Drug-induced liver disease: primer for the primary care physician.

Disease-a-Month 60 (2014) 55–104

Contents lists available at ScienceDirect

Disease-a-Month journal homepage: www.elsevier.com/locate/disamonth

Drug-induced liver disease: Primer for the primary care physician James T. O’Donnell, PharmD, MS, FCP, Donald H. Marks, MD, PhD, Paul Danese, PhD, James J. O’Donnell III, MS, PhD

Introduction: The problem of drug-induced liver disease As a metabolizer of harmful and foreign substances in the blood, the liver is particularly susceptible to drug-induced disease, regardless of the site of action for therapeutic effect. Liver adverse reactions are linked to approximately 1100 drugs, 2 3% of all hospitalizations, and 3% of all jaundiced patients. As the primary cause of drug-induced death and the main reason for the removal of drugs from the pharmaceutical market,1 drug-induced liver dysfunction/disease (DILD) is or should be of utmost concern to pharmacists and physicians. Since symptoms and latency vary widely according to the drug(s) in question, as well as the age, sex, and overall health of the patient, detection can be problematic. For accurate diagnosis, an extensive drug history, a low threshold of suspicion, and the elimination of other liver disease causes are integral.2 An exposition of the metabolic processes promotes understanding of how certain drug interactions occur and what to do about those interactions. DILD can be distinguished from other forms of liver disease by its 4 major pathological patterns of injury: (1) hepatitis, also known as hepatocellular or cytotoxic (most common), (2) cholestatic, (3) vascular, and (4) neoplastic. Which pattern or type occurs is dependent upon many factors, including the specific drug, the details of treatment (dosing and duration), and patient susceptibility. This can rationalize how DILD can result in a wide array of symptoms that at first glance, look like natural liver deterioration.3–8 Given its rarity, drug-related hepatotoxicity may not occur during clinical trials, which are usually limited to a few thousand participants. However, after approval of a drug for use and subsequent marketing, large numbers of patients are exposed and rare toxic effects may emerge.9 The diversity in the clinical pathology of the DILD contributes to the diagnostic difficulties that arise in discovering the role a specific drug plays in a clinical manifestation. Aside from the plethora of classic drug agents that result in hepatotoxicity,10 there are many other agents that need to be taken into account. Even the excipients within a formulation are considered, especially with the increasingly popular herbal medicines, many of which are undocumented.11–13 Illicit drugs such as cocaine and ecstasy (3,4-methylenedioxy-N-methylamphetamine) can also cause 0011-5029/$ - see front matter & 2014 Mosby, Inc. All rights reserved. http://dx.doi.org/10.1016/j.disamonth.2013.11.002

56

J.T. O’Donnell et al. / Disease-a-Month 60 (2014) 55–104

DILD.8 Due to the large variety of agents, drug hepatotoxicity remains a significant challenge for the scientific community, namely, physicians, health authorities, and pharmaceutical firms. The majority of liver adverse drug reactions (ADRs) are present in only a small percentage of patients, which makes it challenging to spot during drug development. Liver dysfunction is mostly recognized due to random reports within the first 2 years of the newly marketed drug. The presence of DILD is weakly documented by only a few studies. However, even in these cases, diagnoses are not absolute.1 Both genetic and environmental factors also play into one’s susceptibility to DILD, and there are many poorly understood aspects needing further research. These future ventures should include studies of the prevalence of liver ADRs in conjunction with specific genetic markers and determining the role that environmental factors play.1 There are few treatment options for DILD. One can discontinue the drug, induce N-acetylcysteine in the case of acetaminophen toxicity, carnitine for valproate-induced mitochondrial injury, or undergo a liver transplant in fulminant (end-stage hepatitis) liver cases. Guidelines for monitoring therapy are set in place with the use of specific drugs such as halothane and methotrexate, and they should be heeded appropriately. Drugs with known hepatotoxic potential should be used sparingly with known alcoholics. Patients with impaired hepatic function should be administered a decreased dose. When confronted with a patient with liver disease, the treating physician must consider the possibility of drug-induced hepatic injury. As contrasted to prescribing physicians, there may be little that the pharmacist can do, other than recognize patients at risk, understand the meanings and values of certain liver tests and interpretations, recommend to patients when and why liver function testing is important, and inform the patient that alcohol may exacerbate the hepatotoxic potential of other drugs. There should be continued awareness for the general public to use hepatotoxic drugs with caution and know the risks for liver toxicity—especially when using acetaminophen or anabolic steroids.2 In the United States, drug overdose is now the leading cause of acute liver failure among transplant recipients. Acetaminophen is the most common culprit in these cases, and the patients often have no prior history of liver disease. When a common drug is found to cause even rare hepatotoxicity, it may be removed from clinical use. Although such a drug poses great danger to only a few patients, its removal leads to the loss of drug availability to many.9 For practicing physicians, drug-related hepatotoxicity is a liability risk; for the pharmaceutical industry, it leads to financial losses; and from a regulatory perspective, it is the most common reason for regulatory actions on the part of the Food and Drug Administration (FDA) (Table 1).9 In a consensus report involving the criteria for detecting drug-induced liver disorders, the term “liver injury” was defined as the increase of greater than or equal to 2 times the upper limit of normal (ULN) alanine aminotransferase (ALT) or conjugate bilirubin (CB). This is also the case for a combined increase in aspartate aminotransferase (AST), alkaline phosphatase (ALP), or total bilirubin (TB).36 Levels of ALT that are greater than twice the ULN should be taken seriously and therefore alternate drugs should be used. If there are no alternates, the physician may choose to challenge the patient diagnostically. Clinically, the approach to adverse hepatic reactions is to classify them as either dose-dependent (type I) or dose-independent (type II). Type I usually involves patients who have ingested a sufficient amount of drug and type II is a rare manifestation of therapeutic drug doses.2 When determining the diagnosis of drug-induced liver disorders, the side effects of non-liver origin may play a significant role. These side effects include rash, fever, and adenopathy in the case of sulfonamide-, phenylbutazone-, quinidine-, and procainamide-associated jaundice. Infection by Epstein–Barr virus and cytomegalovirus are possible diagnoses in this scenario.2 Toxicity caused by non-steroidal anti-inflammatory drugs (NSAIDs) may occasionally include both the liver and the bone marrow. Also, amiodarone, disulfiram, and isoniazid may cause both liver injury and polyneuropathy. Lung fibrosis is a much-feared side effect of methotrexate and nitrofurantoin treatment and is associated with chronic hepatitis. This mostly happens in cases

Table 1 Prescription drugs withdrawn from market for safety reasons in the United States from 1997 to 2011. Bold entries identify drugs withdrawn for hepatotoxicity.81 Sources “Drug Recalls.” U S Food and Drug Administration Home Page. Web. 16 Dec. 2011. 〈http://www.fda.gov/Drugs/DrugSafety/DrugRecalls/default.htm〉. “Drugs@FDA.” U S Food and Drug Administration Home Page. Web. 16 Dec. 2011. 〈http://www.accessdata.fda.gov/scripts/cder/drugsatfda/〉. “List of Withdrawn Drugs.” Wikipedia, the Free Encyclopedia. Web. 16 Dec. 2011. 〈http://en.wikipedia.org/wiki/List_of_withdrawn_drugs〉.

Drug

Brand name

Type of drug

Date approved

Date withdrawn

Time on market Primary health risk

Estimated U.S. sales

Eli Lilly and Co.

Opioid analgesic

9/10/2003

11/19/2010

7 years

Not available

11/21/2001

10/25/2011

10 years

Addictiveness and heart problems Increased bleeding

11/1997 5/2000 10/27/2003

10/08/2010 06/21/2010 6/8/2009

3 years 10 years 5.67 years

Not available Not available $500 million

12/29/2003

5/14/2008

4.42 years

7/24/2002 12/30/1988

3/30/2007 3/29/2007

4.67 years 18.25 years

Heart attack/stroke Veno-occlusive disease Progressive multifocal leukoencephalopathy (PML) Liver and cardiovascular toxicity Adverse cardiovascular effects Valvular dysfunction

9/24/2004

7/2005

10 months

Fatal interaction with alcohol

Not available

1975

3/2005

30 years

Hepatotoxicity

Not available

11/23/2004

2/28/2005

3 months

Not available

Sibutramine Gemtuzumab ozogamicin Efalizumab

Meridia Mylotarg Raptiva

Human activated protein C Abbott Laboratories Oral anorexiant Wyeth/Pfizer, Inc. Monoclonal antibody Genentech, Inc. Monoclonal antibody

Aprotinin

Trasylol

Bayer

Tegaserod maleate Pergolide mesylate

Zelnorm Permax

Hydromorphone hydrochloride Premoline

Palladone

Novartis Valeant Pharm International Purdue Pharma

Basic pancreatic trypsin inhibitor 5-HT4 agonist Dopamine receptor agonist Opioid analgesic Dopamine reuptake inhibitor Monoclonal antibody Cox-2 inhibitor Fluoroquinolone antibiotic Cox-2 Inhibitor Cholesterol lowering

11/1/2002 12/19/1997

4/7/2005 1999

2.5 years 2 years

Progressive multifocal leukoencephalopathy Serious skin reactions Hepatotoxicity

5/21/1999 6/26/1997

9/30/2004 8/8/2001

5.33 years 4.17 years

Heart attack and stroke Rhabdomyolsis

Anesthetic and muscle relaxant Gastrointestinal Heartburn

9/19/1999

3/30/2001

7 months

Bronchospasm

Not available $160 million in 1st year $2.5 billion $554 million in 2000. $23 million

2/9/2000 7/29/1993

11/28/2000 7/14/2000

9 months 7 years

Ischemic colitis Heart rhythm abnormalities

$50.4 million $2.5 billion

Cylert

Eli Lilly and Co.

Tysabri

Valdecoxib Trovafloxacin

Bextra Trovan

Pfizer, Inc. Pfizer, Inc.

Rofecoxib Cerivastatin sodium

Vioxx Baycol

Merck & Co. Bayer

Rapacuronium bromide

Raplon

Organon

Alosetron hydrochloride Cisapride monohydrate

Lotronex Propulsid

Glaxo Wellcome Janssen Pharmaceuticals

Not available Not available Not available

57

Natalizumab

Abbott Laboratories Biogen

a

$200 million


Acetaminophen and Darvocet propoxyphene napsylate Drotrecogin alfa Xigris

Manufacturer

58

Table 1 (continued ) Brand name

Manufacturer

Type of drug

Date approved

Date withdrawn

Time on market Primary health risk

Estimated U.S. sales

Troglitazone

Rezulin

Warner-Lambert

Type 2 diabetes

1/29/1997

3/21/2000

2.16 years

Liver failure

Asternizole

Hismanal

Antihistamine

12/19/1988

5/18/1999

6 months

Heart rhythm abnormalities

Grepafloxacin Mibefradil

Raxar Posicor

Janssen Pharmaceuticals Glaxo Wellcome Roche Laboratories

$2.1 billion worldwide $23 million

11/6/1997 6/20/1997

11/1/1999 6/8/1998

2 years 1 year

Duract Seldane and Seldane-D Pondimin

Wyeth-Ayerst Hoechst Marion Roussel Wyeth-Ayerst

7/15/1997 5/8/1985

6/22/1998 2/27/1998

1 year 2.75 years

Heart rhythm abnormalities Drug interaction and lowered heart rate in women Liver failure Heart rhythm abnormalities

$23.5 million Not available

Bromfenac Terfenadine

Antibiotic Blood pressure and cardiovascular Pain reliever Antihistamine Appetite suppressant

6/14/1973

9/15/1997

23.25 years

Valvular heart disease

Not available

Redux

Wyeth-Ayerst

Appetite suppressant

4/29/1996

9/15/1997

1.42 years

Valvular heart disease

$255.3 million

Fenfluramine hydrochloride Dexfenfluramine hydrochloride a

Returned to U.S. market.

$89.7 million Not available


Drug


59

when the drug is continued in patients with subclinical or undetected liver injury. In rare cases, the liver dysfunction can progress to chronic liver disease such as cirrhosis, despite drug cessation. Other drug examples that are associated with liver cirrhosis include amiodarone, retinol (vitamin A), chlorpromazine, and methyldopa. Anabolic steroids and other drugs have been connected to hepatocellular adenoma and carcinoma, angiosarcoma, and nodular focal hyperplasia.2 Liver damage by non-narcotic analgesics usually manifests into acute hepatitis with a degree of cholestasis. This is often indistinguishable from viral hepatitis and sometimes chronic active hepatitis. Acetaminophen in overdose may cause acute hepatic necrosis which has a chance of progressing to fulminant hepatic failure. Aside from the case of acetaminophen, the mechanisms behind non-narcotic analgesic liver damage remain to be elucidated.42

History of hepatotoxic drug reactions Following a major increase in the 1960s, the incidence of drug-induced liver disease has continued to rise. A larger number of agents now appear to contribute to the total burden of drug-induced liver disease, but the incidence of hepatotoxicity for most drugs is still very low. Several of the best known hepatotoxic drugs have been abandoned or their use is diminishing. Acetaminophen, halothane, isoniazid, and anticancer drugs are among the agents that continue to cause concern. Drugs are an uncommon cause of most types of liver disease. However, in comparison to similar disorders resulting from other etiologies, drug-induced disorders tend to be more severe. This, together with the diversity of hepatic lesions that drug reactions may produce, are reasons why drug-induced liver disease assumes an importance that is disproportionate to its low incidence. The modern era of chemotherapeutics was introduced by the use of sulfonamides, penicillin, agents that correct hormonal disturbances, antituberculous drugs, and eventually cytostatic antileukemic drugs. Thus, reports of hepatic injury and jaundice due to sulfonamides, antithyroid drugs, anabolic steroids, and methotrexate gradually appeared during the 1940s and early 1950s. Many of the drugs used during the first 70 years of the 20th century carried a hepatotoxic potential that was not recognized for many years. One of the most spectacular examples was oxyphenisatin, a component of many popular laxatives, including Carter’s Little Liver Pills. This agent was used for 40 years before Reynolds and colleagues (1971) identified it as the cause of “puzzling cases” of acute or chronic hepatitis.14 During the last 50 years, there has been an exponential increase in the quantity and diversity of therapeutic substances introduced into medical practice. For most of these, hepatotoxicity (or more usually, an adverse hepatic drug reaction) is an exceedingly rare complication; typical levels of risk are 1–10 per 100,000 of those exposed. It is, therefore, not surprising that recognition that a new drug can cause liver disease is often delayed until many years after the commencement of marketing. There may then be a further delay before there is widespread awareness of that drug’s potential to cause a hepatic drug reaction. Typical examples are alphamethyldopa, isoniazid, and erythromycin esters. Alternatively, some physicians may initially refuse to accept that an otherwise useful and safe drug is really the causative agent for liver disease that has been ascribed to it. The prototypical example is the controversy that surrounded halothane as a possible cause of postoperative jaundice. Failure to accept that a drug can cause liver disease will retard efforts to reduce the frequency of the reaction. The entire spectrum of liver injury associated with viral hepatitis, both morphologic and clinical, can result from drug administration. The assumption that a disease resembling viral hepatitis is actually produced by a drug will continue to rest on evidence provided by historical and epidemiologic data. Hepatotoxic reactions can be unpredictable and, with few exceptions, are not produced in animals. By contrast with cholestatic drug reactions, which are common and relatively benign, hepatocellular reactions are uncommon but are associated with significant mortality.

60


Examples of hepatotoxic drugs include halothane, isoniazid, phenytoin, and carbamazepine. The presence of eosinophilia is occasionally helpful in suggesting drug-induced hepatitis, but this finding in the blood occurs in fewer than half the patients. Upon withdrawal of the offending agent, patients generally recover without sequelae. However, a second exposure, as in the case of halothane, may precipitate massive hepatic necrosis.

Overview of DILD pathogenesis, risk factors, presentation, diagnosis, and organization The ability of a wide range of medications to cause DILD is well-known to the medical community. Familiar examples include acetaminophen, methotrexate, and statins. The liver is the principal site of drug metabolism and detoxification in the body. Many medications pass through the liver and remain concentrated there, leading to local higher concentrations of potentially hepatotoxic medications. People are genetically diverse, which affects the way each of us metabolizes drugs and other chemicals. Liver toxicity of medication should be taken in the context of the increasing number of different medications available, the rising number of medications individual patients are concurrently taking, the complexity of drug–drug interactions, and the toxicity of drug metabolites. According to a guidance issued by the FDA in August 2009, drug-induced liver injury (DILI) “has been the most frequent single cause of safety-related drug marketing withdrawals for the past 50 years (e.g., iproniazid), continuing to the present (e.g., ticrynafen, benoxaprofen, bromfenac, troglitazone, and nefazodone). Hepatotoxicity discovered after approval for marketing also has limited the use of many drugs. Several drugs have not been approved in the United States because European marketing experience revealed their hepatotoxicity (e.g., ibufenac, perhexiline, and alpidem). Finally, some drugs were not approved in the United States because premarketing experience provided evidence of the potential for severe DILD (e.g., dilevalol, tasosartan, and ximelagatran).” The FDA report was designed to describe ways to distinguish drugs that have potential hepatotoxicity from those that do not. “Evidence of hepatocellular injury is a necessary, but not sufficient, signal of the potential to cause severe DILI.” A more specific signal of such potential is a higher rate of more marked peak AT elevations (10x-, 15xULN), with cases of increases to 4 1000 U/L causing increased concern. The single clearest (most specific) predictor found to date of a drug’s potential for severe hepatotoxicity, however, is the occurrence of a small number of cases of hepatocellular injury (aminotransferase elevation) accompanied by increased serum total bilirubin (TBL), not explained by any other cause, such as viral hepatitis or exposure to other hepatotoxins, and without evidence of cholestasis, together with an increased incidence of AT elevations in the overall trial population compared to control. Increased plasma prothrombin time, or its international normalized ratio (INR), a consequence of reduced hepatic production of Vitamin K-dependent clotting factors, is another potentially useful measure of liver function that might suggest the potential for severe liver injury. DILD is a relatively uncommon, but severe, cause of idiosyncratic liver damage that requires special consideration as a safety problem. There are approximately 2000 cases of acute liver failure each year in the U.S., with medications accounting for perhaps 25–50% of these.15 Overall, 39% are due to acetaminophen and 13% are idiosyncratic reactions due to other medications. DILD accounts for 2–5% of cases of patients hospitalized with jaundice, approximately 10% of all cases of acute hepatitis, and up to a quarter of all cases of chronic hepatitis. DILD has become the leading cause of acute liver failure among patients presenting for evaluation at liver transplant centers in the United States, and the leading single cause for having to remove approved drugs from the market. Following are some interesting facts concerning drug-induced liver disease (DILD):

DILD may be unaffected or exacerbated by any pre-existing liver diseases. Symptoms can be non-specific, such as nausea, fever, and rash.


61

On aggravation of symptoms, it can be very difficult to differentiate between the deterioration or complication of underlying liver disease and new or worsening DILD. Our understanding at any given time of the potential for new drugs to cause liver toxicity is based upon limited data from licensing trials. It is best advised that new drugs should be used under careful observation until patient data is available for large populations.

Epidemiology of drug hepatotoxicity The literature analysis of the epidemiology of drug-induced liver dysfunction/disease, presents a number of limitations. One such limiting factor involves the method used for the collection of the data. Currently, there are 2 steps in collecting data. Step 1 involves evaluating the new drug’s safety. This includes toxicological studies such as animal, cellular, or tissue models for the purpose of initial screening of the drug agent. Liver safety is also analyzed in healthy patients. This permits the presence of liver anomalies to be discovered. The number of clinical trial subjects range between 2000 and 5000.8 At a few thousand patients, only frequent events are detected. The majority of drugs have a hepatotoxic risk ranging from 1 in 10,000 to 1 in 100,000. For example, in the case of troglitazone, hepatotoxic risk was estimated to be 1 in 50,000, which explains the reason why liver dysfunction was undetected within the 5000 patients in clinical trials.16 For drugs such as antihistamines, hepatotoxicity is very negligibly low (less than 1 in 1,000,000), which makes DILD detection unlikely especially at an early stage as well as the exhibition of a patient susceptibility factor. In some cases, hepatotoxicity takes place years after the drug’s first use. For example, papaverine takes up to 40 years and amiodarone takes up to 25 years.8 Our current knowledge of the epidemiology of DILD is drawn from past studies, in which the number of events is analyzed within a certain period of time.3 There is limited data on the frequency of hepatotoxic drug reactions; however, most studies reflect the prevalence of drug hepatotoxicity in terms of the proportion of the population with this problem. It is quite rare to be given legitimate incident data reflected as number of cases per number of patients/year.2 There was a 10-year-long study within the Liver Unit of Beaujon Hospital in Paris that indicated that of all the patients with acute hepatitis, 10% were connected to drug toxicity.17 A prospective cohort study was performed on 17 tertiary care centers in the United States, and the results indicated that about 50% of the acute end-stage hepatic failure cases were due to medicinal agents. However, in the geriatric patients, medicinal agents accounted for 20% of jaundice and 25% of fulminant hepatic failure.3 The increasing number of drug agents transitioning into clinical trials largely contributes to the increasing presence of patients with DILD. There are over 1100 medicinal agents, herbal remedies, illicit substances, and environmental agents that have been proven to cause DILD. About 10% of mild to moderate liver biochemistry alteration cases are caused by medicinal agents.1 In patients who are older than 50 years of age, this number is increased to 40%. A French national survey that was performed in 1983 accumulated 980 cases of drug-induced hepatitis—63% were women and most of them were older than 50 years.18 Also, the Danish Board of Adverse Reactions to Drugs documented 572 cases of drug-induced hepatitis between 1968 and 1978.19 In patients with drug-induced hepatitis including jaundice, the risk for fulminant hepatic injury is 20% while the risk for patients with acute viral hepatitis and jaundice is 1%. Also, a risk of sub-fulminant hepatic injury is 70% for patients with drug-induced hepatitis with encephalopathy.20 The Boston Collaborative Drug Surveillance Program showed data that indicated 20 hepatotoxic reports out of 66,995 hospital admissions (1 in 3350) in 1981.21 A survey of 1999 hospital admissions to a Danish hospital indicated that 157 of these cases were drug-related with 3 of them being due to liver dysfunction.22 In the cases of general practice, drug hepatotoxicity is scarce, with a British survey showing an incidence of 0.3 cases in 100,000 patients.2

62


Approximately 10–15% of patients with end-stage liver failure are drug-related. Additionally, reporting systems indicate relative severity of drug hepatotoxicity. Epidemiological studies show that hepatotoxic reactions are between 1 in 600 and 1 in 3500, which means 2–3% of all hospital admissions are due to adverse drug reactions.2 Risk factors There are many different ways in which to combine, split, organize, or categorize DILD, these include (1) risk factors, (2) hepatotoxicity of specific medications, (3) hepatotoxicity symptoms (if any), (4) pathologic (liver biopsy) presentation, (5) whether the presentation was predictable (dose related) or unpredictable, and (6) extrahepatic/systemic manifestations (if any). Risk factors include gender, age, nutritional state, body mass index, presence of underlying diseases (diabetes, renal failure, infection with HIV, or hepatitis virus), and pre-existing liver disease.23 The presence and degree to which these risk factors correlate to DILD depends on the specific medication under consideration. Pre-disposing factors Individual traits of a patient such as genetics, gender, age, and alcoholic consumption may increase one’s susceptibility to drug-induced liver injury/dysfunction.

Gender: DILD is more common in females, although the cause for this is unknown. Race: Race can influence the ability of various drugs to be hepatotoxic. Isoniazid (INH), for example, may be more hepatotoxic in African Americans and Hispanics. Age: Liver toxicity is uncommon in children, although acetaminophen toxicity is a familiar exception. Decreased clearance in the liver, drug interactions, diminished hepatic blood flow, and several other factors seem to account for increasing hepatotoxicity with increased age.

In some cases, the dose or drug type plays a large role. For example, the amount of a single dose of acetaminophen or a cumulative dose of methotrexate may influence liver function. Pharmacokinetic and pharmacodynamic interactions also are significant contributors that amplify susceptibility. Illicit drugs such as cocaine, MDMA, or toxic mushrooms directly contribute to end-stage liver failure as well. Oral contraceptives have also emerged as a dosedependent risk factor in a dose-dependent manner as well as patients with diabetes, obesity, or hyperlipidemia. Because of their persistence, long-acting/delayed-release drug formulations may cause more injury than shorter-acting drugs. Generally, patients with mild or moderate liver dysfunction do not have a higher risk of idiosyncratic drug-induced liver disease with exceptions of methotrexate, troglitazone, trovafloxacin*, niacin, tolcapone, or pemoline.* Patients who exhibit severe liver disease (cirrhosis) have an altered metabolism which increases the risk of DILD. Despite this, some liver enzymes are unaffected by liver disease, especially those responsible for Phase II biotransformation reactions. Patients with cirrhosis also are susceptible to renal dysfunction, therefore drug agents that impact the kidney are to be avoided.3 Depending on the liver enzymes impacted by DILD and the specific drug, patients may have a reduced or altered drug-metabolizing capability. These cases have severe uncompensated liver disease with ascites or encephalopathy. The drug dose should be reduced in these patients to achieve normal drug concentrations. Several references24-28 state that patients with alcoholic liver disease are more sensitive to develop drug-induced hepatotoxicity. It is also unknown if this is due to the liver disease itself, drug interactions with alcohol, or other reasons such as malnutrition. Alcoholism or general excessive alcohol intake will more certainly worsen drug hepatotoxicity. *

Removed from the market due to hepatotoxicity


63

In the event that a specific drug will knowingly benefit the patient with liver disease, the dose is adjusted according to the level of impaired liver dysfunction. In this case, there is no reason to deprive the patient of beneficial treatment; however, a heightened vigilance should be given to patients with alcoholic liver disease. Genetic factors The genetic characteristics of the patient highly influence drug-induced hepatotoxicity. Expression of genes involved in the metabolism of medications can have a profound and idiosyncratic influence on DILD. Genetic polymorphisms may increase risk by causing metabolic abnormalities or altered expression of various liver cytochrome P450 enzymes. In a patient with decreased expression of CYP2D6, perhexiline hepatotoxicity becomes a higher risk. Greater than 75% of patients with perhexiline liver toxicity have deficient CYP2D6 activity. The standard assay for evaluating CYP2D6 activity is the rate of metabolism of debrisoquine, a substance hydroxylated by CYP2D6. “Fast acetylators” of isoniazid, for example, were initially believed to have a higher sensitivity to isoniazid hepatotoxicity than “slow acetylators.”29 However, several studies have disproved this, and it is now generally agreed that there is no difference in hepatotoxicity frequency between the slow and fast acetylators.30 Acquired factors

Chronic alcohol intake increases the severity of acetaminophen hepatotoxicity. This is caused by mechanisms involving the induction of Phase I liver enzymes, such as CYP2E which generates the toxic metabolite N-acetyl-p-quinone imine (NAPQI) from acetaminophen. Glutathione acts as an antioxidant or electrophilic scavenger. Due to glutathione depletion, the liver has a lower resistance to the toxic metabolites. Alcoholism also potentiates methotrexate toxicity. Alcohol favors CYP2E1 metabolism, which increases production of NAPQI, a toxic metabolite responsible for hepatocyte death. Chronic alcoholism, through malnourishment, may decrease the amount of glutathione (GSH), which inactivates NAPQI, and spares liver cells. When the substrate is diminished, hepatocyte death increases. A thorough alcohol use history should always be taken when evaluating possible DILD, and consideration should be given to the wisdom of treating some causes of liver disease (for example hepatitis B or C) if alcohol use/abuse is ongoing. Drug–drug interactions can play a role in hepatotoxicity. Enzyme increases can lead to the formation of toxic metabolites from 1 of the 2 drugs taken together. For example, the rifampin–isoniazid combination results in the facilitation of isoniazid biotranformation into toxic molecules. Similarly, enzyme induction by a drug such as phenobarbital can cause liver toxicity to anti-depressant drugs. Enzyme inhibition by drugs can also impact hepatotoxicity. For example, in the case of troleandomycin–estrogen drug–drug interaction, estrogen metabolism is inhibited by the blockade of CYP3A4 causing an accumulation of estrogen and resultant cholestatic effects.1 Extrahepatic disease also plays a role in drug-induced liver toxicity. For example, hyperthyroidism causes halothane hepatitis, and human immunodeficiency virus (HIV) increases the risk of adverse reactions such as cotrimoxazole hepatotoxicity. Other liver diseases might influence hepatotoxicity of drugs such as anti-retrovirals and methotrexate. Anti-retrovirals also can worsen dormant chronic hepatitis B and C by destabilizing or adversely modifying the immune system.31 Co-infection of HIV with hepatitis B or C can leave patients at increased risk for hepatotoxic effects when treated with antiretroviral therapy. Patients with cirrhosis have poor tolerance for hepatotoxic drugs. Patients with HIV can experience drug-related hepatotoxicity from HAART treatment. In addition to direct drug toxicity, mechanisms can include HIV viral hepatitis, co-infection with hepatitis B or C, and insulin resistance from HAART medications contributing to the development of steatohepatitis.75,76

64


Histopathology DILD does not usually present with a characteristic histopathologic picture, regardless of the causative medication. Although there are a number of distinctive pathologic categories of DILD,32 rarely is it possible to diagnose DILD from the histopathologic pattern, although causes sometimes can be ruled out. DILD can present with 1 of several pathologic categories: (1) (2) (3) (4) (5) (6) (7) (8) (9)

Impaired liver function Acute hepatocyte necrosis Fatty liver Granulomatous change Acute cholestasis Cholestatic type Chronic parenchymal injury Vascular change Liver tumor

In the presence of suspected DILD, the pathologic picture may help confirm the diagnosis. Knowing the distinctive pathologic category may or may not help in identifying a cause of liver disease, in that 1 drug can present with different pathologic states, and no pathologic picture seems to be pathognomic for a particular drug or toxin. It is clear that cirrhosis can follow a number of insults produced by drug intake. The drugs involved are those that lead to types of liver disease that in other circumstances also may progress to cirrhosis. Cirrhosis may be the end result of the prolonged administration of drugs that presumably act as direct toxins, such as methotrexate and isoniazid, or those that may be mediated by hypersensitivity, such as oxyphenisatin and methyldopa.33 Cirrhosis is most commonly caused by chronic, excessive alcohol use/abuse.

Predictability DILD can present in either a predictable (dose related) manner, or it can be unpredictable. Individuals with a history of adverse drug reactions are more likely to experience a reaction to another agent. This could be explained on the basis of genetic predisposition to immunoallergic responses such as penicillin or sulfonamide allergy. Alternatively, prior sensitization to 1 agent may confer an increased risk of liver injury after exposure to a chemically related compound. Examples of such cross-sensitivity include halothane with methoxyflurane and enflurane and the phenothiazines. DILD that presents in a predictable manner usually is due to direct hepatotoxicity, is reproducible in animals, tends to damage hepatic lobules, and can be induced by drug metabolites. Acetaminophen is an example of a drug that can cause predictable liver damage. DILD presenting in an unpredictable/idiosyncratic manner (not related to dosage), is typically not reproducible in animals. Examples of drugs known to cause unpredictable DILD include erythromycin, INH, halothane, and chlorpromazine. Unpredictable/idiosyncratic DILD can be either a hypersensitivity or immunoallergic reaction or a metabolic-idiosyncratic reaction. Hypersensitivity immune-related responses, as is seen for phenytoin, have a short latency and are characterized by fever, rash, and eosinophilia. A metabolic-idiosyncratic reaction, as is seen for INH, is caused by indirect metabolites of the offending drug. The response can occur within a week of exposure to the offending drug but can also take up to 1 year to manifest.


65

Clinical manifestations and diagnosis Although DILD may be accompanied by nausea, right upper quadrant abdominal pain, and elevated liver enzymes, DILD may not always present with these symptoms. In fact, DILD can have a widely varied clinical presentation, including the following associated extrahepatic/ systemic manifestations:

Fever, rash, and eosinophilia: chlorpromazine, halogenated anesthetic agents, sulindac, Dapsone (Sulfone syndrome), and anticonvulsants (anticonvulsant hypersensitivity syndrome) Obstructive jaundice: chlorpromazine, erythromycin, and amoxicillin–clavulanic acid Serum sickness: para-amino salicylate, phenytoin, and sulfonamides Muscular syndrome (myalgia, stiffness, weakness, and elevated creatine kinase level): clofibrate Anti-nuclear antibodies: procainamide Bone marrow injury: ribavirin, gold salts, propylthiouracil, chlorpromazine, chloramphenicol, and interferon Associated pulmonary injury: amiodarone and nitrofurantoin Associated renal injury: Gold salts, methoxyflurane, penicillamine, and NSAIDs Fatty liver of pregnancy: tetracycline Bland jaundice (pure cholestatis): contraceptive and anabolic steroids and rifampin Reye’s syndrome (acute encephalopathy, cerebral edema, and fatty infiltration of liver): aspirin Reye’s-like syndrome: sodium valproate

DILD can be asymptomatic, with only mild elevations of liver enzymes. Some helpful generalities concerning AST/ALT in DILD include the following:

In hepatocellular DILD, ALT levels are increased 4 2 ULN, and alkaline phosphatase (alk phos) levels are within the reference range or are minimally elevated. Elevation of AST 4 ALT, especially if more than 2 times greater, suggests alcoholic hepatitis. Elevation of AST o ALT is often observed in persons with viral hepatitis. In both viral and in drug-induced hepatitis, the AST and ALT levels steadily increase and can peak in the low thousands range within 7–14 days of injury. Some medications can cause marked increases in AST. In acetaminophen toxicity, for example, transaminase levels greater than 10,000 IU/L can occur. Normalization of mild liver enzyme elevation has been reported with continued use of some medicines, including NSAIDs and statins. Hepatic cytotoxicity via immune-mediated hypersensitivity reactions appears after weeks to months of exposure to a drug agent. Manifestation of these reactions may include rash, eosinophilia, lymphocytosis, fever, arthralgias, and liver-specialized antibodies, i.e., LKM (liver/kidney microsomal antibody). Blood AST (aspartate transaminase) and ALT (alanine transaminase) levels are mildly elevated; however, autoimmune serum markers (anti-nuclear antibody and anti-smooth-muscle antibody) and hyperglobinemia are also typical.3 Prominent signs of cholestatic injury include jaundice and pruritis. Cholestatis is a condition that is characterized by the decreased flow of bile from the liver to the small intestine (duodenum).34 Secondary signs include anorexia, fatigue, and malaise, but these are more present in chronic cholestasis. Jaundice is caused by hyperbilirubinemia, which is an increased amount of bilirubin (conjugated and unconjugated) in the blood.3

Diagnosis Drug-induced liver dysfunction is diagnosed by the detection of anomalies in the liver enzymes or by the emergence of hepatitis-like symptoms. End-stage liver failure is frequently the first appearance of acute liver injury. Patients might display symptoms such as pruritus, coagulopathy, or malabsorption of vitamins A, D, E, and K. Malnutrition also may contribute to liver dysfunction.3

66


Absolute diagnosis of DILD is usually determined by the deduction of possible causes, onset of action, the state of the symptoms after stopping the drug, and comparison to other reported cases. Liver biopsy can be helpful; however, even this mechanism of diagnosis can fail in determining specifically drug-induced liver dysfunction.35 To assess the causality of DILD, there are a number of methods one can employ using the following factors:

The timeline of drug administration including the time between the beginning treatment and end treatment—onset of drug action. The effects of stopping or continuing a drug and the role of drug and disease in the etiology when this happened. Patient response to the drug re-administration/re-challenge. Lab test results. Background knowledge of drug toxicity.

The main purpose of determining causality is to use quantitative lab results to determine the likelihood of liver toxicity.36 It is also important to know what happens when a drug is stopped or taken in combination with other drugs. It is the responsibility of the physician to keep up-todate with new additions to the known hepatotoxic drug list.34 Determining the cause of drug-induced hepatic injury becomes closely definitive once the above parameters are investigated and applied to specific patients’ cases. In rare circumstances, the cause of liver injury may only be probable or highly probable.77 For example, this is true in cases of drug overdose with acetaminophen, relapse after accidental carbamazepine readministration, and drug hepatitis in relation to iproniazid.1 More commonly, determination of the cause of liver injury is considered possible. In these cases, the liver injury has no distinct features, the timeline hints at the problem and some general causes of liver injury have been excluded. Far too often, the diagnosis of the cause of liver injury is doubtful or unlikely. This means that there are no specific features of hepatic injury and significant pieces of evidence involving timeline are not present. The degree of end-stage hepatic injury needs to be quantified to assess hepatic recovery following either liver transplantation or drug cessation due to poor outcome.1 An excluded diagnosis is evident when another cause has been established such as a viral infection or when the timeline of hepatic injury does not match up with drug administration. This is the case when the symptoms are present before the drug administration or if there is a delay of symptoms greater than 15 days after the end of treatment. Some exceptions to this include hepatitis caused by halothane, of which reactions can take place several weeks after the first exposure.1 Being able to recognize the development of hepatic injury is integral to the prevention of chronic perpetuation of cholestasis or other liver dysfunction. Efficient recognition and withdrawal of the drug will also prevent unnecessary damage-control measures. There are some drugs that do not show the true effects until much later, however. In the case of amoxicillin–clavulanate acid, the onset of liver disease may emerge weeks after the interval of therapeutic administration. Also, recovery after drug withdrawal may take months (3 or more) and in certain circumstances may never occur, for example with progression of vanishing bile duct syndrome (VBDS).34 Hyperbilirubinemia Some helpful generalities concerning bilirubin levels in DILD include the following: ● Normally, the total bilirubin level is less than 1.1 mg/dL and approximately 70% is indirect (unconjugated) bilirubin. – Unconjugated hyperbilirubinemia (4 80% of the total bilirubin is indirect) suggests hemolysis or Gilbert syndrome.


67

– Conjugated hyperbilirubinemia ( 450% of the total bilirubin is direct) suggests hepatocellular dysfunction or cholestasis. – Total bilirubin 425–30 mg/dL suggests that extrahepatic cholestasis is an unlikely diagnosis. The predominantly conjugated bilirubin is water soluble and is easily excreted by the kidney in extrahepatic cholestasis. ● Elevated aminotransferase, accompanied by elevated bilirubin, is suggestive of sub-fulminant or fulminant hepatic necrosis. ● With increasing hepatocellular injury, bilirubin levels are invariably increased, suggesting a worse prognosis.

Hepatic drug metabolism Human metabolism must deal with a variety of substances that, in excess, can cause disease. Most of the liver’s detoxification is performed on substances that are fat soluble, including drugs, certain vitamins, carcinogens, pesticides, and other environmental pollutants. Many endogenous chemicals such as fatty acids, prostaglandins, and sex hormones are also fat soluble and are metabolized by the liver. The duration of the physiological or toxicological action of these lipophilic substances depends on how long they reside intact in the liver. Other substances that are not fat soluble are also metabolized. Toxic intermediary metabolites may be formed by the liver after a conversion reaction that renders the chemicals into a fat-soluble state on which the liver can act. The cytochrome P450 (CYP450) system of enzymes is the primary enzyme system for metabolism of drugs. Hepatic microsomal P450s are the mixed-function oxidases (MFOs) that activate oxygen for incorporation into lipophilic substrates. Oxidized substrates are then either eliminated from the body or metabolized further by allied hepatic enzyme system. Cytochrome P450 This is the collective term for the family of hemo-proteins found in the hepatic endoplasmic reticulum. Each individual P450 exhibits a characteristic range of turnover numbers with specific substrates.39 Several hundred isoenzymes of CYP450 have been identified, and this is a very active area of research with significant relevance to drug metabolism. Nutritional effects on drug metabolism A person’s nutritional state can exert a major influence on his or her capacity to metabolize and eliminate drugs and foreign compounds. Chronic alcohol abuse or malnutrition and starvation may lead to a deficiency of glutathione, a substrate necessary for inactivating hepatic toxic substances. In addition to reducing hepatic glutathione, fasting can enhance activity of P450 2E1 (CYP2E1), which means acetaminophen is converted to its toxic metabolite at a faster rate. The extent to which each of these factors contributes to the increased risk of acetaminophen-induced hepatotoxicity among alcoholics remains unclear. An important patient variable that influences potential acetaminophen toxicity is the nutritional state—whether a patient has been eating a normal diet or fasting. Several reports have implicated fasting as a factor that enhances the prospect of hepatotoxicity associated with therapeutic doses of acetaminophen. Yet, it is reasonable to expect that an individual who takes acetaminophen for symptoms of a cold, for example, or for other minor illness may not be hungry. That person’s decreased food and fluid intake could therefore potentially exacerbate acetaminophen toxicity. Consider also that nausea is one of the early symptoms of acetaminophen-induced toxicity. Patients who feel nauseated typically stop eating and drinking—further enhancing the very reason that made them stop eating and drinking: acetaminophen toxicity. This would further exacerbate the acetaminophen toxicity.

68


Biochemical mechanisms Drugs and toxins can produce liver damage by at least 4 general types of mechanisms: (1) The drug directly impairs the structural and functional integrity of the liver. (2) Metabolism of the drug produces a metabolite, usually an oxidizing or alkylating species, that alters hepatocellular structure and function. (3) A drug metabolite binds to hepatic proteins to produce new antigenic determinants that become the targets of specific immune responses. (4) The drug initiates a systemic hypersensitivity response (drug allergy) that damages the liver. Idiosyncratic liver injury can be the product of several different mechanisms that progress independently, concurrently, or serially. Hepatitis (hepatocellular or cytotoxic injury) is possibly due to free radical formation, reactive oxygen species (ROS) or electrophiles that bind to hepatic proteins during biotransformation and impede ionic transport, and cellular membrane integrity and adenosine triphosphate (ATP) production essential for hepatocellular energy.3 These mechanisms may lead to hepatic death via apoptosis or necrosis. Necrosis involves cellular swelling/lysis due to mitochondrial dysfunction and ATP depletion. Also, an increased concentration of calcium in the cell can activate hydrolases such as proteases, nucleases, and phospholipases, thereby inducing cellular membrane breakdown and inflammation. Apoptosis involves programmed cell death distinguished by cellular shrinkage, nuclear disassembly, and cellular fragmentation. Both necrosis and apoptosis can be either zonal or non-zonal. Zonal injury is limited to specific areas of the liver within the hepatic acinus. This is postulated by the concentration of the hepatic enzyme responsible for metabolizing the drug agent. Non-zonal injury is diffusive and causes the hepatic lobules to collapse. In this case, the liver structure essentially disappears. Most idiosyncratic cases result in non-zonal injury.3 Hepatic cytotoxicity may also be the product of an increase in metabolized xenobiotic. This metabolized xenobiotic is a toxic adduct that can migrate to the hepatocyte membrane to induce an immune reaction. For example, phenytoin and phenobarbital induce CYP450 in the liver and therefore decrease the threshold for liver damage.37 This is because there is an accumulation of toxic metabolites compounded by a depletion of glutathione. Some drug compounds activate macrophages that cause the formation of granulomas or produce fibrosis. Hepatic toxicity that is immune-derived results in cell death by apoptosis and zonal injury and is characterized by regions of inflammation wrapped in fibrous strands.3 In addition to necrotic or apoptotic liver cell death, hepatotoxicity reactions may cause a condition called steatosis. Steatosis is an abnormality where fatty acids are present in the mitochondria. It is caused by the inability of triglycerides to couple with hepatic transport proteins, hence movement of fat from the liver is impeded. Phospholipidosis is characterized by the collection and entrapment of phospholipids in the lysosomes of hepatocytes caused by drug agents.38 Direct liver injury to the cells lining the bile ducts can cause cholangiodestructive cholestasis and is defined by bile duct destruction or vanishing bile duct syndrome.40 Combination hepato-cholestatic injury is the product of a delayed hypersensitivity or is considered an immune reaction with portal vein inflammation or failure of hepatic pumps. The failure of these pumps causes toxic bile acids to accumulate, thereby causing secondary liver injury.3 The drugs that are currently known to cause serious DILD generally are widely prescribed and have a high margin of safety. They include agents such as NSAIDs, antibiotics, anticonvulsants, newer antihypertensive drugs, H2-receptor blockers, and psychotropic agents. For these types of drugs, the absolute incidence is in the order of 1 adverse hepatic drug reaction per 100,000 person years of exposure. Because of the very large number of prescriptions filled for these medications, they are among the most common causes of DILD. Acetaminophen continues to be a drug with widespread prescription and OTC use; however, acetaminophen presents its own set


69

of challenges to the toxicological community. Dose-related toxicity can lead to fatal hepatic failure, and the drug continues to be widely used in suicide attempts. Even when given at therapeutic doses, acetaminophen can still cause hepatotoxicity.

FDA and DILD FDA approval does not ensure that a drug is not hepatotoxic or that DILD will not occur. Registration trials for pharmaceuticals focus on efficacy. Safety is a secondary endpoint. Study sample sizes are typically small, and if the incidence of DILD is low, then it may not become apparent until post-licensure, which is often collected by spontaneously reported postmarketing data. Some drugs in current or recent use, including many familiar examples, but certainly not all drugs capable of causing DILD, have been discussed in this review. The FDA has taken a number of well-publicized actions concerning hepatotoxic medications in recent years, including strengthening of warnings, restrictions on use, and removal from marketing. Examples include glimepiride, interferon beta 1a, trimethoprim/ sulfamethoxazole, pemoline, bromfenac, duloxetine, felbamate, kava kava, troglitazone, tolcapone, trovafloxacin, and zileuton. When the hepatotoxicity of a new drug is particularly common or severe, the agent may be withdrawn by the manufacturer usually with prompting by the FDA; benoxaprofen, bromfenac, and oxyphenisatin (in most countries) are examples. Other drugs, such as chlorpromazine, are still used because their favorable qualities continue to outweigh the small risk of drug-induced liver disease. Recent history of drug withdrawals is presented in Table 1.

Types of liver injury Liver injury caused by therapeutic drugs, though uncommon, is on the rise. Drug-induced liver injury can mimic many different types of liver disease. These conditions include cholestasis, steatosis, granulomas, acute and chronic hepatitis, cirrhosis, vascular disorders, and tumors. When treating a patient with liver injury, the physician must obtain an accurate drug history from the patient.23 Liver injury, or liver dysfunction, occurs when there is an increase of greater than 2 times the normal range in ALT or CB or a total increase in AST, AP (alkaline phosphatase), and TB. There is no other biochemical exam that is this specific to liver dysfunction.36 These alterations in enzymes are important because of the great capacity of the liver to heal injury, with the subsequent development of adaptive tolerance, as frequently seen with initial exposure to drugs such as isoniazid and tacrine. Tests reflecting liver injury alone do not necessarily predict or indicate serious hepatotoxicity. Vague symptoms such as fatigue, anorexia, nausea, discomfort in the right upper quadrant, and dark urine may be the first clues that hepatotoxicity has occurred. Drug-related hepatotoxicity should be considered when such symptoms occur in conjunction with biochemical evidence of liver injury, and especially with concurrent impaired liver function. The regulation of serum enzyme activity is not a function of the liver. Liver function is more accurately assessed according to the levels of total bilirubin or conjugated bilirubin—reflecting the liver’s ability to move bilirubin from plasma into bile. Another measurable liver function is protein synthesis, which is reflected in the albumin concentration and the prothrombin time [or its international normalized ratio (INR)].9 Additionally, liver injury is termed “hepatocellular” when there is a 2-fold increase in ALT alone. This is also the case when the ratio (R) of the serum activity of ALT to AP R is greater than or equal to 5. Liver dysfunction is termed “cholestatic” when there is a 2-fold increase in AP or when R is less than or equal to 2. Finally, liver dysfunction is considered “mixed” when both the ALT and AP are increased. The R value in this case must be between 2 and 5.36

70


We can apply these parameters in order to understand the difference between acute liver injury and chronic liver injury. Acute liver injury is considered to be the case when the above increases have lasted shorter than 3 months. Chronic liver injury, however, occurs when these increases have lasted more than 3 months. In the case of severe liver injury, additional manifestations couple with the increases in liver enzymes. These manifestations include jaundice, prothrombin time prolongation, and hepatic encephalopathy. Finally, the designation fulminant (end-stage) liver injury is used when there is rapid progression (days to weeks) of hepatic encephalopathy and coagulopathy.36 Acute hepatic injury is characterized by an ALT above twice the normal concentration and generally shows indistinct clinical manifestations and appears identical to acute viral hepatitis. This might be linked to overdosing, an individualistic drug reaction, drug sensitivity, or autoimmune causes. This is the most prevalent kind of liver damage induced by drugs. Hundreds of drug agents can cause acute hepatic injury including herbal medications, excipients (inert additives of pills), cocaine, and amphetamine derivatives.1 Approximately one-quarter of acute hepatic failure cases are induced by drug administration. It is estimated that 2% of jaundiced hospitalized patients are there due to drug toxicity. In geriatric patients, the incidence is ten times higher. A study in Japan reported that the number of cases of liver disease reported was 10 times greater in the decade 1964–1973 than in the previous decade. Limited progress has been made in confirming the reasons for different sensitivities to drug toxicity. Terms such as “hypersensitivity” and “idiosyncrasy” have not given a specific understanding of the metabolic mechanisms of liver injury. In some cases, reactive drug metabolites are the obvious culprit; however, the reason behind the variability of incidence and hepatic injury severity has not been discovered. From a research standpoint, hepatic lesions that are found in patients exposed to drugs are not reproduced experimentally. There are many cases of animals in research having a hypersensitivity response by inducing drug-induced hepatic injury with halothane, for example. However, true cholestasis and hepatitis is not found in animals after similar drug challenges.23 Fatty liver disease Fatty liver disease is quite common and is mostly due to obesity and alcohol abuse. Fatty liver disease occurs when large drops of fat fill the cytoplasm of the hepatocyte and displace the nucleus to the periphery of the hepatocyte. Poisons that man is environmentally exposed to may also cause a similar type of fatty liver without extensive necrosis (phosphorus) or zonal necrosis (carbon tetrachloride).23 Therapeutic drug doses theoretically should not cause hepatic steatosis, but the prolonged administration of antimetabolites such as methotrexate can lead to fatty liver coupled with more severe aspects of liver disease such as hepatic necrosis, steatohepatitis, and fibrosis. Steatosis and severe liver disease without necrosis may be produced rapidly by significant doses of tetracycline.23 Refer to Table 2 for drugs that cause fatty liver disease.82 This is a partial list of agents that produce fatty liver. Some of these drugs cause inflammation also. The association of fatty liver with calcium-channel blockers is weak; however, the association of this condition with amiodarone is strong. Drug-induced fatty liver may not have Table 2 Drugs that cause fatty liver disease. Glucocorticoids

Synthetic estrogens

Aspirin

Calcium-channel blockers Tetracycline Valproic acid

Amiodarone Methotrexate Cocaine

Tamoxifen Perhexiline maleate Antiviral agents


71

sequelae (cases caused by glucocorticoids) or may result in cirrhosis (caused by methotrexate and amiodarone).39

Hepatitis The entire spectrum of liver injury seen in conjunction with viral hepatitis, both morphological and clinical, can also result from drug administration. Hepatic reactions are unpredictable with few exceptions. In contrast to drug-induced cholestasis, which is relatively common and benign, hepatitic (inflammatory) reactions are uncommon and have significant mortality.23 In various stages of hepatitis, distinguishing a drug-induced disease from a viral pathogen will continue to depend on evidence from historical and epidemiological data.23

Massive hepatic necrosis (fulminant hepatitis) Massive hepatic necrosis is a rare complication of viral hepatitis that involves the destruction of the entire liver lobule. This form of injury can be accurately mimicked by drugs such as halothane, iproniazid, and isoniazid. When exposed to these agents, hepatocytes within entire hepatic lobules appear to progress through coagulative necrosis. The lesion may begin in the central zones of the lobule and progress to include the entire lobule.23 Both the clinical and morphological sides of liver disease can be generated by a number of drugs, many of which can produce massive hepatic necrosis. Some of these drugs include halothane, isoniazid, phenytoin, phenelzine, sertraline, naproxen, diclofenac, kava kava, or ecstasy and iproniazid. The presence of eosinophilia may be indicative of drug-induced hepatitis; however, this only occurs in less than half the patients. Upon drug cessation, patients generally recover, however a second exposure might manifest massive liver necrosis.23

Chronic active hepatitis Prolonged administration of a number of drugs, such as the laxative oxyphenisatin and isoniazid, might produce both clinical and histological manifestations of chronic active hepatitis. Withdrawal of the drug generally, but not consistently, ameliorates the liver disease and associated disorders.23

Chronic persistent hepatitis The clinical and histological features of chronic persistent hepatitis may be produced by drugs that are associated with chronic liver injury such as oxyphenisatin, alpha-methyldopa, and isoniazid.23

Cirrhosis The drugs that are attributed to cirrhosis are those that lead to types of liver disease which in other situations also might progress to cirrhosis. These types of pre-cirrhosis liver diseases include chronic active hepatitis, subacute hepatic necrosis, and healed massive hepatic necrosis. Cirrhosis can be the end result of long-term administration of drugs that act as direct hepatic toxins such as methotrexate and isoniazid or drugs that are mediated by hypersensitivity such as oxyphenisatin and alpha-methyldopa. It is also possible that cases of cryptogenic cirrhosis, which is cirrhosis with unknown etiology, may indicate drug-induced injury.23

72


Vascular disorders—Budd–Chiari syndrome One of the most common causes of the Budd–Chiari syndrome (occlusion of the hepatic veins) is the use of oral contraceptives. The mechanism is not well established, but it is most likely similar to the mechanism responsible for other thromboembolic phenomena connected to oral contraceptives and is more closely linked to estrogenic than to progestational effects. The first lesion in the liver is probably an endophlebitis of the hepatic veins, which can lead to secondary thrombosis. Occlusion of the hepatic veins leads to congestion, hemorrhage, necrosis, and fibrosis. The classical Budd–Chiari triad is abdominal pain, ascites, and hepatomegaly. Congestive cirrhosis may also become evident if the patient lives long enough.23 Peliosis hepatitis and hepatocellular carcinoma Peliosis is characterized by irregular, blood-filled spaces in the liver that are lined by hepatocytes. The blood-filled spaces freely communicate with the hepatic sinusoids. Initially this lesion was associated with tuberculosis and tumors, but now it is attributed to drug administration such as sex steroids for both males and females. The primary culprits include anabolic steroids such as oxymethalone and methyl testosterone and oral contraceptives especially with large quantities of estrogen (Rubin, 1980). There are also a number of reports of hepatocellular carcinoma due to the usage of anabolic steroids and the oral administration of contraceptives.23 Drug-induced cholestasis Cholestatis is a syndrome resulting in reduced bile flow. Mechanisms are either (1) hepatocellular, where an impairment of bile formation occurs, or (2) obstructive, where impedance to bile flow occurs after it is formed. The typical histopathologic features of hepatocellular cholestasis include the presence of bile within hepatocytes and canalicular spaces, in association with generalized cholestatic injury. Obstructive cholestasis is characterized by bile plugging of the interlobular bile ducts, portal expansion, and bile duct proliferation in association with centrilobular cholestatic injury.41 To clinically identify cholestasis, there has to be combination of liver enzyme changes in the blood along with a histological assessment. Enzymatic changes include such observances as an increase in serum alkaline phosphatase SAP (2-fold) coupled with a normal level of aminotransferase (AT) or an increase in both ALT and SAP with a ratio less than 2:1. In cases where a 3-fold SAP and 5–8-fold AT increase are seen, a biochemical cholestasis diagnosis is likely. With this said however, none of the measures is absolutely fool-proof.35 In both hepatocellular (liver swelling, which compresses the intrahepatic channels) and in obstructive cholestatic disease, the bilirubin is elevated. Hepatocellular injury is often accompanied by aminotransferase 4 500 and alkaline phosphatase level normal to 3 ULN. In obstructive injury, the aminotransferase is normal to mildly elevated and alkaline phosphatase can be up to 4 ULN.40 Biopsies of the liver may be performed, but not in every case of presumed drug-induced cholestasis. The histological assessment of the liver may prove to be quite helpful in determining the cause of liver dysfunction and suggesting possible drug culprits. Biopsies also can prove useful in developing a patient’s prognosis and determining the possible progression of chronic cholestasis.41 There are number of drug agents that have been known for many years to cause liver cholestasis. These drugs include estrogens and anabolic steroids, chlorpromazine, erythromycin, and oxypenicillins. Similar drugs such as tamoxifen may also bring about cholestasis. Sulindac or octreotide can lead to extrahepatic cholestasis secondary to biliary sludge or calculi. An obstructive cholestasis can result from vanishing bile duct syndrome,40 which has been associated with chlorpromazine, flucloxacillin, amitriptyline, and trimethoprim-sulfamethoxazole. Modern drugs


73

such as ticlopidine, statins, and some herbal remedies have also been associated with cholestatic liver dysfunction.34 Treatment for drug-induced cholestasis is “supportive.” This means that the drug presumably causing the dysfunction should be withdrawn from the treatment regimen promptly to prevent the progression of the hepatic dysfunction. Drugs that are used to ameliorate drug-induced hepatic damage include cholestyramine, which alleviates pruritus by binding bilirubin in the blood and facilitating excretion, and rifampin or opioid antagonists, which are administered when the first-line therapy fails. Additionally, nutritional support is instilled for patients who have an extended cholestatic situation, which puts them at risk for biliary cirrhosis and liver failure. For patients with prolonged cholestasis, a liver transplant may be necessary.34 Cholestasis with bile duct injury Drugs that induce cholestatic hepatitis can also cause cholestasis with bile duct injury. These drugs include flucloxacillin, chlorpromazine, and carbamazepine. There are some drugs such as dextropropoxyphene, carbamazepine, and methylenedianiline that are also associated with jaundice and hepatic damage in food contamination and in certain industrial settings.34 Cholestasis with hepatitis Hepatitis can complicate the cholestatic condition. There are diverse degrees of hepatic parenchymal and portal vein inflammation that add to the morbidity of cholestasis. Symptoms usually emerge within 6 weeks after the introduction of the drug to the patient’s system. Influenza-like symptoms begin the condition, followed by anorexia, vomiting, right upper quadrant pain, jaundice, and pruritus. Pruritus in particular emerges in 70% of these patients. Clinically, this may present itself similar to acute cholangitis. An increase in serum AP (3N or more) and a moderate increase in AT (2–5N) is also prevalent. Termination of the cholestatic progression occurs within 3 months after drug cessation; however, this may take longer with certain drugs such as chlorpromazine or oxypenicillins.34 Pure cholestasis consists of pruritus and jaundice. Liver transaminases can be slightly increased but can also have normal levels. This is seen with drugs such as sex steroid derivatives, cytarabine, and azathioprine. ACH is defined by cholestasis coupled with pain and hepatic tenderness, and often hypersensitivity reactions. This form of hepatitis is the second most common type of liver injury and is connected to several hundred agents.8 Drug-induced vanishing bile duct syndrome Case report A 37-year-old black woman (Ms. W) presented with difficult-to-control diabetes (DM), morbid obesity, hypertension, and chronic skin abscesses. During admission for control of skin abscesses, she received trimethoprimsulfamethoxazole (TMP-SMZ), which was continued for 30 days after discharge. She then required admission for UTI and was treated with antibiotics. Metformin and glimepiride were added at the time of discharge, as parenteral insulin was poorly controlling the DM. Shortly thereafter she developed a cholestatic drug reaction, which progressed to end-stage liver disease (ESLD) with the histologic features of vanishing bile duct syndrome (VBDS). Ms. W’s initial clinical course was characteristic for medication-induced cholestatic liver disease. Typically, slow resolution should occur with discontinuation of the offending drug. When recovery did not occur as expected, the presence of VBDS was identified, leading to liver transplant.40 Initially, this patient’s workup seemed to indicate a hepatocellular mechanism for the jaundice, as the cause was felt to be drug-induced. She had been started on both TMP-SMZ and glimepiride around the same time several months before her problem manifested clinically,

74

J.T. O’Donnell et al. / Disease-a-Month 60 (2014) 55–104 Table 3 Drugs in current use that have been reported in association with the VBDS. Amoxicillin–clavulanic acid Ampicillin Amitryptiline and imipramine Azathioprine Barbiturates Carbamezipine and phenytoin Chlorthiazide Cotrimoxazole Clindamycin Chlorpromazine Cimetidine Cyproheptadine Erythromycin esters

Estradiol and norendrostenolone Flucloxacillin and dicloxacillin Glycyrrhizin Haloperidol Ibuprofen D-penicillamine Prochlorperazine Tetracycline Terbinafine Thiabendazole Tiopronin Tolbutamide Methyl testosterone

and both have been reported associated with cholestatic jaundice. VBDS is a rare disorder, affecting the intrahepatic bile ducts. Most patients are asymptomatic, but some may present with pruritus and, rarely, jaundice. The alkaline phosphatase level usually is elevated, along with GGT, which may exceed 600 IU/L. A wedge biopsy of the liver often is necessary to make the diagnosis. On physical examination, frontal bossing and triangular facies may be noted, and additional tests can reveal butterfly vertebrae and posterior embryotoxon of the eye, but these findings were not present for this patient. Alagille syndrome and progressive familial intrahepatic cholestasis can also lead to VBDS, as can cystic fibrosis, systemic mastocytosis, histiocytosis-X, and Hodgkin disease, but some cases remain idiopathic. VBDS has been reported in association with chlorpromazine, flucloxacillin, amitriptyline, and TMP-SMZ. More than 30 drugs have been shown to cause VBDS (Table 3). Both TMP-SMZ and glimepiride had a temporal association with this patient’s cholestatic jaundice. TMP-SMZ but not glimepiride has been reported to be causally related to VBDS. Using standard accepted methods of determining causation (Marks, 2012) TMP-SMZ, the common denominator for both the cholestatic jaundice and subsequent VBDS, was probably causative. Ms. W presented with medication-induced cholestatic jaundice which progressed to VBDS. Although the biochemical markers pointed to an obstructive etiology, in this case the cholestatis was drug-induced (TMP-SMZ) but of an obstructive nature. Physicians prescribing the commonly used medications described above should always consider the possibility of medication-induced liver disease when liver enzymes become abnormal, and immediately withdraw any potentially offending agents if necessary.40 The clinical and biochemical characteristics of VBDS are similar to PBC (primary biliary cirrhosis); however, antimitochondrial antibodies that are present in PBC are not detectable in VBDS. The pathological progression of this condition is not known. Potential mechanisms include immune-mediated bile duct damage indicated by the presence of white blood cells and Stevens– Johnson syndrome in some circumstances. Another mechanism includes the presence of toxicity from secondary bile acid retention.34

Drug-induced sclerosing cholangitis Floxuridine-induced sclerosing cholangitis emerges in many metastatic colorectal carcinoma cases. Intra-arterial floxuridine infusions are also used in carcinoid tumors within the liver. Biliary strictures are also a complication. The common hepatic, right and left hepatic ducts are affected and the lesions resemble primary sclerosing cholangitis.34


75

Other types of liver injury Drugs cause less than 1% of chronic hepatitis and liver cirrhosis cases, but for rare hepatic lesions, drugs remain a significant contributor. For example, estrogens may cause hepatic adenoma, and thiopurines contribute to hepatic vascular disorders (nodular regenerative hyperplasia and peliosis hepatitis).4

Specific DILD product areas and case discussions Actos hepatic effects There have been post-marketing reports of fatal and non-fatal hepatic failure in patients taking Actos (pioglitazone hydrochloride), although the reports contain insufficient information necessary to establish the probable cause. There has been no evidence of drug-induced hepatotoxicity in the Actos controlled clinical trial database to date. Patients with type 2 diabetes may have fatty liver disease or cardiac disease with episodic congestive heart failure, both of which may cause liver test abnormalities, and they may also have other forms of liver disease, many of which can be treated or managed. Therefore, obtaining a liver test panel [serum ALT AST, alkaline phosphatase, and total bilirubin] and assessing the patient is recommended before initiating Actos therapy. In patients with abnormal liver tests, Actos should be initiated with caution. Measure liver tests promptly in patients who report symptoms that may indicate liver injury, including fatigue, anorexia, right upper abdominal discomfort, dark urine, or jaundice. In this clinical context, if the patient is found to have abnormal liver tests (ALT greater than 3 times the upper limit of the reference range), Actos treatment should be interrupted and investigation done to establish the probable cause. Actos should not be restarted in these patients without another explanation for the liver test abnormalities. Patients who have serum ALT greater than 3 times the reference range with serum total bilirubin greater than 2 times the reference range without alternative etiologies are at risk for severe drug-induced liver injury and should not be restarted on Actos. For patients with lesser elevations of serum ALT or bilirubin and with an alternate probable cause, treatment with Actos can be used with caution. Several mechanisms of drug-induced toxicity have been discussed, most of which demonstrate different toxic mechanisms, involve special age groups, and may have resulted in removal or threats of removal of drugs from the market. This is certainly demonstrative of the serious nature of drug-induced liver toxicity. Many case series of DILD have involved litigation following some serious injury or fatality related to a drug exposure.

Non-steroidal anti-inflammatory drugs (NSAIDs) use and hepatoxicity The class termed non-steroidal anti-inflammatory drugs (NSAIDs) includes anthranilic and arylalkanoic acid derivatives. Serious forms of hepatotoxicity are not currently a common problem with this group of drugs, but drugs such as ibufenac and benoxaprofen had to be withdrawn for this reason. Literature reports, extent of use, and patterns of clinical, biochemical, and pathological anomalies have shown that analgesics of this class seem to have a greater chance of hepatotoxicity; however, liver damage is not a dominant feature of acute NSAID overdose.42 NSAIDs have been associated with rare adverse reactions in the liver, including fulminant hepatitis and cholestasis.43 These reactions are idiosyncratic, mostly independent of the dose administered and are host-dependent. Among the NSAIDs (excluding acetaminophen), a rank order of relative risk cannot be established, and the incidence in relation to use is not known.42

76


NSAIDs have anti-inflammatory effects along with inhibition of cyclooxygenase and therefore prostaglandin formation. These drugs are heterogenous when it comes to adverse effects on the liver. A few NSAIDs are thought to cause drug-induced hepatitis and nearly all NSAIDs are implicated in case-reported liver disease. The risk for NSAID users who also are in contact with other potential hepatotoxic drugs has been shown to be at a much greater risk than from additively combining the risks for NSAID and other drug exposure individually, suggesting a synergistic toxicity of combining NSAIDs with other hepatotoxic drugs.71 The mechanisms that are responsible for NSAID-induced hepatotoxicity remain poorly understood and have mostly been diagnosed from clinical manifestations. Some emerging evidence indicates that many acidic NSAIDs are metabolized to reactive acyl glucuronides which form covalent adducts with plasma proteins and hepatocellular proteins. In vitro research utilizing hepatocytes co-cultured with lymphocytes indicates that NSAID-altered proteins can become antigenic.43 NSAIDs have been known to cause both hepatocellular and cholestatic injury. Cholestatic hepatitis in particular is a product of sulindac toxicity and is predominantly in women older than 50 years of age. Other NSAIDs such as ibuprofen have been associated with VBDS in both children and adults.36 Incidences of fulminant hepatitis are not known because not all are reported and because the association with a specific liver injury might be difficult to prove. Clinically manifested hepatic toxicity connected with diclofenac, naproxen, or piroxicam ranges is estimated to be 0.05%.43 Idiosyncratic hepatotoxic injury is a major clinical concern because a prediction cannot be made based upon preclinical safety studies and because the reactions are host-dependent. Even though the incidence of such injury is low, the resulting injury can be quite significant, sometimes resulting in fulminant liver failure.43 An emphasis is placed on using minimal analgesic, rather than anti-inflammatory doses of short-acting NSAIDs and, where possible, avoiding their use in high-risk patients.44

Reye’s syndrome Young patients are especially sensitive to the effects of salicylates and have been reported to have severe to fatal liver damage as a result of being treated with aspirin or other salicylates. In some cases, this has been associated with acidosis, hypoglycemia and hyperammonemia, and encephalopathy with cerebral edema. These effects resemble Reye’s syndrome, which is a pediatric condition with a high mortality rate. Reye’s syndrome is usually caused by the combination of an illness resembling the flu or varicella and the use of salicylates.42 Diclofenac The nature of diclofenac-related liver injury is idiosyncratic based on the varying onset of injury after long periods of drug administration. Delayed recurrence of injury after readministration of the drug indicates that metabolic abnormalities may be at play. This postulation is supported by the lack of fever, rash, or eosinophilia in most reported cases. Age appears to be a factor in susceptibility to diclofenac-induced hepatic injury similar to that associated with other NSAIDs. Patients older than 60 years of age make up a greater proportion of those with liver injury compared to drug users in general. Also, female patients have a higher prevalence of DILD, making up two-thirds of reported cases related to dilofenac.43 Of the marketed NSAIDs, diclofenac has been identified to have a greater potential for hepatotoxicity.

Naproxen Naproxen has been the culprit in a few cases of liver injury. Hepatocellular jaundice, fulminant hepatic failure, cholestatic jaundice, and indeterminate jaundice have been associated with naproxen.47


77

Fig. 1. Distribution of reported patient outcomes within the FDA Adverse Event Reporting System where liver injury is co-reported with sulindac as a suspect medication.

Sulindac Sulindac was selected for a study because several hundred cases of hepatic injury that had been reported to the FDA’s Adverse Reaction Voluntary Reporting System between 1978 and 1986 had a large number associated with sulindac. The resulting report contained analysis of clinical, biochemical, and histological characteristics of hepatic injury and postulates mechanisms of liver injury.45 Hypersensitivity markers such as fever, rash, or eosinophilia were reported in 35–55% of patients. This was most connected to patients with a short latent period (less than 4 weeks). Even in patients with a latent period of greater than 8 weeks, the incidence of fever and rash was greater than 20%.45

Fig. 2. Top 25 adverse reactions co-reported with liver injury when sulindac is a suspect medication.

78


Fig. 3. Age distribution of patients within the FDA AERS where liver injury is co-reported with sulindac as a suspect medication.

Liver sections were taken from 15 patients. The liver injury was mostly hepatocellular in 9 patients and cholestatic in 6 patients. Hepatocellular injury contained spotty lesions and panacinar lesions in 8 patients and zonal lesions in 1 patient. In 4 of the 8, there was spotty necrosis with mild injury characterized by scattered ballooning, apoptosis, and focal necrosis. Portal inflammation was minimal with a high presence of eosinophils in 2 of the 4 cases.45 From 338 reports that were submitted to the FDA, 247 were inadequate or ambiguous for sulindac toxicity. Overall, 91 cases of drug reaction were analyzed and 15 had histological data available. The results of this study indicated 4 deaths—3 with severe hypersensitivity and 1 with fulminant hepatic failure. Two-thirds had hypersensitivity reactions. Female to male ratio was 3.5 to 1; 69% of the patients were over 50 years of age. Of the patients, 67% had jaundice, 43% had cholestasis, and 25% had hepatocellular injury. Eosinophilia was more frequent in patients with cholestatic injury (40%) than in those with cholestasis (Figs. 1–3). Previous reports and “package inserts” indicate that hypersensitivity is the most common mechanism for sulindac-induced hepatic injury. Some reactions to sulindac are mediated through other mechanisms. For example, metabolic rather than immunological idiosyncrasy is suggested by hepatic injury not connected to hypersensitivity. These cases also appear after longer drug exposure periods.45 In conclusion, sulindac hepatic injury has a higher prevalence in females rather than males. This injury can lead to cholestatic or hepatocellular injury mostly due to immunological sources. In some patients, the mechanism may be due to metabolic idiosyncrasy. Overall, this study exposes the practicality and significant purpose to the analysis of adverse reaction reports when characterizing drug-induced liver injury.45 Other NSAIDs Occasional reports have been found of liver damage or abnormal liver function tests in patients who have taken niflumic acid, ibuprofen, naproxen, flurbiprofen, and fenbufen. With more recent drugs, there is simply not enough information to determine a specific clinical syndrome. With older agents such as ibuprofen, liver damage is uncommon with clinical manifestation being variable.3 Isoniazid (INH)-induced hepatotoxicity Introduction INH is an important hepatotoxin, especially with the resurgence of tuberculosis and the public health clinics’ protocols for prophylactic use in exposed persons. Close monitoring of patients for liver enzyme elevations and signs of fatigue and jaundice can detect liver toxicities and save the patient from developing an end-stage liver disease. When isoniazid (isonicotinic acid hydrazide) was introduced in 1952, it represented a major advance in the treatment of tuberculosis. It has remained the mainstay of antituberculous


79

chemotherapy. Isoniazid is potent, bactericidal, and has good clinical efficacy. In addition, it is inexpensive and can be administered orally. Initially, isoniazid was thought to have a low incidence of side effects. This apparent safety, together with the high level of patient acceptability, led to isoniazid becoming extremely popular during the 1960s. This included its use as a chemoprophylactic agent for tuberculin-reactive individuals, such as household contacts of patients with recently diagnosed tuberculosis, who were thought to be at heightened risk of active tuberculosis. Studies from the USPHS and others showed that between 10% and 36% of patients have asymptomatic elevations of ALT levels during isoniazid therapy. In most patients, these abnormalities occur within 10 weeks, are relatively minor, and resolve spontaneously. They are often associated with trivial histological changes in the liver. The frequency of clinically significant hepatotoxicity is about 1% of all patients exposed. There is, however, a striking age dependence, with isoniazid hepatitis being rare in childhood and occurring in more than 2% of adults over the age of 50 years. Several observations indicate that the longer the period of isoniazid treatment prior to the recognition of hepatotoxicity, the more severe the liver injury at the time of diagnosis. Isoniazid hepatitis appears to be more severe in African-American women, in those who habitually excessively drink alcohol, and in those treated with combined therapy with rifampicin. Deaths still occur from isoniazid-induced liver injury. Most, if not all, deaths due to isoniazidinduced liver injury could be prevented by immediate interruption of isoniazid treatment at the onset of the first symptoms of hepatotoxicity. The role of monitoring liver enzymes during the first 3 months of isoniazid therapy is less clear, although it may indirectly serve a useful function by keeping the specter of dangerous hepatotoxicity in the minds of prescribers and recipients of isoniazid. Clinical features and laboratory findings Presenting symptoms. Symptoms typically resemble viral hepatitis. About one-third of patients present with a prodrome characterized by malaise and fatigue; this is often attributed by the patient to a “viral illness.” Gastrointestinal symptoms are also present in about half of these cases. About 10% of patients present with jaundice alone. The remainder present with predominantly digestive complaints, such as anorexia, nausea, vomiting, and abdominal pain. Fever, arthralgia, and rash are noted in less than 10% of patients with isoniazid-induced hepatitis. Physical signs. Hepatomegaly—with or without hepatic tenderness—is found in one-third of patients admitted to the hospital with isoniazid-induced liver injury. Jaundice is usual in severe cases, but splenomegaly and signs of chronic liver disease are rare. Liver tests. These resemble acute viral hepatitis, except that in about half of the reported patients, the level of AST exceeds the ALT. Values over 500 U/L are common, and they may exceed 2000 U/L, but the level of AT abnormality does not appear to reflect prognosis. SAP is also elevated in most patients, and about 10% of patients have a biochemical “mixed” hepatocellularcholestatic hepatitis by liver test criteria. However, this does not usually correlate with a clinical syndrome of cholestasis. The serum bilirubin concentration is typically increased. Values greater than 250 pmol/L are common and indicate a poor prognosis. Prolongation of the PT is present in one-third of cases; in 1 series, 60% of individuals with this indicator of severe liver injury died. Mechanism. Isoniazid-induced hepatotoxicity is an idiosyncratic form of drug-induced liver injury. It is independent of the total or incremental dose of isoniazid, and several studies have shown no relationship to isoniazid blood levels. Drug allergy is unlikely to be involved. Thus, the onset of the reaction tends to be later than the usual interval for drug-induced allergic reactions (fever, rash, arthralgia, and eosinophilia are uncommon), and drug-induced antibodies or autoantibodies have not been found.5

80


Fig. 4. Distribution of patient gender within the FDA Adverse Event Reporting System where liver injury is co-reported with isoniazid as a suspect medication.

Adverse drug reactions. An examination of the FDA’s Adverse Event Reporting System suggests that liver injury is strongly associated with isoniazid-related adverse reactions. Specifically, 35% of the 1925 adverse event cases where isoniazid is reported as a suspect medication report some form of hepatic injury (677 cases). Figure 4 shows that both males and females are roughly equally affected by isoniazid-related liver injury. In contrast to gender, patient age appears to play a role in the seriousness of isoniazid-related liver injury, since serious patient outcomes, such as death, are skewed toward more elderly patients (Fig. 5). As noted above, isoniazid cases that report hepatitis as an adverse reaction are somewhat rare in children, with only 2.4% of isoniazid hepatitis cases coming from patients aged 10 years or less. Also as noted above, onset of isoniazid-related adverse reactions can frequently occur months after initial treatment with isoniazid (Fig. 6). Acetaminophen Acetaminophen is an analgesic (pain reliever) and anti-pyretic (controls fever) medication that is sold over-the-counter—without a prescription. More than 200 pain relievers and cold remedies under various trade names, including Tylenol, contain acetaminophen. It may not be considered a “real drug,” and a patient may not include it in their drug history reported to hospitals and doctor’s offices. Although hepatoxicity is mostly associated with overdose, liver injury can occur with therapeutic concentrations.2

Fig. 5. Comparison of the average age of patients in isoniazid-related liver injury adverse event cases as a function of whether death is reported as a patient outcome.


81

Fig. 6. Distribution of number of days to onset of liver-injury-related adverse reactions after commencement of isoniazid therapy.

Acetaminophen differs from NSAIDs because it is not anti-inflammatory. Although considered relatively non-toxic and taken by as many as 100 million Americans each year, a number of acetaminophen-related liver injuries and deaths are reported every year to poison control centers. These include cases of accidental overdose but also cases of what we call “adverse effects.” Acetaminophen is mostly converted in the liver to inactive compounds by conjugation with sulfate and glucuronide. If more acetaminophen is ingested than the body can detoxify, the sulfate and glucuronide pathways become saturated. When this happens, more acetaminophen is shunted to the cytochrome P450 system, which oxidizes acetaminophen to produce a highly reactive, toxic intermediary metabolite, N-acetyl-p-benzo-quinone imine (NAPQI). Under normal conditions, NAPQI is detoxified in the liver by conjugation with glutathione. If hepatocellular supplies of glutathione run out, the toxic NAPQI is left free to react with molecules in the liver cell membrane. This can result in widespread hepatocyte (liver cell) damage and may lead to acute hepatic necrosis (liver death). Animal studies have shown that 70% of hepatic glutathione must be depleted before hepatotoxicity occurs. The initial symptoms of acetaminophen overdose can at first seem minimally serious, but they can be quite deceiving and after several days can progress to irrevocable liver failure and death. The Matthew Nomogram is commonly used to access risk of liver injury from acetaminophen toxicity and is based on time passed since ingestion of the acetaminophen and the plasma level of acetaminophen.48,72,73 As described in the literature, the clinical course after a toxic ingestion follows 4 stages. ● Stage 1 occurs 12–24 h post-ingestion. Symptoms present include nausea, vomiting, diaphoresis, and anorexia. Children frequently have episodes of vomiting even without toxic levels. Those patients with plasma levels in the toxic range have a mean onset of symptoms by 6 h, with 100% showing symptoms by 14 h. Laboratory studies (liver enzymes as a measure of liver inflammation and clotting time as a measure of hepatic synthetic capacity) are typically normal during this time. ● Stage 2 occurs 24–48 h post-ingestion. By that time, symptoms have decreased but laboratory abnormalities begin to appear with a rise of liver enzymes (AST and ALT), bilirubin, and prothrombin time. ● Stage 3 occurs between 48 and 96 h post-ingestion and is when the peak abnormalities are seen. AST levels as high as 30,000 may be seen, with the definition of hepatotoxicity related to acetaminophen has been accepted as AST levels in excess of 1000. Well under 1% of patients in Stage 3 will develop fulminant hepatotoxicity. ● Stage 4 occurs during the first week after ingestion with hepatic abnormalities returning to near normal by 7 or 8 days.

82


Early therapeutic intervention is crucial. A number of unfortunate persons with acetaminophen toxicity do not have their symptoms immediately recognized and do not receive appropriate therapy for a number of days. Acetaminophen is considered safe at therapeutic doses; however, if an overdose occurs acute hepatic cell death may occur due to the formation of a highly toxic intermediate generated by the hepatic cytochrome P450-dependent microsomal enzymes.52 There has been a prevalence of increased susceptibility to acetaminophen-induced liver injury in patients who are taking drugs that increase CYP450 hepatic enzymes or are ingesting large quantities of alcohol. In this scenario, even common therapeutic levels of acetaminophen can cause hepatotoxicity.37 Problems with current labeling Acetaminophen is hepatotoxic if taken in overdoses, but for adults, more than 7.5–10 g/d are considered an overdose (2002 FDA Advisory Meeting). The currently recommended maximal therapeutic dose is 4 g/d; however, instructions for use are often confusing. One product states that up to two 500 mg extra strength tablets can be taken every 4–6 h as required but not more than 4 g/ d. If the condition for which acetaminophen is taken extends in duration to more than 18 h, there is a chance to exceed the recommended daily dose even with the longer recommended dosing interval (6 h). A hypothetical treatment course over 1 day based on these drug insert recommendations is as follows: 1 g at start of the therapy, 1 g at 6 h, 1 g at 12 h, 1 g at 18 h, and finally a dose at 24 h, equaling 5 g in a 24-h period. If taken every 4 h, the total goes up to 7 g/d if following the directions for use but not limiting to a maximum dose per day. This by itself would not necessarily be a problem if the toxicity would really start at 7.5–10 g/d. However, there have been several reports where doses between 4 and 7.5 g/d have been associated with hepatoxicity and fulminant hepatitis. A more precise definition of these doses would be “more than the recommended dose.” By labeling these events “overdose,” the impression is created that toxicity is a non-unexpected effect (of the overdose). It is just as reasonable to conclude that the therapeutic window for acetaminophen may be much smaller than claimed. For a drug with a narrow dosing safety margin, dosages only slightly exceeding the recommended dose could be considered an overdose. By labeling liver toxicity at minimally elevated doses of acetaminophen as due to an “overdose,” the impression is created that the toxicity is the predictable effect of an overdose rather than the more problematic consequence of high-normal acetaminophen doses. Watkins et al.74 studied 145 healthy volunteers at 2 U.S. medical centers. They were given a placebo, extra strength Tylenol, and prescription painkillers that contain acetaminophen, such as Percocet (acetaminophen plus oxycodone). Patients took the medication or placebo every 6 h for 14 days. The liver enzyme AST, which is indicative of liver inflammation, was measured daily for 8 days and at regular intervals after that. All patients were on the same diet. Out of 106 patients, 41, or 39%, taking acetaminophen alone or in combination with another drug saw their liver enzymes increase to more than 3 times the upper limit of normal. Overall, 27 patients had enzyme levels exceeding 5 times normal, and 8 patients had 8 times the normal amount of enzyme. Of the 39 patients on a placebo, only 1 had enzymes that exceeded twice the normal level. Is the drug acetaminophen, especially with its current labeling, really safe for OCT use? One point that has to be considered for OTC use is that consumers must be able to correctly dose a drug. If the therapeutic window is smaller than commonly appreciated, it can be expected that a subpopulation of normal users, because of biologic diversity and individual conditions, may actually be more susceptible to adverse effects with therapeutic doses. Move of FDA toward revised labeling In the interests of safe and reliable use, all products containing acetaminophen should display labeling with the following additional precautions: ● Add comparative dose information, so that if a particular product has more (double strength) or less acetaminophen than might be expected caution can be raised.


83

● How many milligrams, and percentage of maximum daily intake, will occur with maximum daily recommended dose. ● Caution consumers to not concurrently take multiple products containing acetaminophen and provide a list of common products containing acetaminophen, irrespective of the manufacturer. ● Caution consumers that signs and symptoms of liver toxicity (list them) may not appear for several days after acetaminophen was last used, and ask consumers to consider the development of liver toxicity as an alternate explanation to persistent cold symptoms. ● Caution consumers not to use alcohol with acetaminophen and to maintain adequate food and fluid intake.

Microsomal enzyme induction A study was performed on patients with microsomal enzyme induction due to a single dose of acetaminophen while taking either anticonvulsants or rifampin and also with healthy volunteers. The induction of drug-metabolizing enzymes was found as expected in the patients taking anticonvulsants and rifampin by observation of the significantly shortened half-life and increased clearance with no effect on the volume of distribution. Phase II glucuronide conjugation of acetaminophen was also found to be increased in patients compared to the healthy volunteers. In the healthy volunteers, the plasma concentrations of acetaminophen were lower and the half-life was shorter. However, the plasma concentrations of acetaminophen’s glucuronide conjugate and the area under the curve (AUC) ratio of the conjugate to the unmetabolized drug was increased.42 Mercapturic acid and cysteine conjugates are also generated by the conversion of acetaminophen to its hepatotoxic metabolite. Surprisingly, patients with this enzyme induction due to anticonvulsant and rifampin treatment do not seem to be at a higher risk for acetaminophen hepatotoxicity because of these reactive metabolites.42 1. Drug Master Plus Interaction Report47: Acetaminophen (oral) and phenytoin Clinical significance: moderate Action: studies show that phenytoin may increase the hepatic metabolism and decrease the oral bioavailability of acetaminophen. Acetaminophen-related hepatotoxicity might be increased, while the analgesic and anti-pyretic effectiveness may be decreased. Recommendations: although no special precautions are generally necessary, this interaction should be kept in mind if patients taking phenytoin are also taking large doses of acetaminophen or are on prolonged therapy. 2. Linden and Rumack describe the interaction between phenytoin and acetaminophen: Prior exposure to drugs that stimulate the hepatic microsomal P450 mixed-function oxidase enzyme system (including phenytoin) may enhance acetaminophen toxicity.48 The chronic use of drugs such as antihistamines, phenytoin, barbiturates, and other sedatives that stimulate the P450MFO system may enhance APAP toxicity.49 3. Martindale, the Extra Pharmacopoeia (acetaminophen interactions—review of drug interactions involving acetaminophen): gastrointestinal absorption may be delayed by drugs that decrease gastric emptying, such as anticholinergic agents or opioid analgesics. The likelihood of toxicity may be increased by the concomitant use of enzyme-inducing agents, such as alcohol or anti-epileptic drugs.50 4. Drug interaction facts51: This issue was elevated from a possible warning in 1986 to a suspected warning in 1990, which says hydantoins alter acetaminophen metabolism and several references are provided in the handbook monograph. A reference reports that in 6 patients on chronic-administration therapy, acetaminophen clearance was 46% higher and its half-life was 28% shorter in the group receiving anticonvulsants when compared to 6 controls. Other studies have confirmed this observation. 5. In most laboratory-animal species, the hepatotoxicity of acetaminophen is increased by pretreatment with microsomal enzyme inducers, such as phenobarbital and ethanol. There

84


had been reports of liver damage following the therapeutic use of acetaminophen in chronic alcoholics, and after overdose, the severity of liver damage appears to be greater in chronic alcoholics and patients who have previously been taking drugs likely to cause induction (such as phenytoin). The observation suggests that microsomal enzyme induction might increase the production of the reactive metabolite of acetaminophen and thus enhance its hepatotoxicity.52 This report appeared in the literature as far back as 1981. 6. Drugs, such as phenytoin and phenobarbital (butalbital), which induce the cytochrome P450 enzyme system, could lower the threshold for hepatic damage by increasing the production of the toxic metabolite and the utilization of glutathione. Increased susceptibility to acetaminophen-induced liver damage has been reported in patients taking drugs that induce hepatic enzymes or in chronic alcohol ingestion, which may even cause hepatotoxicity with therapeutic doses of acetaminophen.37 Therapeutic research center’s warning and FDA report The FDA convened a meeting of “stakeholders” to discuss how to address the public health problem of liver injury related to the use of acetaminophen in both over-the-counter (OTC) and prescription (RX) products. The FDA recognizes that acetaminophen is an important drug used to treat pain and fever in both settings and is not seeking to remove it from the market. The risk of developing liver injury to the individual patient who uses the drug according to directions is very low. However, acetaminophen-containing products are used extensively, making the absolute number of liver injury cases a public health concern. For additional information about acetaminophen: http://www.fda.gov/Drugs/DrugSafety/ InformationbyDrugClass/ucm165107.htm. Background. Acetaminophen is one of the most commonly used drugs in the United States,1 yet it is also an important cause of serious liver injury. Acetaminophen is the generic name of a drug found in many common brands of over-the-counter (OTC) products, such as Tylenol, and prescription (Rx) products, such as Vicodin and Percocet. Acetaminophen is an important drug, and its effectiveness in relieving pain and fever is widely known. Unlike other commonly used drugs to reduce pain and fever [e.g., non-steroidal anti-inflammatory drugs (NSAIDs), such as aspirin, ibuprofen, and naproxen], at recommended doses, acetaminophen does not cause adverse effects, such as stomach discomfort and bleeding, and is considered safe when used according to the directions on its OTC or Rx labeling. However, taking more than the recommended amount can cause liver damage, ranging from abnormalities in liver function blood tests, to acute liver failure, and even death. Many cases of overdose are caused by patients inadvertently taking more than the recommended dose (i.e., 4 g/d) of a particular product or by taking more than 1 product containing acetaminophen (e.g., an OTC product and an Rx drug containing acetaminophen). The mechanism of liver injury is not related to acetaminophen itself but to the production of a toxic metabolite. The toxic metabolite binds with liver proteins, which causes cellular injury. The ability of the liver to remove this metabolite before it binds to liver protein influences the extent of liver injury. In a study that combined data from 22 specialty medical centers in the United States, acetaminophen-related liver injury was the leading cause of acute liver failure for the years 1998 through 2003.2 Patients in this study were found to have taken too much acetaminophen from OTC, Rx products, or both. Almost half of these cases involved overdose in which the patient had not intended to take too much acetaminophen (unintentional overdoses), although many cases of liver injury with acetaminophen result from self-harm, i.e., intentional self-poisoning. The high percentage of cases of liver failure related to unintentional acetaminophen overdose was also observed in a study published in 2007.3 The extent of liver failure cases reported in the medical literature provides an important signal of concern. However, the types of databases available to identify cases make it difficult to determine the full extent of the problem or whether interventions have been successful.


85

Why acetaminophen overdoses occur There are few data available describing consumer behavior with acetaminophen products or consumer understanding of acetaminophen toxicity. However, based on the prevalence of liver injury, it appears that there are distinct factors associated with acetaminophen and acetaminophen products that contribute to this public health problem. These factors are listed below. ● Taking just a small amount of acetaminophen over the recommended total daily dose (4 g/d) may lead to liver injury.4 Currently recommended doses and tablet strengths of acetaminophen leave little room for error and the onset of liver injury can be hard to recognize. There is scientific agreement that taking a large amount of acetaminophen over a short period of time causes liver injury, but there is limited agreement as to the specific threshold dose for toxicity. In addition, the onset of symptoms associated with acetaminophen liver injury can take several days, even in severe cases. The symptoms of liver injury may not be readily identified by an individual because they may be non-specific and mimic flu symptoms. The antidote for acetaminophen poisoning, N-acetylcysteine, is less effective when liver injury has progressed too far. ● Some individuals may be especially sensitive to liver injury from acetaminophen. The maximum safe dose may not be the same for all persons. Individuals with increased sensitivity may experience toxic effects at lower acetaminophen doses. Available information suggests that some individuals, such as those who use alcohol or have liver disease, may have a greater sensitivity to the effects of the toxic metabolite because they produce more or are unable to clear it from the body as easily. More research is needed to understand whether ethnicity, genetics, nutrition, or other factors might have a role in making some individuals more sensitive. ● There is a wide array of OTC and Rx acetaminophen products used in a range of doses for various indications. For some people, it may be difficult to identify the appropriate product to use. Acetaminophen is in many widely used OTC single-ingredient products, such as those to treat headaches, and multiple-ingredient (combination) products, such as those to treat symptoms of the common cold like aches and fever. Acetaminophen is also a component of a number of Rx drug products in combination with narcotic pain medicines. So, consumers may reasonably attempt to treat different conditions or symptoms with multiple choices among products containing acetaminophen, but may not realize that acetaminophen is an ingredient common to each. ● It can be difficult to identify acetaminophen as an ingredient. Rx products that contain acetaminophen (usually with codeine or oxycodone) are often labeled as containing “APAP” on pharmacy-dispensed containers.5 Without clear labeling, patients may take more than 1 product containing acetaminophen (e.g., a Rx product and an OTC product) without realizing it, and in some cases may take a harmful overdose. ● Multiple products exist for children containing different strengths. Liquid acetaminophen formulations intended for use in infants are typically more concentrated (i.e., stronger) to enable proper dosing using less liquid. However, failure to distinguish between the 2 strengths of liquid can result in an accidental overdose where the parent gives a higher dose of the concentrated drops to a younger child. ● The association between acetaminophen and liver injury is not common knowledge.6 Consumers are not sufficiently aware that acetaminophen can cause serious liver injury, and their perceptions may be influenced by the marketing of the products. Finding ways to educate consumers about the risk of liver injury from acetaminophen has been difficult. Current labeling on OTC products may be overlooked, as can the patient information provided with dispensed prescriptions. Programs to educate the public about safe use of acetaminophen have been small and encountered a number of obstacles. Advertisements of OTC drugs often emphasize the effectiveness of products, but are not subject to the same requirements to offset such messages by providing warning information as prescription products. Also, acetaminophen is available in retail outlets in large quantities (e.g., 500 tablets per bottle), which may contribute to the perception that the ingredient is unlikely to be harmful.

86


FDA’s previous actions In the late 1990s, research began to show that acetaminophen was a major cause of acute liver failure in the United States, with up to half of the cases due to accidental overdose. Responding to these concerns, the FDA took a number of steps to reduce the incidence of liver injury related to acetaminophen. In 1998, the FDA finalized a regulation that required all OTC acetaminophen products to include an alcohol warning in labeling. The warning stated: Acetaminophen. “Alcohol Warning”: “If you consume 3 or more alcoholic drinks every day, ask your doctor whether you should take acetaminophen or other pain relievers/fever reducers. Acetaminophen may cause liver damage.” In 2002, the FDA convened an Advisory Committee meeting to discuss unintentional liver toxicity related to the use of OTC acetaminophen.7 The Advisory Committee recommended a specific liver toxicity warning and distinctive labeling on OTC packages so that products containing acetaminophen could be more easily identified. The FDA and manufacturers were also advised to educate consumers and health professionals about the risk of liver injury from acetaminophen. In early 2004, the FDA launched a public education campaign to help consumers use acetaminophen more safely. By most standards, the campaign would be considered small, due to budgetary constraints. It was also limited by reluctance on the part of some commercial outlets to provide a venue for the FDA’s message about acetaminophen toxicity as the product was sold or promoted in those outlets. Nonetheless, the FDA has continued to expand efforts to improve public education about acetaminophen overdosing and liver injury and has recently updated the acetaminophen information on its Web site. In 2004, the FDA sent letters to every state board of pharmacy asking them to consider requiring labeling on the immediate container of Rx products containing acetaminophen that is as follows: (1) uses the term acetaminophen, not APAP, (2) instructs patients to avoid concurrent use of other acetaminophen-containing drugs, (3) instructs patients not to exceed the maximum daily recommended acetaminophen dose, and (4) instructs patients to avoid drinking alcohol during prescription use.8 FDA was informed by the National Association of Boards of Pharmacy that, as of February 2008, no states had implemented regulations related to the request. In December 2006, FDA issued proposed regulations for OTC labeling for acetaminophencontaining products to require inclusion of new safety information and that the container and outer carton identify acetaminophen when it is an ingredient.9 The final version of the regulation is currently under review. In 2007, the Director of FDA’s Center for Drug Evaluation and Research (CDER) convened a multidisciplinary working group in CDER to continue to evaluate the issues associated with acetaminophen-related liver injury and consider additional steps FDA could take to decrease the number of cases of acetaminophen-related liver injury. The working group considered detailed reviews of the issues from the Office of Nonprescription Products, the Office of Surveillance and Epidemiology, and the Division of Anesthesia and Analgesic and Rheumatology Drug Products as part of its deliberations. The working group considered the full range of options proposed and made recommendations to the Center Director regarding which should be considered for implementation. Given the complex nature of the underlying problem of acetaminophen liver toxicity, the Center Director and the Working Group agreed that the options should be presented for public discussion prior to taking further action. The report of the Working Group was made available by or around May 22, 2009, at the 2009 Meeting Materials web page, click on the year 2009 and scroll down to the appropriate advisory committee link (http://www.fda.gov/Drugs/ DrugSafety/InformationbyDrugClass/ucm165107.htm). Adverse drug reactions. There are 5824 adverse event cases within the FDA’s Adverse Event Reporting System (AERS 1997–2012) in which some form of liver injury is reported and acetaminophen (or an acetaminophen-containing pharmaceutical) is reported as a suspect medication. The role of acetaminophen dosage and liver injury is well established, and as expected, a large percentage (39%) of these acetaminophen-related liver injury cases are explicitly associated with acetaminophen overdose (intentional or accidental). Interestingly, there appears to be a significant skewing of these adverse event cases toward female patients,


87

Fig. 7. Distribution of patient gender within the FDA Adverse Event Reporting System Cases where liver injury is coreported with acetaminophen as a suspect medication.

with an almost 2:1 ratio of females to males in liver injury cases where overdose is reported (Fig. 7). Furthermore, children under the age of 10 years represent a significant portion of acetaminophen-related liver injury cases (Fig. 8).

Macrolide antibiotics Macrolide-induced hepatotoxicity is characterized by cholestatic hepatitis. The best example of this is erythromycin estolate. The onset of this form of DILD is 2–25 days after macrolide exposure. The clinical symptoms include anorexia, nausea, vomiting, and significant right upper quadrant pain. Jaundice also occurs in 50% of patients. Liver biopsies indicate cholestasis with a plethora of white blood cells in the portal inflammatory infiltrate. Newer macrolide antibiotics have caused cholestatic jaundice. Examples include roxithromycin, clarithromycin, and azithromycin.34 Biaxin (Clarithromycin)-Induced Fulminant Hepatic Failure A 63-year-old man was treated prophylactically to eradicate Helicobacter pylori with Biaxin (Clarithromycin, Abbott). The patient was hospitalized and treated for gastric ulcer and was

Fig. 8. Distribution of patient ages in adverse event cases where liver injury is co-reported with acetaminophen and acetaminophen is considered a suspect medication. Data obtained from the U.S. Food and Drug Administration’s Adverse Event Reporting System.

88


prescribed Biaxin 500 mg TID (3 times a day) for 2 weeks. H. pylori was identified following Giemsa staining. Initially, the patient’s gastroenterologist had recommended Prilosec 20 mg 2 times daily for 7 days, Biaxin 250 mg 2 times daily (500 mg/d) for 7 days, and Flagyl/ metronidazole 500 mg 2 times daily for treatment of the H. pylori infection. The patient’s internist, however, prescribed Biaxin 500 mg TID (3 times daily) for 2 weeks. In addition, the patient was also taking 60 mg of Procardia, 30 mg of Prevacid, and 20 mg of Mevacor per day. He was admitted to the hospital with increasing abdominal pain and an ashen color. The patient had mild abdominal pain for 3 days, but it significantly worsened in the preceding 24–48 h. He became hypotensive on the ward, was transferred to ICU, coded, and was found to be profoundly acidotic; probably secondary to sepsis-WBC 19,000 range. No source for the sepsis was noted. A surgical note indicates that the patient was started on Biaxin for H. pylori. He was recently discharged from the hospital a few days earlier after treatment for gastric ulcer, was placed on Biaxin, apparently, with active bleeding. The patient could not be resuscitated and died. The surgeons described the following: “At surgery, the liver seemed somewhat tense and firm. Wedge biopsy of the liver was procured. A pathologist’s report for the liver biopsy report indicated ‘severe central lobular necrosis and hemorrhage’…. The possibility of shock or cardiovascular failure is mentioned. In addition, the possibility of DIC, sepsis, drugs or toxins is mentioned. These findings should be correlated clinically.” The patient’s pretreatment liver function tests were normal, and a pre-Biaxin CT of the abdomen did not demonstrate any abnormal liver findings. The death certificate lists the cause of death as hepatic failure, drug reaction (Biaxin). A published article by Wallace et al.,53 and publications that followed, warned about highdose risk in the elderly and suggested caution. A December 1995 FDA Advisory Committee recommendation for a 1500 mg/d dosage of Biaxin for treatment of H. pylori would predictably increase the use of Biaxin. An FDA Anti-Infective Drug and Gastrointestinal Drugs Advisory committee recommendation was as follows: “… a combination therapy of the antisecretory medication Prilosec and the antibiotic Biaxin be approved for the treatment of Helicobacter pylori (H. pylori) infected patients with active duodenal ulcer to eradicate H. pylori, the bacteria now believed to cause approximately 89 percent of peptic ulcers. . . . The recommended dosage of eradicating H. pylori is Prilosec 40 mg once daily and Biaxin 500 mg three times daily for the first 14 days….” This FDA Advisory Committee recommendation would predictably increase the use and the dosage of clarithromycin (used for H. pylori infection in the stomach). Literature describing the dose-related phenomenon of hepatotoxicity preceded the patient’s Biaxin death. Brown et al.54 and also previously Wallace et al.53 described abnormal liver enzyme levels during high-dose clarithromycin monotherapy for Mycobacterium avium complex or Mycobacterium abscesses. Portions of the article are reproduced here for discussion: The elevation in enzyme levels was seen in five (36%) of 14 elderly patients receiving clarithromycin at a dose of 2,000 mg/day, and it was associated with unexpectedly high serum drug levels. In all five patients, both the transaminases and the alkaline phosphatases were elevated. With discontinuation of the high dose, the GOT and/or the SGPT in the five patients fell by at least 50% with all values 100 IU/liter or less by the next enzyme determination (one or two weeks). The alkaline phosphatase and GGT were slower to decline, and three of five patients remained at approximately the same level for at least four weeks. Healthy, elderly volunteers have been noted to have higher peak and trough levels of both the parent drug and its major metabolite compared with younger volunteers. This drug dose (1,000 mg/day) should be used with caution in any patient over age 55. This strongly suggests that clinical trials of mycobacterial lung disease in older age populations should use no more than 1,000 mg per day except under special circumstances, with attention being paid to body mass and renal function, to minimize side effects.


89

The article by Brown et al.54 continues:

In addition, since these data were published, we have noted two other HIV-seronegative patients who developed abnormal liver enzyme levels during clarithromycin therapy…. The patients who develop these signs of hepatotoxicity are typically elderly or have reduced body mass, are receiving clarithromycin doses of 2,000 mg/day, and are often asymptomatic…. The ability of patients to tolerate lower doses of clarithromycin ….(four of four tested strongly suggest that this is a toxicity that is serum-level-related and not due to hypersensitivity. The authors recommend “reductions in doses may be required in patients with reduced renal function and/or reduced body mass to prevent hepatic or gastrointestinal toxicity.” The elderly fall into both of these categories as a result of their elevated age (normal aging process reduces lean body mass and deceases renal function as a function of age). Central zonal necrosis is an imminent liver toxic sign, frequently associated with a few drugs acting like liver toxins (e.g., acetaminophen). Clarithromycin was the only drug being taken by the patient discussed in this case at the time of his fatal liver toxicity that has had central zonal necrosis associated with its use. Conventional knowledge describes macrolide toxicities, including clarithromycin, to be associated with occasional abnormalities in liver tests, and occasional cholestatic jaundice, considered idiosyncratic, and therefore not dose related. Most are described as mild and transient. Shaheen and Grimm55 described a 25-year-old man who developed fulminant hepatic failure following clarithromycin use; he subsequently required liver transplantation. The authors wrote, “the potential for severe liver injury associated with clarithromycin is of growing concern in the light of recent reports to the FDA. We believe that this case, although not proof of causation, supports a true association.” Clarithromycin along with many other macrolide antibiotics has the capacity to cause liver cholestasis, however more uncommon is hepatocellular injury. In less than 1% of patients, there is an elevation of AST, ALT, bilirubin, and alkaline phosphatase. Animal studies have indicated that large doses of clarithromycin can cause reversible liver dysfunction, however in humans, there has been no dose relation with hepatotoxicity. Hepatic necrosis is even more rare when associated with macrolide erythromycin.55 The Food and Drug Administration’s AERS database includes 878 reported cases of liver toxicity where clarithromycin is considered a suspect, including 20 cases where pancreatitis is also reported. Figure 9 shows the age distribution of patients reporting clarithromycin-related liver injury cases, with an average patient age of 54.5 years.

Fig. 9. Distribution of patient ages in adverse event cases where liver injury is co-reported and clarithromycin is considered a suspect medication.

90


Rezulin (troglitazone) hepatotoxicity This hypoglycemic agent was launched in 1997, and it was recognized pre-approval that the drug had a hepatotoxic risk. Special precautions were included in the labeling calling for close monitoring and frequent liver function tests. A number of lawsuits were filed against physicians who failed to monitor patients, pharmacists who failed to counsel patients of the need to monitor, and the company marketing Rezulin, Parke Davis, which failed to adequately warn about the product. Case report The following is a case report by one of the authors (J.T.O.D.) regarding a woman who developed Rezulin-associated cirrhosis and was placed on a liver transplant list. Rezulin therapy initiated on 12-10-97 Rezulin discontinued 10-98 Glynase was changed to Amaryl because of uncontrolled blood glucose. Dr. Physician’s New Patient Evaluation of 11/19/97 was as follows: In regard to the diabetes, patient does selfglucose monitoring 4 times a day. She mentioned in the morning it is 126-140 and after lunch sometimes it can drop down to 40-50 and sometimes up to 200, and this happens 2-3 times a week. 11/19/97 Amaryl was then changed to Rezulin. Physician’s note of 12-10-97 was as follows: “The patient relates that she is still having some problems with hypoglycemia with Amaryl.” PLAN: “My plan is to stop the Amaryl. We will try some Rezulin 200 mg daily for two weeks and then increasing to 400 mg daily.” The lab results for the Liver Function Test (LFT) show elevated LFT enzymes as early as 1224-97. There is a rise in ALT in April 1998 that persists through 07-98. The Rezulin is discontinued in October 1998 “because of elevated liver enzymes.” The October, 1998, lab results list AST as 33 and ALT as 35. There is no record of the LDH or GGT for Oct 98; ALK PHO is 99. In the 10/30 visit with a second physician, the following is recorded: Assessment: Did have some elevation of liver enzymes last visit. Will stop Rezulin. Start Glucophage 500 mg, on bid for two weeks; if sugars drop below 120, cut insulin in half. If they remain below 120 will stop insulin. If one bid does not result in this, go up to two bid. Recheck in two weeks. This change was apparently prompted by the July 3 and July 8 LFTs that showed AST 62 and ALT 69 on both dates (two separate, dated reports). Almost four months passed before these tests got action to discontinue the Rezulin. July 3-Oct 30. Examination of the LFTs for the patient shows a progressive and continuous elevation of the LFTs. Subsequently, she developed a chronic hepatitis condition. Discussion The catastrophic idiosyncratic liver damage did not appear as an acute fulminant hepatitis, but rather as a chronic active hepatitis. The purpose of frequent LFTs was to observe the onset of liver enzyme release as a result of the Rezulin exposure. With the exception of a single borderline lab study, or in the absence of an alternative treatment for diabetes, 2 LFTs in the abnormal range raised an alarm and caused the Rezulin to be discontinued. Since the damage may be asymptomatic, the tests were the patient’s only protection from slowly evolving liver damage, perhaps from some mildly toxic metabolite of Rezulin. The FDA advisory committee discussed monitoring LFT in Rezulin therapy. There was testimony that the liver damage was not an allergic reaction but an idiosyncratic reaction. ALT is described as a poor measure of liver damage, but the best available. Liver damage can be


91

estimated by monitoring a rise above normal limits. The elevation of these LFTs can be marginal, but 1 and one-half to 3 times normal is unmistakable. Dr. Seeff, a member of the FDA panel discussing Rezulin, however, stated that the purpose of the tests was to identify the unlucky patient who was experiencing hepatic damage and that more frequent testing and prompt discontinuation was the best bet. The testimony by Dr. Graham, FDA Epidemiology, suggested that the mean time of injury after starting Rezulin therapy was about 6 months; and that by 1 year, most patients who had not suffered the idiosyncratic injury were not likely to do so in the future. It therefore appeared that the first 4–6 months of therapy were crucial to monitor for elevation of LFT and to discontinue Rezulin in the face of AST numbers above normal, unless no alternative therapy was available to treat diabetes. An interesting editorial in the October 1999 issue of Pharmacy Today goes further, calling for the withdrawal of Rezulin.56 The inescapable conclusion was that Rezulin should be withdrawn from the market and that if the FDA chose to permit the continued marketing of Rezulin, its use should have been restricted to those patients who were already taking it and tolerating it satisfactorily. Most would have been better served if they were switched from Rezulin to Avandia or Actos (other glitazone agents with similar activity in treating diabetes). Since the withdrawal of Rezulin (3/ 21/00), there have been only a few reports of liver failure with these 2 alternate drugs, and strengthening of the package insert warnings has followed. However, both Avandia and Actos were found to have significant cardiac risks, leading to congestive heart failure, and their use has been sharply curtailed/limited by the FDA and practicing physicians.

An enhanced black box warning was issued by the FDA WARNINGS Hepatic

Rare cases of severe idiosyncratic hepatocellular injury have been reported during marketed use (see ADVERSE REACTIONS). The hepatic injury is usually reversible, but very rare cases of hepatic failure, leading to death or liver transplant, have been reported. Injury has occurred after both short- and long-term troglitazone treatment. During all clinical studies in North America, a total of 48 of 2510 (1.9%) Rezulin-treated patients and 3 of 475 (0.6%) placebo-treated patients had ALT levels greater than 3 times the upper limit of normal. Twenty of the Rezulin (troglitazone removed from the US market 3/21/ 00) -treated and one of the placebo-treated patients were withdrawn from treatment. Two of the 20 Rezulin-treated patients developed reversible jaundice; one of these patients had a liver biopsy which was consistent with an idiosyncratic drug reaction. An additional Rezulintreated patient had a liver biopsy which was also consistent with an idiosyncratic drug reaction. (See ADVERSE REACTIONS, Laboratory Abnormalities.) It is recommended that serum transaminase levels be checked at the start of therapy, monthly for the first six months of therapy, every two months for the remainder of the first year of troglitazone therapy, and periodically thereafter. Liver function tests also should be obtained for patients at the first symptoms suggestive of hepatic dysfunction, e.g., nausea, vomiting, abdominal pain, fatigue. anorexia, dark urine. Rezulin therapy should not be initiated if the patient exhibits clinical or laboratory evidence of active liver disease (e.g., ALT4 3 times the upper limit of normal) and should be discontinued if the patient has jaundice or laboratory measurements suggest liver injury (e.g., ALT4 3 times the upper limit of normal). Despite the warning, patients continued to report significant liver toxicity. Despite the actions of the company to defend their blockbuster drug and keep it on the market, the drug was withdrawn, and there has been extensive litigation associated with Rezulin. An examination of the FDA’s AERS/MedWatch database identifies 2202 cases of liver injury where Rezulin was considered a suspect. Consistent with Rezulin’s original prescribing

92


Fig. 10. Top 25 adverse reactions co-reported with liver injury when Rezulin is a suspect medication.

information, a large portion of these cases co-report some form of abnormal liver function test result (Fig. 10). The demographic distribution of patients within this group indicates that Rezulinassociated liver injury is somewhat skewed toward females (Fig. 11) and older patients (Fig. 12). The cause of the preponderance of female patients in this group is unclear. In contrast, the bias toward older patients most likely mirrors the age distribution of patients with type 2 diabetes. Methotrexate (MTX) hepatotoxicity Methotrexate for psoriasis Methotrexate hepatotoxicity in psoriatic patients is well established; however, there are differences in the incidences and risks for hepatotoxicity.57 A study on 104 patients in Nova

Fig. 11. Distribution of patient genders within the FDA Adverse Event Reporting System where liver injury is co-reported with Rezulin as a suspect medication.


93

Fig. 12. Distribution of patient ages in adverse event cases within the FDA Adverse Event Reporting System where liver injury is co-reported with Rezulin as a suspect medication.

Scotia was performed to look deeper into the connections between MTX hepatotoxicity and risks such as obesity, diabetes, and alcohol consumption.58 MTX is commonly used for severe psoriasis and psoriatic arthritis and has an established efficacy and significant degree of patient compliance. Side effects associated with MTX have been shown to greatly decrease by using low, intermittent doses. However, even in this case, if used long term, there is a potential for hepatic fibrosis and cirrhosis. The incidence of cirrhosis in this case can be as much as 25.6%.58 With the increasing usage of methotrexate (MTX), a well-defined toxicology profile has been developed, with hepatotoxicity being the most significant adverse reaction that can occur with long-term therapy. Hepatic fibrosis became evident in children and adults with leukemia who were given daily doses of MTX. In the late 1960s, case reports included psoriasis patients who developed cirrhosis that were also treated with MTX. Taken together, this information suggested that MTX had the means to cause hepatotoxicity in patients with nonmalignant conditions that were treated with low doses of methotrexate.59 Roenigk and coworkers generated some guidelines on the dosage indications of MTX on psoriatic patient assessments in 1972. There were further revisions in 1973, 1982, and 1988. These guidelines made recommendations upon the clinical and laboratory assessments of renal, hematopoietic, and hepatic function in these patients as well as liver biopsy analyses. This analysis would be performed before MTX treatment and repeated annually while on the drug. The risks of hepatotoxicity are dose-dependent and duration-dependent. There are also other risk factors that apply to the presence and intensity of liver toxicity. These include obesity, diabetes, alcohol, age, and any previous liver dysfunction.58 In psoriatic patients undergoing MTX therapy, the most significant side effects are liver fibrosis and cirrhosis. Liver function monitoring does not always exhibit the accurate slow progression of underlying MTX-induced fibrosis. However, many studies have indicated that monitoring the histology of the liver is the best way to accurately monitor the MTX-induced hepatotoxicity and reducing the risk of cirrhosis. Even with these recommendations, however, the use of liver biopsy monitoring is controversial because of its invasiveness. Many need more extensive justification for using this technique.58 Methotrexate for rheumatoid arthritis Methotrexate (MTX) is a folic acid antagonist that was approved for use in the treatment of rheumatoid arthritis in 1988. There has been a long-standing issue of its role in hepatotoxicity in these patients, and it has also been seen to be a problem in psoriatic patients. The guidelines to establish such diagnoses were written in 1982 and revised in 1988 by the Psoriatic Task Force (PTF). The community of rheumatology adopted the PTF’s findings themselves due to lack of research on the subject of MTX-induced hepatotoxicity for their respective patients; however, these patients were eventually found to have a lower incidence of serious liver disease than psoriasis patients when treated weekly with MTX.60

94


Methotrexate (MTX) is a highly significant and ubiquitous drug treatment for rheumatoid arthritis. Patients with RA usually respond very favorably to low-dose methotrexate therapy, and the probability of patients continuing their treatment beyond 5 years is greater than for other slow-acting antirheumatic drugs. Thus, given its sustained clinical utility and relatively predictable toxicity profile, low-dose methotrexate is a useful addition to the therapy of RA.61 An evaluation of methotrexate has been performed by the American College of Rheumatology in order to establish the risks of developing liver disease. This was done in order to optimize recommendations given to patients about liver toxicity monitoring and to evaluate the role of the surveillance of liver biopsies. The result of this evaluation was to recommend running liver blood tests for ALT, AST, alkaline phosphatase, albumin, and bilirubin and perform hepatitis B and C serologic studies and other common tests such as blood cell count and serum creatinine tests before staring methotrexate treatment. Liver biopsies should only be considered when the patient has a history of alcoholism, abnormal AST values, or chronic hepatitis. Every month or 2 LFT samples should be taken to monitor AST, ALT, and albumin levels; however, liver biopsies should not be routine (only for pretreatment if applicable). Liver biopsies at this stage are only recommended if liver blood tests are persistently abnormal. Abnormalities of liver blood tests include elevations in AST greater than 50% of the time in a 12-month period or a decrease in serum albumin. The patients who are selected for liver biopsy are those seemingly at a high risk for developing CSLD (clinically significant liver disease). Vigilant monitoring is mandatory to decrease the risk of CSLD.62 In a study performed by an Ad Hoc Committee on methotrexate, the cumulative dose of methotrexate was shown to be a risk factor for cirrhosis progression. The investigators first discovered this progression of CSLD after a cumulative methotrexate dose of 1.5 g.62 With patients who have a history of high alcohol consumption, it is recommended that the patient consult with a gastroenterologist to consider a pretreatment liver biopsy. The decision for this biopsy is at the discretion of the physician and at the agreement of the patient. This is also true for patients with hepatitis B or C and any case of unexplained elevations in serum transaminase levels. Outside of these patients, the risks and costs of a liver biopsy overshadow the benefits of getting one.62 Patients who are about to begin methotrexate treatment have a requirement to abstain from alcohol consumption; however, exceptions include special occasions. Regular alcohol consumption is strictly discouraged due to the lack of data on what is a safe amount of alcohol when taking methotrexate. Patients need to be counseled and guided regarding the risks of CSLD especially involving alcohol consumption, dose, and age. They should be educated on the reasons and logic behind routine blood monitoring (4–8 weeks), the need for liver biopsy, and also potential complications; this conversation should be indicated in the patient’s records.62 Judicious screening prior to MTX institution, laboratory monitoring, and appropriate clinical care are the most cost-effective and advantageous to the RA patient receiving MTX therapy.63 In cases where there are persistent elevations in serum transaminase levels, it is recommended to change the NSAID used or decrease in MTX dose followed by temporary discontinuation of MTX. If serum transaminase levels are still elevated despite these adjustments, then a liver biopsy should be considered. This is the case of AST or ALT levels that are within the range in 5 of 9 annual liver blood tests or if the serum albumin level decreases below the normal limits in the scenario of well-controlled rheumatoid arthritis, before MTX is subsequently continued. The frequency of liver blood tests is crucial if a non-invasive route is to be used to determine the risk of CSLD.62 Clinical features and laboratory findings Presenting symptoms. MTX-induced hepatic injury and liver enzyme elevations have been demonstrated after treatment of leukemia and gestational disease and during treatment of psoriasis and rheumatoid arthritis. A 40-year-old man with a long-standing history of rheumatoid arthritis was treated with MTX over a 6-month period and developed an overwhelming hepatic necrosis. He was successfully transplanted.64


95

Screening for higher risk. In a retrospective Canadian study evaluating MTX hepatotoxicity in psoriatics, the incidence of severe hepatotoxicity was shown to be high: 23.1% (24 of 104 patients). This study showed diabetic patients to be at particular increased risk of MTX hepatotoxicity. Occasional alcohol consumption was not associated with increased risk. Three patients who developed cirrhosis over 2 years of standard MTX therapy may represent a subset of psoriatics with increased hepatic susceptibility to MTX. In another 3 patients, severe hepatic fibrosis had regressed upon discontinuation of MTX, but they developed accelerated recurrence of the severe hepatic fibrosis upon resumption of MTX therapy, this suggests the possibility of unusual sensitivity to the drug. These cases reemphasize the need for continuing surveillance, with regular liver biopsies, of psoriatic patients on MTX.58 Liver function tests. Serial measurement of liver enzymes is useful to detect liver toxicity due to methotrexate in patients with rheumatoid arthritis or other rheumatic diseases. A Spanish series of 141 adult patients who were treated with methotrexate and studied retrospectively from 1988 to 1991. The more common diagnoses included rheumatoid arthritis (120 cases) and psoriatic arthritis (12 cases). In periodic studies carried out every 2–3 months, a transient increase in transaminase values associated with methotrexate in 13 patients (9.2%) was observed. Two patients developed a viral infection during therapy, one due to cytomegalovirus and the other due to the Epstein–Barr virus. Both patients had a favorable outcome once methotrexate was withdrawn.65 In using liver enzymes as a means of monitoring for liver damage, less than 49% of the time, abnormal AST values correlated with 97% specificity for a normal biopsy grade. Regular AST measurements are useful markers of hepatic histologic outcome, within the range of mostly normal histology, in patients with RA receiving long-term weekly MTX.66 The value of dynamic hepatic scintigraphy (DHS) and serum aminoterminal propeptide of type III procollagen (PIIINP) were investigated as screening methods for early detection of MTXinduced hepatic damage in 25 patients. These relatively non-invasive procedures were compared with the liver biopsy classification. DHS appeared to be very promising as a screening test to differentiate between the presence or absence of MTX-induced hepatic damage but was unsuitable to reliably grade the severity of hepatic damage. Although a global relationship was demonstrated between serum PIIINP concentration and hepatic damage, single measurements in individual patients were not reliable. The combination of PIIINP measurements with DHS had only a limited additional value above DHS alone. DHS has great promise for the detection of early MTX-induced hepatic damage. These results reinforce the need for regular liver biopsies to ensure the safe prolonged use of MTX in psoriasis patients.57 Mechanism of damage. The mechanism by which MTX causes liver damage is unknown. Methotrexate is eliminated almost entirely by the kidneys. The risk of methotrexate toxicity is therefore increased in patients with poor renal function, most likely as a result of drug accumulation. Declining renal function with age may thus be an important predictor of toxicity to methotrexate. Up to 60% of all patients who receive methotrexate for rheumatoid arthritis (RA) discontinue taking it because of adverse effects, most of which occur during the first year of therapy. Gastrointestinal complications are the most common adverse effects of methotrexate, but hepatotoxicity, hematological toxicity, pulmonary toxicity, lymphoproliferative disorders, and exacerbation of rheumatic nodules have been reported. Decreased renal function as a result of disease and/or aging appears to be an important determinant of hepatic, lymphoproliferative, and hematological toxicity. Concomitant use of low doses of folic acid has been recommended as an approach to limiting toxicity. Interactions between methotrexate and several non-steroidal antiinflammatory drugs have been reported, but they may not be clinically significant. However, caution is advised in the use of such combinations in patients with reduced renal function.61 Adverse drug reactions Examination of the FDA’s Adverse Event Reporting System shows that, of the 2512 cases reporting methotrexate as a suspect medication with liver injury, there is a bimodal distribution of patient ages,

96


Fig. 13. Distribution of patient ages in adverse event cases within the FDA Adverse Event Reporting System where liver injury is co-reported with methotrexate as a suspect medication.

with a minor mode highlighting patients aged less than 20 years (Fig. 13). The gender distribution of those suffering liver injury is heavily skewed toward females (62% for females vs. 38% for males), which likely mirrors the gender composition of patients who are most likely to be prescribed methotrexate. As noted above, declining renal function is commonly associated with methotrexate and liver injury, with 15% of these cases reporting some form of renal function disorder. Estrogens and oral contraceptive steroids Estrogen-induced cholestatic liver damage is another prominent condition. In this case, ethnicity plays a role. Ethnic diversity involving this condition suggests that there are genetic factors that govern one’s sensitivity to estrogen-induced cholestasis. This condition is high in Chile and Scandinavia where 1 in 4000 women are affected. In other countries, the frequency is 1 in 10,000. Additionally, 50% of women who have had pregnancy-derived benign cholestasis develop cholestasis with oral contraceptives. The estrogen-induced liver dysfunction usually terminates within days or weeks of estrogen cessation, but there have been reported cases of chronic cholestatic progression. Other agents that have produced a similar response include 17-alkylated anabolic steroids, which may also cause chronic cholestasis, and tamoxifen. Estrogens at physiological doses are pertinent in therapies such as hormone replacement therapy (HRT) for menopause and do not cause cholestasis and can be used effectively in patients with chronic cholestasis. This is particularly important because these patients are at risk for osteoporosis.34 Statins Statins are 3-hydroxy-3-methyl glutaryl-coenzyme A reductase inhibitors that commonly cause a rise in angiotensin. The liver disorders that are connected with statin use include mild hepatic injury and end state, or fulminant, hepatic failure. However, fulminant hepatic failure is uncommon. Cholestasis is a side effect of this condition and is associated with the statins lovastatin, atorvastatin, and pravastatin. Liver histology indicates acute cholestatic hepatitis and liver blood test alterations approach normal levels after statin discontinuation.34 Ticlopidine Ticlopidine is a platelet inhibitor that is used for the treatment of vascular-related conditions. In early clinical trials, the liver blood tests indicated anomalies in 4.4% of the patients, and greater than 30 ticlopidine-induced cholestasis patients have been discovered. These individuals were mostly comprised those of ages 55 years and above, possibly in connection with atherosclerotic vascular disease. Symptoms begin 2–12 weeks after ticlopidine administration


97

and include pruritis, jaundice, abdominal discomfort, and fever. AP levels increase to 3 times the normal amount with a moderate rise in AT. Liver blood tests indicate normalization of these levels within 3–6 months after drug cessation, but full recovery may take greater than 1 year.34 ACE Inhibitors and AT II receptor antagonists Angiotensin-converting enzyme (ACE) inhibitors are also involved in cholestasis, and these include captopril. The onset of cholestatic symptoms takes about 8 weeks after the drug is taken and symptoms include acute cholangitis with abdominal pain, fever, and jaundice. Angiotensin II receptor antagonists such as irbesartan constitute a new class of antihypertensives that have also been known to cause cholestatic hepatitis.34 Fluoroquinolones Fluoroquinolones are used ubiquitously for a variety of clinical infections, and with this in mind, hepatic dysfunction is lower than expected. A total of 14 cases have been discovered with 8 of them being cholestasis.46 Drug examples include ciprofloxacin, norfloxacin, ofloxacin and enoxacin, with clinical symptom onset between 3 and 12 days after drug administration. The hepatic threat was resolved in all of the patients who presented with cholestasis.67,68 However, this drug class has had its hepataotoxic notoriety. A member of this class, Temafloxacin, was removed from the Market shortly after approval due to severe liver toxicity. In June 1999, the U.S. FDA advised physicians to limit the prescription of Trovan (trovafloxacin —Pfizer) after it had been found “strongly associated” with 14 cases of acute liver failure, 6 deaths, and 4 patients who required liver transplantation. The FDA had received over 100 reports of liver problems in people taking Trovan, which was at that time being prescribed at a rate of 300,000 patients per month in the United States. The FDA determined that there were 14 reports, “but we don’t know how many cases were not reported,” said Dr. Goldberger. It is likely that 14 cases is an underestimation of the actual number of cases. The FDA acknowledged the severity of the more extreme hepatotoxicity cases and responded by having strong warnings added to labels. A public health advisory was sent to doctors, and the drug company complemented that with a “Dear Doctor” letter. Availability of the oral form of the drug was drastically reduced, and in the intravenous form, it is available in hospitals as a second-line therapy for only severe indications. Between its launch in February 1998 and initial reports of hepatotoxicity in 1999, over 2 million prescriptions for both the oral and intravenous forms of the drug had been filled.80 Trovan was subsequently withdrawn from the market. Terbinafine Terbinafine is an antifungal used in the treatment of onychomycosis, which is a fungal infection of the nails. Of treated patients, 10% have had elevated AST and bilirubin levels and 4 cases of cholestasis have been reported. The onset of hepatic threat occurs between 20 and 40 days after therapy commencement. After drug cessation, liver tests normalize between 3 and 6 months. Histological analysis indicated hepatocellular and canalicular cholestasis with different stages of portal vein inflammation. Three cases have shown cholangitis and 1 case has shown ductopenia. Other drugs that fit this category include ketoconazole and fluconazole.69 Adverse drug reactions Analysis of the FDA Adverse Event Reporting System identifies 953 cases where liver injury is reported with terbinafine as a suspect medication, with the majority of these cases reporting some form of abnormal liver enzyme test result (Fig. 14). The AERS database indicates that terbinafine-related liver injury reactions primarily occur within approximately 40 days from the beginning of terbinafine therapy (Fig. 15).

98


Fig. 14. Top 25 adverse reactions co-reported with liver injury when terbinafine is a suspect medication.

Chlorpromazine Chlorpromazine-induced liver dysfunction occurs in 0.2–2% of patients. The emergence of this form of DILD takes 1–6 weeks after the administration of the drug. Over half of the patients initially experience fever and other systemic complications, then later on experience jaundice and pruritis. Liver blood tests indicate an increased ALT (greater than 3 times the normal level) with an increased AST level. Eosinophilia is also present in 60–80% of patients. Liver biopsies

Fig. 15. Distribution of number of days to onset of liver-injury-related adverse reaction after commencement of terbinafine therapy.


99

result in histological observations including centrilobular cholestasis, portal vein inflammation, parenchymal injury, and bile duct damage. The termination of this drug-induced liver damage progression occurs within 12 weeks after drug cessation; however, about 7% of patients contract VBDS. Other drugs that progress in a similar manner include prochlorperazine, haloperidol, pimozide, sulpiride, and clozapine.34 Herbal and environmental hepatotoxicity—Chaparral, Kava, and Aegeline The use of herbal remedies has received heightened attention in its potential ability to treat liver disease; however, there has been concern regarding their adverse hepatic effects. Environmental hepatotoxins have also been of speculation especially in rural environments.36 Chaparral Chaparral is one such herbal remedy and is derived from the creosote bush Larrea tridentata leaves. This is an evergreen desert shrub that can be found in southwestern parts of the United States and also in Mexico. Chaparral has been used in treating cutaneous disorders, cancer, and many other illnesses; however, some liver toxicity has been linked to this remedy. This includes acute and chronic hepatitis, cholestasis, fulminant hepatic failure as well as cirrhosis.70 Kava Kava comes from the underground portion of Piper methysticum, a shrub native to the South Pacific. In the USA, kava was sold in concentrated form as a dietary supplement to counter anxiety, insomnia, and stress. Reports from Germany, Switzerland, and the USA cited more than 25 cases of liver damage linked to Kava products from 1999 to 2002, according to the FDA and Centers for Disease Control and Prevention (CDC). U.S. cases included a previously healthy 45year-old woman who needed a liver transplant after taking a kava supplement for 8 weeks and a 14-year-old girl who needed a liver transplant after taking a supplement for 44 days. The Food and Drug Administration warned in 2002 of a possible link between kava and liver failure and has asked doctors nationwide to report such problems in kava users. Germany, where herbal supplements are regulated similarly to prescription drugs and must be licensed, was the first nation to ban kava in 2001. Switzerland, the United Kingdom, Singapore, Australia, Canada, Ireland, and France have banned kava or limited sales after linking it to liver failure and other maladies.79 Aegeline Shortly before this manuscript was submitted to the Editor, the FDA announced a recall of a recall and destruction of inventories of a dietary inked to dozens of cases of acute liver failure and hepatitis, including 1 death and illnesses so severe that several patients required liver transplants.83 In addition to the recall of certain OxyElite Pro products, USPLabs assured FDA officials that it will destroy warehouse stocks of the supplement, with a retail value of about $22 million. FDA will oversee the destruction of the product. FDA used new enforcement tools provided by the FDA Food Safety Modernization Act to act quickly in the face of a potential danger to public health. The supplement was advertised as an aid to losing weight and building muscles. FDA warned the company on October 11, 2013, that certain OxyElite Pro products and another supplement, VERSA-1, are considered adulterated because they contain a new dietary ingredient, aegeline, for which the company did not provide evidence of safety. Non-synthetic aegeline is an alkaloid extract from leaves of the Asian bael tree (Agele marmelos). Evidence of danger On September 13, 2013, FDA learned of a cluster of 7 Hawaii residents with acute liver failure/ non-viral hepatitis.

100


A joint investigation by the Hawaii Department of Health and CDC revealed that the patients all had consumed OxyElite Pro products. The FDA meanwhile identified patients outside of Hawaii with similar liver dysfunction after using OxyElite Pro. On October 11, 2013, the FDA warned the company that certain OxyElite Pro and VERSA-1 products were deemed adulterated and that failure to immediately cease distribution of both products could lead to enforcement actions. The FDA also outlined its findings of harm linked to OxyElite Pro. As of the end of October 2013, there were 56 cases of acute liver failure or acute hepatitis linked to OxyPro Elite, 43 of them in Hawaii. The investigation continues. While manufacturers of dietary supplements are not required to provide proof of safety and effectiveness prior to marketing, they are required to notify the FDA of plans to include a new dietary ingredient. They are also required to submit evidence that the dietary ingredient would reasonably be expected to be safe under the conditions of use recommended or suggested in the supplement labeling. A new dietary ingredient is defined as one not marketed in the United States before October 15, 1994. Companies are required to provide evidence of safety of the new dietary ingredient 75 days before the product goes to market. This notification was not made by USPLabs before it began using DMAA, a new dietary ingredient, in OxyElite Pro. The FDA was likewise not informed when the company, no longer formulating with DMAA, began using the new dietary ingredient aegeline. The FDA Food Safety Modernization Act has been instrumental in FDA’s enforcement actions regarding the OxyElite Pro dietary supplements. Antiviral therapy for HIV-AIDS While not primarily responsible for this class of drug treatment in these patients, the primary care physician will, however, be managing the general care of the AIDS patient. Many antiviral medications are associated with hepatotoxicity, and therefore, the awareness of this risk is important. Review of a current microbiology review reference demonstrates that most antivirals used in the treatment of HIV-AIDS are associated with a variety of DILDs, from minor elevations in transaminases to full-blown hepatitis.78

Summary and conclusions Conclusions and future perspectives Non-narcotic analgesics can result in a plethora of hepatic lesions, however, clinically significant liver damage is uncommon with normal doses. Hepatotoxicity that is caused by salicylates, NSAIDs, acetaminophen, and pyrazolones varies, but some of these drugs cause generalized liver reactions. The risks of liver injury may vary depending on factors such as age, sex, dosage, and treatment duration. Hepatotoxicity from salicylates and most NSAIDs has been frequently associated with females who have collagen diseases. Acetaminophen-induced liver dysfunction almost always is due to overdose. Other than the fatty changes in hepatocytes in Reye’s syndrome patients who are attributed to salicylate usage and the acute necrosis due to acetaminophen in overdose, the pathological changes in liver reactions to non-narcotic analgesics are variable. Liver damage has been reported with most NSAIDs and pyrazolone analgesics, but reliable patterns of hepatotoxicity are still largely undiscovered. Therefore, the degree of relative risk cannot be elucidated and the incidence connected to drug usage is not known. Other drugs associated with this group are glafenine, diclofenac, clometacin, sulindac, and pirprofen, which have high risk of hepatotoxicity but the mechanisms are unknown.42 Analysis of NSAID-induced idiosyncratic adverse reactions indicates that a number of acidic NSAIDs are connected to hepatocellular and cholestatic (rare) liver injury. The risk of


101

unpredictable liver injuries due to NSAIDs at therapeutic levels is minimal. Current NSAID usage has been associated with a 2-fold increase in the risk of liver injury and an additional risk of 5 per 100,000 person years. Fulminant hepatitis can progress in susceptible patients, and an intriguing task for the future is to pinpoint some of the risks that can be important in producing toxicity. For example, there is evidence that NSAID-induced liver injury is more prevalent in elderly patients compared to other age groups. It is unclear whether this is due to increased usage in elderly patients or a decrease in kidney function, which could enhance susceptibility through drug accumulation, or increased serum levels of drug glucuronides, which would lead to higher levels of a drug irreversibly binding to plasma proteins.43 The liver is a vital organ for life and is vital for metabolism of drugs and other substances consumed by patients. When the drugs consumed for a variety of therapeutic purposes cause injury to the patient, it creates a quandary for the therapist. Special monitoring and evaluation of continued benefit over risk may lead to selection of an alternative but less-toxic agent. Sometimes, the FDA and the manufacturer must make the difficult choice to not bring the product to market, restrict its use after release, or eventually withdraw the product from the market. Drug interactions Basically, the scientific identification of a hepatotoxic drug interaction is when there is an absence of hepatotoxicity when the 2 drugs are given separately and a presence of hepatotoxicity when they are given together. In many cases, this criterion is either unrealistic or unethical. Additionally, it is a well-established fact that the more drugs a patient receives, the higher the risk for adverse drug events; however, it is not known is this applies to DILD.2 Treatment In the event that there is an emergency overdose of acetaminophen, the use of Nacetylcysteine can prevent or limit liver dysfunction. Other than N-acetylcysteine there is no defined treatment for DILD but merely treatments for the symptoms depending on the level of liver impairment. The only avenue to ameliorate the poor liver prognosis of drug-induced endstage liver injury in certain cases is to conduct an emergency liver transplantation.2 Morbidity and mortality A mere 25% of all patients with acute idiosyncratic end-stage liver failure survive a few weeks with no liver transplant. The cause of death includes multi-organ failure, sepsis, cardiac arrest, cerebral edema, and respiratory failure. In the case of mild to moderate liver dysfunction, mortality is less common than severe morbidity. As expected, the patient prognosis becomes worse the longer the hepatotoxin is taken.3 Prevention The majority of DILD cases are reversible; however, there are very few prophylactic options available. Patients should be prescribed drugs based on their clinical need, utilizing a thorough risk-benefit analysis, considering their predisposing risk factors. A non-hepatotxin should be selected in favor of a hepatotoxin, if one is available.3 Ubiquitously accepted rules for drug hepatotoxicity prevention have only been established for a select few drugs. Methotrexate treatment of psoriasis and rheumatoid arthritis can cause dose-dependent steatosis and liver cirrhosis. In this case, the postulation of liver function is an unreliable means of predicting hepatotoxicity. The best strategy is to provide a weekly dosing regimen of 15 mg in conjunction with liver biopsies every 6 months.2 Manufacturers generally

102


provide recommended guidelines for tracking liver enzymes during treatment with known hepatotoxic agents such as methotrexate. The only therapy known to significantly ameliorate drug-induced hepatitis is N-acetylcysteine (NAC) for patients with acetaminophen (Tylenol) toxicity, which is typically given within 24 h of acetaminophen ingestion. The patients who do not respond to primary treatments or supplemental therapy are usually patients with fulminant (end stage) hepatic failure, and therefore require liver transplantation to survive. Sadly, only 50% of patients with DILD actually receive this transplantation.2 There is no evidence to show that, despite instructions and warnings on drug labels, routine monitoring of liver enzymes prevents clinically significant hepatotoxicity, most of which is unpredictable and quite uncommon. Thus, an argument can be made that a more effective and efficient method of detecting and preventing hepatotoxicity would involve vigilance on the part of the patients themselves in recognizing symptoms, followed by prompt medical evaluation. Admittedly, such an approach may not apply to all drugs.9 It is well-accepted to monitor patients who are at a special risk such as patients with alcoholic liver disease. Cost–benefit analysis does not absolve the continuous monitoring of serum transaminases in otherwise healthy individuals. The best method of severe hepatotoxicity prevention is to educate and alert clinicians to be mindful of the DILD diagnosis when a patient develops clinical signs of liver disease.2 References 1. Larrey D. Epidemiology and individual susceptibility to adverse drug reactions affecting the liver. Semin Liver Dis. 2002;22:145–155. 2. Dossing M, Sonne J. Drug-induced hepatic disorders. Drug Saf. 1993;9:441–449. 3. MacLaren R. Hepatic and cholestatic diseases. In: Tisdale JE, Miller DA, eds. Drug-Induced Diseases: Prevention, Detection, and Management. Bethesda, MD: American Society of Health System Pharmacists; 2005:515–535. 4. Stricker BHCH. Drug-Induced Hepatic Injury. 2nd ed. Amsterdam: Elsevier; 1992. 5. Farrell GC. Drug-induced acute hepatitis. In: Farrell GC, editor. Drug-Induced Liver Disease. Edinburgh: Churchill Livingstone; 1994:247–299. 6. Pessayre O, Larrey D, Biour M. Drug-induced liver injury. In: Bircher J, Benhamou JP, McIntyre N, Rizzetto M, Rodés J, eds. Oxford Textbook of Clinical Hepatology. 2nd ed. Oxford: Oxford University Press; 1999:1261–1315. 7. Zimmerman HJ. Drug-induced liver disease. In: Zimmerman HJ, editor. Hepatotoxicity: The Adverse Effects of Drugs and Other Chemicals on the Liver. 2nd ed. Philadelphia: Lippincott Williams & Wilkins; 1999:427–456. 8. Larrey D. Drug-induced liver diseases. J Hepatol. 2000;32:77–88. 9. Navarro VJ, Senior JR. Drug-related hepatotoxicity. N Engl J Med. 2006;354:731–737. 10. Biour M, Poupon R, Grange JD, Chazouilleres O. Hépatotoxicité des medicaments. Gastroenterol Clin Biol. 2000;11: 1052–1091. 11. Larrey D. Hepatotoxicity of herbal medicines. J Hepatol. 1997;26:47–51. 12. Schuppan D, Jia JD, Brinkhaus B, Hahn EG. Herbal products for liver diseases: a therapeutic challenge for the new millennium. Hepatology. 1999;30:1099–1104. 13. Seeff LB, Lindsay KL, Bacon BR, Kresina TF, Hoofnagle JH. Complementary and alternative medicine in chronic liver disease. Hepatology. 2001;34:595–603. 14. Mackay IR. Immune mechanisms and liver toxicity. In: Cameron RG, Feuer G, de la Inglesia FA, eds. Drug-Induced Hepatotoxicity. Berlin: Springer; 1996:221–247. 15. Bernal W, Wendon J. Acute Liver Failure. N Engl J Med. 2013;369:2525–2534. 16. Lee WM. Assessing causality in drug-induced liver injury. J Hepatol. 2000;33:1003–1005. 17. Larrey D. Hépatites médicamenteuses: aspects épidémiologiques, cliniques, diagnostiques, et physiopathologiques en 1995. Rev Med Int. 1995;16:752–758. 18. Lazarou J, Pomeranz BH, Corey PN. Incidence of adverse drug reactions in hospitalized patients: a meta-analysis of prospective studies. J Am Med Assoc. 1998;279:1200–1205. 19. Dossing M, Andreasen BP. Drug-induced liver disease in Denmark. An analysis of 572 cases of hepatotoxicity reported to the Danish board of adverse reaction to drugs. Scand J Gastroenterol. 1982;17:205–211. 20. Bernuau J, Benhamou JP. Fulminant and subfulminant liver failure. In: Bircher J, Benhamou JR, McIntyre N, Rizzetto M, Rodés J, eds. Oxford Textbook of Clinical Hepatology. 2nd ed. Oxford: Oxford University Press; 1999:1341–1372. 21. Jick H, Walker AM, Porter J. Drug-induced liver disease. J Clin Pharmacol. 1981;21:359–364. 22. Hallas J, Gram LF, Grodum E, et al. Drug related admissions to medical wards: a population based survey. Br J Clin Pharmacol. 1992;33:61–68. 23. Rubin E. Iatrogenic hepatic injury. Hum Pathol. 1980;11:312–331. 24. Pateria P, de Boer B, MacQuillan G. Liver abnormalities in drug and substance abusers. Best Pract Res Clin Gastroenterol. 2013;27(4):577–596. 25. Bunchorntavakul C, Reddy KR. Acetaminophen-related hepatotoxicity. Clin Liver Dis. 2013;17(4):587–607. 26. Chalasani N, Björnsson E. Risk factors for idiosyncratic drug-induced liver injury. Gastroenterology. 2010;138(7): 2246–2259.


103

27. Myers RP, Shaheen AA, Li B, Dean S, Quan H. Impact of liver disease, alcohol abuse, and unintentional ingestions on the outcomes of acetaminophen overdose. Clin Gastroenterol Hepatol. 2008;6(8):918–925. 28. Larson AM, Polson J, Fontana RJ, Davern TJ, Lalani E, Hynan LS, Reisch JS, Schiødt FV, Ostapowicz G, Shakil AO, Lee WM. Acute Liver Failure Study Group. Acetaminophen-induced acute liver failure: results of a United States multicenter, prospective study. Hepatology. 2005;42(6):1364–1372. 29. Mitchell JR, Thorgeirsson UP, Black M, et al. Increased incidence of isoniazid hepatitis in rapid acetylators: possible relation to hydranize metabolites. Clin Pharmacol Ther. 1975;18:70–79. 30. Ellard GA, Mitchison DA, Girling DJ, Nunn AJ, Fox W. The hepatic toxicity of isoniazid among rapid and slow acetylators of the drug. Am Rev Respir Dis. 1978;118:628–629. 31. Krähenbühl S. Mitochondria: important target for drug toxicity? J Hepatol. 2001;34:334–336. 32. Kanel GC, Korula J. Liver Biopsy Evaluation—Histologic Diagnosis and Clinical Correlations. Philadelphia: W.B. Saunders; 2000. 33. Black M, Mitchell JR, Zimmerman HJ, Ishak KG, Epler GR. Isoniazid-associated hepatitis in 114 patients. Gastroenterology. 1975;69:289–302. 34. Chitturi S, Farrell GC. Drug-induced cholestasis. Semin Gastrointest Dis. 2001;12:113–124. 35. Lewis JH. Drug-induced liver disease. Med Clin North Am. 2000;84:1275–1311. 36. Standardization of definitions and criteria of causality assessment of adverse drug reactions. Drug-induced liver disorders: report of an international consensus meeting. Int J Clin Pharmacol Ther Toxicol. 1990;28:317–322. 37. Minton NA, Henry JA, Frankel RJ. Fatal paracetamol poisoning in an epileptic. Hum Toxicol. 1988;7:33–34. 38. Jaeschke H, Gores GJ, Cederbaum AI, Hinson JA, Pessayre D, Lemasters JJ. Mechanisms of hepatotoxicity. Toxicol Sci. 2002;65:166–176. 39. Angulo P. Nonalcoholic fatty liver disease. N Engl J Med. 2002;346:1221–1231. 40. Marks DH. Case report: drug toxicity leading to vanishing bile duct syndrome and cholestatic jaundice. Internet J Gastroenterol. 2009;8:p6. 41. Degott C, Feldmann G, Larrey D, et al. Drug-induced prolonged cholestasis in adults: a histological semi quantitative study demonstrating progressive ductopenia. Hepatology. 1992;15:244–251. 42. Prescott LF. Liver damage with non-narcotic analgesics. Med Toxicol. 1986;1:44–56. 43. Boelsterli UA, Zimmerman HJ, Kretz-Rommel A. Idiosyncratic liver toxicity of nonsteroidal anti-inflammatory drugs: molecular mechanisms and pathology. Crit Rev Toxicol. 1995;25:207–235. 44. Nuki G. Pain control and the use of non-steroidal analgesic anti-inflammatory drugs. Br Med Bull. 1990;46:262–278. 45. Tarazi EM, Harter JG, Zimmerman HJ, Ishak KG, Eaton RA. Sulindac-associated hepatic injury: analysis of 91 cases reported to the Food and Drug Administration. Gastroenterology. 1993;104:569–574. 46. Lee WM. Drug induce hepatotoxicity. N Engl J Med. 2003;349:474–485. 47. Rapha Group Software, Inc. 1996. 48. Linden CH, Rumack BH. Acetaminophen overdose. Emerg Clin North Am. 1984;2:103–119. 49. Linden CH, Rumack BH. Acetaminophen poisoning. In: Tintinalli JE, Krome RL, Ruiz E, eds. Emergency Medicine: A Comprehensive Study Guide. 3rd ed. New York: McGraw-Hill; 1992:593–597. 50. Reynolds JEF. Martindale: The Extra Pharmacopoeia. 29th ed. London: The Pharmaceutical Press; 1989; p 32. 51. Facts and Comparisons Drug Information. St. Louis, MO: Facts and Comparisons; 1986. [Print]. 52. Prescott LF, Critchley JA, Balali-Mood M, Pentland B. Effects of microsomal enzyme induction on paracetamol metabolism in man. Br J Clin Pharmacol. 1981;12:149–153. 53. Wallace RJ Jr., Brown BA, Griffith DE. Drug intolerance to high-dose clarithromycin among elderly patients. Diagn Microbiol Infect Dis. 1993;16:215–221. 54. Brown BA, Wallace RJ Jr., Griffith DE, Girard W. Clarithromycin-induced hepatotoxicity. Clin Infect Dis. 1995;20: 1073–1074. 55. Shaheen N, Grimm IS. Fulminant hepatic failure associated with clarithromycin. Am J Gastroenterol. 1996;91: 394–395. 56. Hussar DA. They may call it an “express,” but it is on the wrong track (Call for Rezulin Market Withdrawal). Editorial. Pharmacy Today. October 1999, American Pharmaceutical Association. 57. vanDooren-Greebe RJ, Kuijpers AL, Buijs WC, et al. The value of dynamic hepatic scintigraphy and serum aminoterminal propeptide of type III procollagen for early detection of methotrexate-induced hepatic damage in psoriasis patients. Br J Dermatol. 1996;134:481–487. 58. Malatjalian DA, Ross JB, Williams CN, Colwell SJ, Eastwood BJ. Methotrexate hepatotoxicity in psoriatics: report of 104 patients from Nova Scotia, with analysis of risks from obesity, diabetes and alcohol consumption during long term follow-up. Can J Gastroenterol. 1996;10:369–375. 59. West SG. Methotrexate hepatotoxicity. Rheum Dis Clin North Am. 1997;23:883–915. 60. Erickson AR, Reddy V, Vogelgesang SA, West SG. Usefulness of the American College of Rheumatology recommendations for liver biopsy in methotrexate-treated rheumatoid arthritis patients. Arthritis Rheum. 1995;38: 1115–1119. 61. Tett SE, Triggs EJ. Use of methotrexate in older patients. A risk–benefit assessment. Drugs Aging. 1996;9: 458–471. 62. Kremer JM, Alarcón GS, Lightfoot RW Jr., et al. Methotrexate for rheumatoid arthritis. Suggested guidelines for monitoring liver toxicity. American College of Rheumatology. Arthritis Rheum. 1994;37:316–328. 63. Newman ED, Scott DW. The use of low-dose oral methotrexate in the treatment of polymyositis and dermatomyositis. J Clin Rheumatol. 1995;1:99–102. 64. Hakim AJ, Machin SJ, Isenberg DA. Autoimmune thrombocytopenia in primary antiphospholipid syndrome and systemic lupus erythematosus: the response to splenectomy. Semin Arthritis Rheum. 1998;28:20–25. 65. Montilla Morales C, López Longo FJ, Moreno Zazo M, Monteagudo Sáez I, Moreno García AC, Carreño Pérez L. Hypertransaminemia and methotrexate: not always a toxic effect? Rev Clin Esp. 1998;198:822–824.

104


66. Kremer JM, Furst DE, Weinblatt ME, Blotner SD. Significant changes in serum AST across hepatic histological biopsy grades: prospective analysis of 3 cohorts receiving methotrexate therapy for rheumatoid arthritis. J Rheumatol. 1996;23:459–461. 67. Fuchs S, Simon Z, Brezis M. Fatal hepatic failure associated with ciprofloxacin. Lancet. 1994;343:738–739. 68. Grassmick BK, Lehr VT, Sundareson AS. Fulminant hepatic failure possibly related to ciprofloxacin. Ann Pharmacother. 1992;26:636–639. 69. Samarasinghe DA, Farrell GC. Drug-induced liver disease. In: Gitnick G, editor. Current Hepatology, 14; 1994:117–167. 70. Sheikh NM, Philen RM, Love LA. Chaparral-associated hepatotoxicity. Arch Intern Med. 1997;157:913–919. 71. Boelesteri UA, Zimmerman HJ, Kretz-Rommel A. Idiosyncratic liver toxicity of nonsteroidal anti-inflammatory drugs: molecular mechanisms and pathology. Crit Rev Toxicol. 1995;25:207–235. 72. Rumack B. Acetaminophen overdose in children and adolescents. Pediatr Clin North Am. 1986;33:691. 73. Rumack B. Acetaminophen overdose in young children. Am J Dis Child. 1984;138:428. 74. Watkins PB, Kaplowitz N, Slattery JT, et al. Aminotransferase elevations in healthy adults receiving 4 grams of acetaminophen daily: a randomized controlled trial. J Am Med Assoc. 2006;296(1):87–93. 75. Nunez M. Hepatotoxicity of antiretrovirals: incidence, mechanisms and management. J Hepatol. 2006;44(suppl 1): S132–S139. 76. Heil EL, Townsend ML, Shipp K, Clarke A, Johnson MD. Incidence of Severe Hepatotoxicity Related to Antiretroviral Therapy in HIV/HCV Coinfected Patients. AIDS Research and Treatment, vol. 2010, AIDS Research and Treatment. Article ID 856542. 77. Marks DH. Evaluation of medical causation. In: O’Donnell JT, editor. Drug Injury: Liability, Analysis and Prevention. 3rd ed. Tucson: L&J Publications; 2012. 78. Lurain NS, Thompson KD. Antiviral agents. In: Versalovic J, Carroll KC, Funke G, Jorgensen JH, Landry ML, Warnock, eds. Manual of Clinical Microbiology. 10th ed.. Washington DC: ASM Science; 2011. 79. 〈http://www.cdc.gov/mmwr/preview/mmwrhtml/mm5147a1.htm〉 Accessed 09.11.13. 80. Moellering RC. Hepatotoxicity of antimicrobials: the trovafloxacin story. Program and Abstracts of the 40th Interscience Conference on Antimicrobial Agents and Chemotherapy. Toronto, Ontario, Canada; 2000. Abstract 1880. 81. O’Donnell JT. Liability, litigation, and lessons in new drug development (Chapter 13). In: O’Donnell JT, editor. Drug Injury: Liability, Analysis, and Prevention. 3rd ed.. Tucson, AZ: Lawyers & Judges Publishers; 2012. 82. Marks DH, O’Donnell JT. Drug induced liver disease (Chapter 13). In: O’Donnell JT, editor. Drug Injury: Liability, Analysis, and Prevention. 3rd ed.. Tucson, AZ: Lawyers & Judges Publishers; 2012. 83. 〈http://www.fda.gov/Food/RecallsOutbreaksEmergencies/Outbreaks/ucm370849.htm〉 Accessed 22.11.13.

Glaucoma. A primer for primary care physicians.

Travel medicine for the primary care physician.

The new primary-care physician.

Viscosupplementation for osteoarthritis: a primer for primary care physicians.

Novel Anticoagulants in Atrial Fibrillation: A Primer for the Primary Physician.

Advanced therapy for inflammatory bowel disease: a guide for the primary care physician.

Physician associates in primary care.

Physician management in primary care.

Solutions to the primary care physician shortage.

Bariatric Surgery: A Brief Primer for Primary Care Physicians.

The gyn oncologist as primary care physician.

Ambulatory opiate detoxification and primary care: a role for the primary care physician.

Neonatal endocrine emergencies: a primer for the emergency physician.

Health reform and physician-led accountable care: the paradox of primary care physician leadership.

Primary care physician smoking screening and counseling for patients with chronic disease.

Dietary counselling for cardiovascular disease prevention in primary care settings: results from a German physician survey.

Primary care physician-led health reform--reply.

Physician-patient expectations in primary care.

Primary care physician-led health reform.

Physician Acceptance of a Physician-Pharmacist Collaborative Treatment Model for Hypertension Management in Primary Care.

Primary care physician insights into a typology of the complex patient in primary care.

nephrologist partnership in treating chronic kidney disease.

State-of-the-art anterior cruciate ligament tears: A primer for primary care physicians.

Pre-exercise screening: role of the primary care physician.