strong support for the hypothesis that mitochondrial dysfunction and DNA damage are critical events in the mechanism of cell necrosis after APAP overdose in patients.81 In a study that reviewed 352 poisoning episodes involving APAP overdoses, the accidental overdose category constituted 16 (9%) of 182 hospital index visits with no apparent gender differences noted. Within the intentional overdose group, 1.2% developed ALF versus 25% of the accidental group. Young females with intentional overdoses accounted for most of the cases.82 This pattern of mortality has been noted in other studies. In a study of 663 patients, unintentional APAP overdose was associated with increased mortality compared with intentional overdoses, despite lower APAP concentrations.83 APAP is considered a model intrinsic or predictable hepatotoxicant. However, in recent years, drug candidates that cause intrinsic liver injury such as APAP are now efficiently identified and discontinued from further development in the preclinical testing state. The preponderance of reported adverse liver reactions is idiosyncratic in nature, usually occurring 5–90 days after the causative medication was last taken.84 Based on their clinical characteristics, idiosyncratic adverse drug reactions in general appear to be immune-mediated, and a substantial body of evidence further suggests that most although not all are caused by chemically reactive species.85 Thus, DILI
Downloaded from het.sagepub.com at UNIVERSITE LAVAL on July 6, 2014
Amacher
7
associated with recently marketed drugs is predominantly if not exclusively idiosyncratic in nature.
Acute liver failure ALF is characterized by coagulopathy and encephalopathy. In a multicenter, prospective study involving 1198 subjects enrolled at 23 sites in the United States, Reuben et al. concluded that DILI is an uncommon cause of ALF that evolves slowly and affects a disproportionate number of women and minorities.86 In another study, the incidence of ALF, which again occurred more frequently in women, was about 2000 cases per year with greater than 50% due to therapeutic drug use.6 In yet another study of 308 consecutive patients with ALF admitted to 17 tertiary care centers in the US, 73% were women.87 APAP overdose (39% of cases) and idiosyncratic drug reactions (13% of cases) were the most frequent apparent causes of ALF. In a multicenter, prospective US study of APAP-induced ALF over a 6-year period, the medium dose ingested was 24 g, which is equivalent to 48 extra-strength tablets.88 In a study based on the UNOS liver transplant database from 1990 to 2002 to estimate a US population-based rate of liver transplantation due to ALF from drugs, liver transplantation due to drug hepatotoxicity accounted for 15% of transplants for ALF during that period.89 Seventy-six percent of the recipients were females with a high representation of African-American women undergoing transplantation for non-APAP DILI. Moreover, a study of 30 non-APAP druginduced fulminant liver failures among adults in Scotland has indicated it is more than twice as common in females as males.90 Together, these studies indicate a greater susceptibility in women for drug-induced ALF following either intrinsic or idiosyncratic toxicity.
Pharmacovigilance studies that suggest gender influences the susceptibility to drug-induced liver injury APAP is an intrinsic or predictable hepatotoxicant that causes dose-dependent liver injury in exposed individuals. Most other examples of human DILI involve idiosyncratic DILI that are peculiar to the individual and therefore are not readily predictable by the pharmacology of the drug or by preclinical testing. Idiosyncratic DILI is further defined as either immunologic (hypersensitivity) or metabolic idiosyncrasy.91 Histopathologically, the hepatocellular type
of DILI is more common in younger patients, whereas the cholestatic pattern increases with older age. Long-term follow up studies in a hepatotoxicity registry has indicated that cholestatic/mixed liver damage is more likely to become chronic while the severity is greater with the hepatocellular pattern.92 An analysis of a computerized diagnoses database in an out-patient hepatology clinic in a Swedish University Hospital during the period 1995–2005 revealed that DILI cases constituted 6% of all outpatients as well as 3% of referrals and occurred more frequently in women.93 In an analysis of 603 DILI cases submitted to the Spanish Registry between April 1994 and August 2007, fulminant liver failure/liver transplantation was encountered as an outcome more frequently in women (P < 0 0.01); however, it was concluded that female sex was not a predisposing factor to overall DILI.94 In a study that included data collected from April 1994 to August 2004, 461 out of 570 submitted cases involving 505 drugs were found to be associated with DILI. Anti-infective drugs were more frequently implicated, with amoxicillin–clavulanate accounting for the 12.8% of the total series. Although there was no difference in sex distribution for DILI, female sex was a significant factor associated with the development of fulminant hepatic failure.95 The susceptibility to amoxicillin–clavulanate-induced liver injury has been linked to the adaptive immune response and class I and II HLA genotypes specifically.96 The inflammatory hepatic disease autoimmune hepatitis (AIH) primarily affects women and is characterized by elevated transaminase levels and the presence of autoantibodies and elevated levels of immunoglobulin G.97 The frequency of druginduced autoimmune-like hepatitis (DIAIH) among patients with classical features of autoimmune hepatitis has been reported to be 9%.98 Specifically, minocycline and nitrofurantoin were implicated in 90% of cases. A search using the Mayo Clinic diagnostic medical index to identify and compare AIH and DIAIH patients indicated that a similar number of DIAIH cases have positive antinuclear antibodies and smooth muscle antibodies compared to the AIH patients.99 Acute onset, female predominance, and absence of cirrhosis at presentation were important clinical manifestations. It has been established that sexual dimorphism exits within the immune system, and females have higher levels of immunoglobulins, greater antibody response to antigens, and higher incidence of autoimmune diseases than males.100
Downloaded from het.sagepub.com at UNIVERSITE LAVAL on July 6, 2014
8
Human and Experimental Toxicology XX(X)
In an extensive analysis of gender and age differences in the reporting of drug-induced hepatic failure (HF) to the spontaneous ADR report system VigiBase™101 from 2000 to 2009, female predominance was found in the largest proportion of cases at age
